Caspase cleavage of the APC tumor suppressor and release ... - Nature

1 downloads 0 Views 306KB Size Report
Feb 5, 2007 - Caspase cleavage of the APC tumor suppressor and release of an amino-terminal domain is required for the transcription-independent function ...
Oncogene (2007) 26, 4872–4876

& 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc

SHORT COMMUNICATION

Caspase cleavage of the APC tumor suppressor and release of an amino-terminal domain is required for the transcription-independent function of APC in apoptosis J Qian1,2, K Steigerwald1, KA Combs1, MC Barton3 and J Groden2 1

Department of Molecular Genetics, Biochemistry and Microbiology, University of Cincinnati College of Medicine, Cincinnati, OH, USA; 2Department of Molecular Virology, Immunology and Medical Genetics, The Ohio State University College of Medicine, Columbus, OH, USA and 3Department of Biochemistry and Molecular Biology, University of Texas MD Anderson Cancer Center, Houston, TX, USA

The adenomatous polyposis coli (APC) tumor suppressor is inactivated by mutation in most colorectal tumors. APC is a component of the Wnt signaling pathway and is best known for its ability to downregulate b-catenin and consequent effects on transcriptional regulation. Previous work demonstrated that APC accelerates apoptosis-associated caspase activity independently of transcription, and suggested novel tumor suppressor functions of APC. In this work, we have mapped the APC apoptosis-accelerating region to amino acids (aa) 1–760 by testing a series of nonoverlapping APC segments. Interestingly, this segment corresponds to a stable group II caspase cleavage product of APC released during apoptosis that includes the aminoterminal aa1–777. Mutation of the APC aspartic acid residue at position 777 to an alanine completely abolished in vitro cleavage of APC by a recombinant group II caspase and rendered the full-length protein unable to accelerate apoptosis in vitro. A truncated APC protein associated with familial and sporadic colorectal cancer, also unable to accelerate apoptosis in vitro and in vivo, is resistant to group II caspase cleavage. These results demonstrate that cleavage of APC and the subsequent release of an aminoterminal segment are necessary for the transcriptionindependent mechanism of APC-mediated apoptosis. Oncogene (2007) 26, 4872–4876; doi:10.1038/sj.onc.1210265; published online 5 February 2007 Keywords: APC; apoptosis; transcription-independent; caspase; colon cancer

Introduction Mutation of APC gene is a rate-limiting event in the development of most colorectal tumors and is associated Correspondence: Dr J Groden, Department of Molecular Virology, Immunology and Medical Genetics, The Ohio State University College of Medicine, 400 W 12th Street, Columbus, OH 43210-2207, USA. E-mail: [email protected] Received 20 June 2006; revised 31 October 2006; accepted 17 November 2006; published online 5 February 2007

with dysregulation of several physiologic processes (Goss and Groden, 2000). Several experimental lines of evidence suggest a role for APC in the regulation of cell death, including our previous work that used an in vitro assay to separate transcription-independent from transcription-dependent functions of APC in apoptosis (Steigerwald et al., 2005). This assay uses a cell-free Xenopus laevis egg extract to monitor the time course of effector caspase cleavage of a colorimetric substrate; exogenous proteins can be added or subtracted to change this time course. This assay is qualitatively reproducible although the time at which caspase is activated varies from extract to extract and makes it impossible to average the results from separate experiments in a meaningful way (Fasschon et al., 1997). It has been used successfully by a number of laboratories to define the roles of several apoptotic mediators (Newmeyer et al., 1994; Fasschon et al., 1997). In our previous experiments, full-length APC (APC FL) protein accelerated the apoptotic time course, whereas mutant or truncated APC protein was unable to do so (Steigerwald et al., 2005). In this study, we further investigate the transcription-independent apoptosis-associated function of APC. In order to map the in vitro apoptosis-accelerating region of APC, we generated a series of non-overlapping recombinant APC segments, which together comprise the full-length protein. His-tagged APC FL and four His-tagged non-overlapping segments were expressed in yeast and purified using nickel chromatography (Figure 1a and b). Each of the APC proteins shown in Figure 1b was tested for its effects on the in vitro apoptotic time course. Addition of 1 mM purified recombinant aminoterminal APC (aa1–760) protein, as APC FL, accelerates the required time course for caspase activity, whereas other recombinant APC segments, aa761– 1339, aa1340–2140 and aa2141–2843, do not have this effect (Figure 1c). The effect of APC aa1–760 on apoptotic events is independent of transcription, as similarly prepared Xenopus egg extracts are quiescent for RNA polymerase II-dependent transcription (Barton et al., 1993), and independent of translation, as cyclohexamide is added during the preparation of

APC N terminus promotes transcription-independent apoptosis J Qian et al

4873

Figure 1 The amino-terminal APC segment accelerates apoptosis-associated caspase activity in vitro. (a) Schematic diagram of five APC segments containing unique amino acids and functional domains. The yeast expression vectors encoding His-tagged APC segments were generated by cloning cDNAs into pYES2/NT vector (Invitrogen, Carlsbad, CA, USA) using BamHI and NotI. cDNAs for each APC segment were generated by polymerase chain reaction (PCR) using the high fidelity polymerase Pwo (Roche Applied Sciences, Indianapolis, IN, USA). Primers used to generate APC segments were listed in Supplementary Table 1. MCR represents the mutation cluster region. (b) Western blots of APC FL and APC segments purified from a yeast expression system. Recombinant proteins were expressed using a yeast expression system as before (Lillard-Wetherell et al., 2004). APC FL was separated by 3% agarose gel electrophoresis and identified using a monoclonal antibody specific for the carboxyl-terminal of APC (Ab-2). APC segments and lacZ protein were separated by 6% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and identified using a polyclonal anti-His antibody. EV represents protein preparations made from yeast transformed with an empty vector. (c) Representative results from an in vitro apoptosis assay (typical of at least three experiments). Xenopus laevis egg extract for in vitro apoptosis studies was prepared and reconstituted as described previously (Steigerwald et al., 2005). Apoptosis is measured by cleavage of a colorimetric substrate, Z-DEVD-pNA, by activated caspase 3 and is displayed as an increase in absorbance at 405 nm versus time in hours. The assay uses Xenopus egg cytoplasmic extract and a membrane fraction titrated to induce caspase activity at 4–6 h. All protein samples were added to a final concentration of 1 mM. (d) Expression and purification of the carboxyl-terminal fragment of APC (aa761–2843). Purified protein was separated by 6% SDS–PAGE and stained with Coomassie bright blue. The arrow identifies this APC segment. (e) Representative results from an in vitro apoptosis assay (typical of at least three experiments) comparing the effects of the amino-terminal of APC (aa1–760) and the carboxyl-terminal of APC (aa761–2843) on the caspase activity time course.

extracts. To exclude the possibility that the larger carboxy-terminal segment of APC (aa761–2843) has the same function as the amino-terminal APC (aa1–760) segment although the non-overlapping segments do not, a His-tagged carboxy-terminal fragment of APC was purified from yeast (Figure 1d) and tested as described above. Figure 1e shows that the carboxy-terminal segment of APC (aa761–2843) is unable to accelerate an apoptotic time course. To validate these in vitro results, the colon cancer cell line SW480 was transiently transfected with a similar

series of mammalian expression vectors encoding nonoverlapping APC segments tagged with green florescent protein (GFP) and the percentage of apoptosis in the GFP-positive cell population was determined. All the GFP-tagged APC segments were expressed in SW480 (Supplementary Figure 1). Forty-eight hours following transfection, the percentage of apoptotic cells transfected with the amino-terminal APC segment (aa1–760) significantly increased compared to that of cells transfected with GFP vector only (Figure 2a), whereas the percentage of apoptotic cells transfected with other vectors, Oncogene

APC N terminus promotes transcription-independent apoptosis J Qian et al

4874

APC aa761–1339, APC aa1340–2140 and APC aa2141– 2843, did not increase. SW480 cells contain mutant APC that is unable to bind and downregulate b-catenin (Korinek et al., 1997). To eliminate the possibility that the APC segments used in the cell culture experiment accelerate apoptosis by downregulating b-catenin in the Wnt signaling pathway and altering transcription, GFPtagged non-overlapping APC segments were introduced into SW480 cells together with TOPFLASH-luc, a bcatenin-sensitive transcriptional reporter plasmid containing TCF/LEF-1-binding sites. None of the APC segments altered LEF-1-dependent transactivation, as each plasmid generated results not significantly different

to mock-transfected cells. Dominant-negative LEF-1 (DLEF-1) was used as a positive control to reduce TOPFLASH-luc transactivation activity (Figure 2b). These cell culture experiments confirm that the apoptosis-accelerating function of APC resides in the aminoterminal APC segment, and that this effect is independent of b-catenin-mediated transcription. An amino-terminal segment of APC that includes aa1–777 is released as a stable product during apoptosis in vivo by caspase cleavage of APC (Browne et al., 1998; Webb et al., 1999). Group II caspases, such as caspase 3, cleave APC immediately following a conserved motif

Figure 2 The amino-terminal segment of the APC tumor suppressor increases apoptosis transcription-independently in colon cancer cell lines. (a) The amino-terminal segment of APC increases apoptosis in SW480 cells. SW480 colorectal carcinoma cells were transiently transfected with GFP-tagged APC vectors. After 24, 48 or 72 h following transfection, the percentage of apoptotic cells within the GFP-positive cell population was determined by flow cytometry using the M30 CytoDEATH antibody (Roche) that specifically recognizes the caspase-cleaved product of cytokeratin. Camptothecin (CAM) (25 nM) was used as a positive control to induce apoptosis. (b) After transient cotransfection with GFP-tagged APC vectors and dual luciferase reporters, SW480 cells were lysed and luciferase activity measured. Transcription activation mediated by b-catenin was assayed using the TOPFLASH system (Promega, Madison, WI, USA). Cell culture experiments were carried out in triplicate and the results expressed as the mean 7s.d. from three separate experiments. Asterisks denote significant differences (Po0.05) versus GFP only.

Figure 3 Cleavage of APC by caspase 3 at aa777 is necessary for accelerating apoptosis-associated caspase activity in vitro. (a) Mutant APC cDNA in the pYES2/NT vector was generated using the PCR-based QuickChange site mutagenesis kit (Stratagene, La Jolla, CA, USA). APC proteins were expressed and purified from yeast. Purified proteins were separated by 6% SDS– PAGE and stained with Coomassie bright blue. The arrow identifies APC proteins. Two different amounts of each sample were loaded. (b) APC FL and APC D777A were incubated with recombinant caspase 3 or caspase 8 for 2 h at 371C. The proteolytic products were separated by 6% SDS–PAGE and identified by Western blotting with an anti-His antibody. Caspase 8 digestion releases a protein of approximately 150 kDa (indicated by arrow) from both APC proteins. Caspase 3 cleavage releases a protein segment from APC FL that is approximately 90 kDa (indicated by arrow). (c) Representative results from an in vitro apoptosis assay (typical of at least three experiments) comparing the effects of the wild-type APC and APC D777A on the apoptotic time course.

Oncogene

APC N terminus promotes transcription-independent apoptosis J Qian et al

4875

DNID that is located adjacent to the armadillo repeat domain (Webb et al., 1999). As the amino-terminal 777 amino acids of APC released by group II caspase cleavage roughly correspond to the same amino-terminal segment identified in our experiments as that responsible for promoting apoptotic activity, we next tested whether APC cleavage by a group II caspase is necessary for accelerating apoptosis. The aspartic acid residue at

position 777 in APC was changed to an alanine by site-directed mutagenesis, and the recombinant mutant protein (APC D777A) was expressed and purified from yeast (Figure 3a). Incubation of wild-type APC FL with caspase 3 resulted in the release of an approximately 90 kDa protein, whereas mutation of Asp777 to alanine completely abolishes cleavage (Figure 3b). Addition of the caspase 3-cleavage-resistant APC to our in vitro apoptosis assay did not accelerate the apoptotic time course in comparison to wild-type APC (Figure 3c). These data suggest that cleavage of APC by a group II caspase is required for the transcriptionindependent function of APC in apoptosis. A frequently observed truncation of the APC protein results from a 5-bp deletion in APC at codon 1309 and is associated with both familial APC and sporadic colorectal cancer, respectively (Groden et al., 1993). As most truncated mutants of APC include the amino-terminal 777 amino acids released by group II caspase cleavage, we tested whether APC 1309, a mutant version of APC, can be cleaved by a group II caspase. Following incubation with caspase 3, APC 1309 and normal APC are cleaved to produce approximately 90 kDa aminoterminal fragments (Figure 4a). However, the 90 kDa band released from cleavage of APC 1309 by caspase 3 is only one-fifth of that released from wild-type APC, as measured by densitometry. This result suggests that the cleavage of truncated APC by caspase 3 is less efficient than wild-type APC. Our previous work has shown that APC 1309 delays apoptosis in an in vitro apoptosis assay using Xenopus Figure 4 The truncated mutant of APC (APC 1309) is resistant to caspase 3 cleavage and unable to increase apoptosis in colon cancer cell lines. APC FL and APC 1309 were incubated with recombinant caspases 3 or 8 for 2 h at 371C. The proteolytic products were separated by 6% SDS–PAGE and identified by Western blotting with an anti-His antibody. Caspase 8 digestion releases a protein segment at approximately 150 kDa (indicated by arrow) from wild-type APC protein, which is close in size to mutant APC (APC 1309). Caspase 3 cleavage releases a protein segment at approximately 90 kDa in both APC FL (indicated by arrow) and APC 1309. The 90 kDa protein released from digestion of APC 1309 is significantly less than that released from wild-type APC. (b and c) SW480 (b) and HCT116 (c) colorectal carcinoma cells were transiently transfected with GFPtagged APC segments. After 24, 48 or 96 h following transfection, the percentage of apoptotic cells within the adherent GFP-positive cell population was determined by immunofluorescence microscopy using the M30 CytoDEATH antibody that specifically recognizes the caspase-cleaved product of cytokeratin. Apoptotic cells were observed and scored for the number of CytoDeath-positive cells within the GFP-positive cell population (n>200/experiment). Results are expressed as the mean 7s.d. from three separate experiments. Asterisks denote significant differences (Po0.05) in comparison to results with GFP-tagged APC 761–1339. (d) APC 1–777 protein level increases in floating HCT116 cells in a time-dependent manner. Expression of GFP-tagged APC in floating cells was dissolved with 4–20% gradient gel (Invitrogen) and detected with an anti-GFP antibody. Apoptosis in these floating cells was evaluated with an anti-caspase 3 antibody that identifies both pro-caspase (p34) and activated caspase subunit (p20), and with an anti-PARP antibody that identifies full-length PARP (p116) and its cleaved product (p90). Equal loading of the different samples was evaluated with an anti-actin antibody. (e) All three GFPtagged APC segments were expressed in adherent HCT116 cells following transfections. Expression of GFP-tagged proteins in adherent cells was evaluated with an anti-GFP antibody. Oncogene

APC N terminus promotes transcription-independent apoptosis J Qian et al

4876

egg extract (Steigerwald et al., 2005). To validate the in vitro result, the colon cancer cell lines SW480 and HCT116 were transiently transfected with mammalian expression vectors encoding APC aa1–777, APC aa761– 1339 or APC1309 tagged with GFP. Forty-eight hours following transfection, the percentage of apoptotic cells transfected with the amino-terminal APC segment (aa1– 777) significantly increased compared to those cells transfected with APC aa761–1339. The percentage of apoptotic cells transfected with APC 1309 did not significantly increase (Figure 4b and c). HCT 116 cells, derived from a human colon carcinoma, contain only APC FL but mutant b-catenin. HCT 116 cells attached to the flask (adherent cells) have a low frequency of apoptosis, whereas cells detached from the flask and floating in the medium have a high proportion of apoptotic cells (95%) (Browne et al., 1998; Webb et al., 1999). To determine whether transfection of APC induces HCT 116 cells to detach and float, we collected the floating cells following transfection at each of the indicated times. In floating cells, pro-caspase 3 (p34) is cleaved to generate activated caspase (p20), whereas poly(ADP-ribose) polymerase (PARP), a 116 kDa caspase substrate, is cleaved to generate a fragment of approximately 90 kDa (Figure 4d). Consistent with the apoptotic pattern of adherent GFP-positive cell determined by using M30, the GFP-tagged APC amino terminus (APC aa1–777) increased in the floating cells in a time-dependent manner. Neither the GFP-tagged APC aa761–1339 nor the GFP-tagged APC 1309 was detected in floating cells (Figure 4d). All three GFP-tagged APC proteins were detected in adherent HCT 116 cells

following transfection (Figure 4e). These cell culture studies demonstrated that APC 1309 is not associated with the acceleration of apoptosis, whereas the aminoterminal 777 amino acids of APC is associated with the acceleration of apoptosis. Our data show that that cleavage of APC and the subsequent release of an amino-terminal segment are necessary for the transcription-independent mechanism of APC-mediated apoptosis. This amino-terminal segment contains a homodimerization domain and the armadillo repeat region of the protein that mediates interactions with other proteins relevant to growth and signaling. For example, the multiple armadillo repeat domain within b-catenin is essential for the transcription-independent induction of apoptosis (Kim et al., 2000). Thus, it may be that the transcription-independent apoptotic effects of APC depend on protein/protein interactions with other pro- or anti-apoptotic protein(s) that interact with the conserved armadillo repeats. Conformational changes of mutant APC may obscure the caspase cleavage site(s) required for release of aa1– 777, or another latent functional domain, responsible for accelerating apoptosis. We are currently characterizing the subcellular localization of the APC segments released by group II caspase cleavage and identifying the associated or colocalized protein(s) to provide a clearer understanding of the mechanism of APC-mediated transcription-independent apoptosis. Acknowledgements This work was supported by NIH CA 63517 (JG).

References Barton MC, Madani N, Emerson BM. (1993). The erythroid protein cGATA-1 functions with a stage-specific factor to activate transcription of chromatin-assembled beta-globin genes. Genes Dev 7: 1796–1809. Browne SJ, MacFarlane M, Cohen GM, Paraskeva C. (1998). The adenomatous polyposis coli protein and retinoblastoma protein are cleaved early in apoptosis and are potential substrates for caspases. Cell Death Diff 5: 206–213. Fasschon DM, Couture C, Mustelin T, Newmeyer DD. (1997). Temporal phases in apoptosis defined by the actions of Src homology 2 domains, ceramide, Bcl-2, interlukin-1b converting enzyme family protease, and a dense membrane fraction. J Cell Biol 137: 1117–1125. Goss KH, Groden J. (2000). Biology of the adenomatous polyposis coli tumor suppressor. J Clin Oncol 18: 1967– 1979. Groden J, Gelbert L, Thliveris A, Nelson L, Robertson M, Joslyn G et al. (1993). Mutational analysis of patients with adenomatous polyposis: identical inactivating mutations in unrelated individuals. Am J Hum Genet 52: 263–272. Kim K, Pang KM, Evans M, Hay ED. (2000). Overexpression of b-catenin induces apoptosis independent of its transacti-

vation function with LEF-1 or the involvement of major G1 cell cycle regulators. Mol Biol Cell 11: 3509–3523. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Kinzler KW et al. (1997). Constitutive transcriptional activation by a beta-catenin-Tcf complex in APC / colon carcinoma. Science 275: 1784–1787. Lillard-Wetherell K, Machwe A, Langland GT, Combs KA, Behbehani GK, Schonberg SA et al. (2004). Association and regulation of the BLM helicase by the telomere proteins TRF1 and TRF2. Hum Mol Genet 13: 1919–1932. Newmeyer DD, Farschon DM, Reed JC. (1994). Cell-free apoptosis in Xenopus egg extracts: inhibition by Bcl-2 and requirement for an organelle fraction enriched in mitochondria. Cell 79: 453–464. Steigerwald K, Behbehani GK, Combs KA, Barton M, Groden J. (2005). The APC tumor suppressor promotes transcription-independent apoptosis in vitro. Mol Cancer Res 3: 78–89. Webb SJ, Nicholson D, Bubb VJ, Wyllie AH. (1999). Caspasemediated cleavage of APC results in an amino-terminal fragment with an intact armadillo repeat domain. FASEB 13: 339–346.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc).

Oncogene