- Email: [email protected]
Purification of catalytically active caspase-12 and its biochemical characterization
Archives of Biochemistry and Biophysics 502 (2010) 68–73
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi
Purification of catalytically active caspase-12 and its biochemical characterization Hyun-Jung Lee a,1, Sung Haeng Lee b,1, Sung-Hee Park a, Md. Golam Sharoar c, Song Yub Shin b,c, Jung Sup Lee d, Byungyun Cho e, Il-Seon Park b,c,* a
Department of Biology, Ewha Womans University, Seoul, Republic of Korea Department of Cellular and Molecular Medicine, School of Medicine, Chosun University, Gwangju 501-759, Republic of Korea c Department of Bio-Materials Engineering, Chosun University, Gwangju, Republic of Korea d Department of Biotechnology, College of Natural Sciences, Chosun University, Gwangju, Republic of Korea e Department of Chemistry and Medicinal Chemistry, Yonsei University, Wonju, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 5 April 2010 and in revised form 7 July 2010 Available online 18 July 2010 Keywords: Caspase-12 Caspase-7 m-Calpain Apoptosis Purification
a b s t r a c t Caspase-12, mainly detected in endoplasmic reticulum (ER), has been suggested to play a role in ER-mediated apoptosis and inflammatory caspase activation pathway. Cleavage of the prodomain by caspase-3/7 at the carboxyl terminus of Asp94 or m-calpain at the carboxyl terminus of Lys158 was reported to be a part of caspase-12-involved apoptosis. We biochemically characterized the prodomain-free forms of caspase-12 and the equivalent enzymes; Dpro1(G95-D419), rev-Dpro1[(T319-N419)-(G95-D318), a reverse form of Dpro1] and rev-Dpro2[(T319-N419)-(T159-D318)]. The three variants showed comparable activities which were dependent on salt concentration and pH. Auto-proteolytic cleavage was observed at two sites (carboxyl termini of Asp318 and Asp320) in Dpro1. Constitutively active forms of caspase-12 (rev-Dpro1 and rev-Dpro2) could induce cell death in cells transfected with the corresponding expression vectors, but no cleavage of caspase-3, DFF45 or Bid was observed, indicating caspase-12 may mediate a distinct apoptotic pathway rather than caspase-8 or -9-mediated cell death. Ó 2010 Elsevier Inc. All rights reserved.
Introduction Caspases are crucial enzymes for apoptosis that is essential for maintaining homeostasis in multicellular organisms [1–3]. Among the 15 known caspases, caspase-8 is known to be involved in receptor-mediated apoptosis in which the enzyme is activated upon formation of death-inducing signaling complex (DISC)2 with Fas or TNF receptor [4]. Caspase-9, on the other hand, forms a protein complex called apoptosome with Apaf-1, cytochrome c and dATP, which plays a critical role in mitochondria-mediated apoptosis [5]. As a result of the both processes, the executioner caspase-3 and -7 are activated to cleave their substrates such as DFF45 [4,5]. The third class of apoptotic pathway has been reported to be induced by stimuli to endoplasmic reticulum (ER) [6]. Caspase-12 that shares a high homology with caspase-1, -4, -5 and -11 [6–9] is a main constituent of the ER-mediated apoptosis. Cytotoxic agents that induce the activation of the enzyme include amyloid b (Ab) [6], ERdamaging agents such as thapsigargin (inhibitor of the intracellular calcium–ATP transport) and brefeldin A (inhibitor of ER–Golgi trans* Corresponding author at: Department of Cellular and Molecular Medicine, School of Medicine, Chosun University, Gwangju 501-759, Republic of Korea. Fax: +82 62 227 8345. E-mail address: [email protected] (I.-S. Park). 1 These authors contributed equally to this work. 2 Abbreviations used: ER, endoplasmic reticulum; DISC, death-inducing signaling complex; DMEM, Dulbecco’s Modified Eagle Media; FBS, fetal bovine serum. 0003-9861/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2010.07.013
port) [7], and respiratory syncytial virus [10]. Moreover, it is also related to the cytokine processing group-I caspases [11]. Previously, caspase-7 was shown to cleave procaspase-12 at the carboxyl termini of Asp94 and Asp341 [12]. The report also suggested that Asp341 was the self-cleavage site. A recent report, however, indicated a single auto-proteolytic cleavage after Asp318 [13]. Caspase-3 which shares the substrate specificity with caspase-7 also processed procaspase-12 after Asp94 [14,15]. Moreover, m-calpain was also shown to cleave procaspase-12 to generate the active enzyme upon ER-stress by agents such as A23187 (calcium ionophore), thapsigargin and brefeldin A [6,7], wherein the cleavage occurred at the carboxyl termini of Lys158 [7]. On the other hand, substrates of caspase-12 have not been reported, although it was proposed to activate caspase-9 in a cytochrome c-independent manner [16]. The current study was performed to provide biochemical background for the caspase-12-involved cellular processes. One obstacle for the investigation was the absence of purified caspase-12. In the previous studies, crude preparation of caspase-12 in bacterial cell lysates was often used rather than purified enzyme [7,13]. Our initial effort to purify active caspase-12 was not successful because the protein was primarily localized in the inclusion bodies and the solubilization did not produce active enzyme, which might be reason why the enzyme in Escherichia coli cell lysate was used in the previous study [13]. Thus, we initially developed an optimal purification procedure for active caspase-12 and its variants. Through
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73
the subsequent characterization using the purified enzymes, we show that activity of caspase-12 depends on salt concentration and pH like other caspases [17] and the auto-proteolytic cleavage between the large and small domains is essential for the activation. Furthermore, we suggest that caspase-12 may function in a distinct cell death pathway rather than being involved in the pathway mediated by caspase-8 or -9. Materials and methods Materials Dulbecco’s Modified Eagle Media (DMEM), fetal bovine serum (FBS), penicillin and streptomycin were purchased from Life Technologies, Inc. (Gaithersburg, MD). Caspase-3 and DFF45 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and Transduction Laboratories (Lexington, KY), respectively. Polyclonal antibody against Bid was generated by injecting the purified protein into mouse. All other materials were purchased from Sigma. Construction of recombinant caspase-12 To purify active caspase-12, we tested 5 different recombinant caspase-12 constructs including 3 reverse forms (rev-) in which the small subunit (T319-D419) is followed by the large subunit. The recombinant enzymes were the full length (A2-N419), Dpro1 (G95-N419), Dpro2 (T159-N419), rev-full length [(T319-N419)(A2-D318)], rev-Dpro1 [(T319-N419)-(G95-D318)] and rev-Dpro2 [(T319-N419)-(T159-D318)] (see Fig. 1 except rev-full length). For the construction of caspase-12 full length, Dpro1 and Dpro2, polymerase chain reaction (PCR) were performed using pCAGGSmCASP12 plasmid DNA [kindly provided by Mrs. Martine Vanhoucke, BCCM/LMBP plasmid collection (http://www.belspo.be/ bccm/lmbp.html)] as a template with the following primers: full length, TACGGATCCGGCGGCCAGGAGGACACAT (sense), GGCCTCG
69
AGATTCCCGGGAAAAAGGTA (antisense); Dpro1, CGCGGATCCAGG ACCTCAGAAGATATGT (sense), GGCCTCGAGATTCCCGGGAAAAAG GTA (antisense); Dpro2, CGCGGATCCAACAGAAAGGGCAAAAGAG (sense), GGCCTCGAGATTCCCGGGAAAAAGGTA (antisense). The PCR products were cloned in pET28b(-) or pET21b(-). For the construction of the rev-full length and rev-Dpro2, the small subunit was amplified with the primers, GGAATTCCATATGACAGATGAG GAACGTGTG (sense), TACGGATCCATT CCCGGGAAAAAGGTA (antisense) and the PCR product was cloned in the NdeI and BamHI sites of pET21b(-). Next, the large subunits of rev-full length (Ala2Asp318) and rev-Dpro2 (Thr159-Asp318) were amplified with the following primers: (Ala2-Asp318), TACGGATCCGATGAAGTT GATGGTGCGGCCAGGAGG (sense), CCGCTCGAGATCAGCAGTGGCT ATCCC (antisense); (Thr159-Asp318), TACGGATCCGATGAAGTTGAT GGTACAGAAAGGGCAAAAGAG (sense), CCGCTCGAGATCAGCAGTG GCTATCCC (antisense). The PCR products were inserted after the small subunit into the BamHI and XhoI sites. The rev-Dpro1 was generated in a similar way. The small subunit was amplified with the following primers: GGAATTCCATATGACAGATGAGGAACGTGTG (sense), CGCGGATCC AAACCATCAGCAGTGGCATTCCCGGGAAAAAG GTA (antisense). The PCR product was cloned in pET21b(-). (Gly95-Asp318) was amplified with CGCGGATCCAGGACCTCAGAA GATATGT (sense), (Ala2-Asp318) antisense. The PCR product was cloned after the small subunit. Several site-directed mutants of procaspase-12 (D24E/D94E, D94E, D115E, D318E, D320E and D318E/D320E) and Dpro1 C298S were constructed using overlap PCR techniques [18,19]. The PCR products were cloned in pET28b (-) or pET21b(-). Purification of recombinant caspase-12 and caspase-7 The Dpro1, rev-Dpro1, rev-Dpro2, Dpro1 D318E/D320E and Dpro1 C298S cloned in pET21b(-) were overexpressed in E. coli BL21(DE3)pLysS with 0.4 mM isopropyl-b-D-thiogalactopyranoside at 30 °C for 3 h. The proteins were purified using a Ni-NTA column
Fig. 1. (A) Schematic diagrams of caspase-12 and its variant constructs used in this study. The putative cleavage sites of procaspase-12 by calpain (x) and other caspases (* and **, see Fig. 3 for details) are indicated. It is noted that rev-Dpro2 was constructed to have DEVD instead of ATAD in rev-Dpro1 to check cleavage by caspase-3. The numbers of amino acid residues were according to mouse caspase-12. (B) Purification of caspase-12 variants. Proteins were subjected to SDS–PAGE and stained with Coomassie Blue. Lane 1, HiTrap Q chromatographic fraction of Dpro1; lane 2, Ni–NTA chromatographic fraction of the Dpro1 C298S mutant; lane 3, Ni–NTA chromatographic fraction of the rev-Dpro1; lane 4, Ni–NTA chromatographic fraction of the rev-Dpro2; lane 5, HiTrap Q chromatographic fraction of the Dpro1 D318E/D320E. Relative molecular weights (kDa) were indicated in the left. The arrows in the right indicate purified caspase-12 protein or its subunits.
70
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73
(Qiagen, Chatsworth, CA) and a Hi-TrapQ column (Amersham Pharmacia Biotech, Piscataway, NJ) as described before [20]. About 1 mg of each was obtained from 1 L culture. Caspase-7 was purified essentially in a same way as caspase-12. Assay of caspase activity Substrates were prepared by in vitro transcribing and translating each plasmid DNA in the presence of 1 ll [35S]-Met (1175.0 Ci/ mmole, 10.2 mCi/ml, NEN Life Science Products, Inc., Boston, MA) using a TNT kit (Promega, Madison, WI) as described in the manufacturer’s manual. Reactions were performed in buffer A (20 mM HEPES, pH 7.0, 20 mM NaCl, 1.5 mM MgCl2, 1 mM EDTA, 1 mM EGTA and 10 mM dithiothreitol) at 30° C for 2 h. The products were analyzed by 13.5% sodium dodecylsulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and autoradiography. To determine kcat/ KM of caspase-12, [35S]-labeled Dpro1 C298S protein mixed with non-labeled, purified protein was used as a substrate, in which concentration of the non-labeled substrate was varied. Amounts of cleaved protein products were determined by multiplying the fraction of the cleaved protein by amount of the added protein substrate. The kinetic parameter was determined by using the Michaelis–Menten equation [21,22]. Cell culture, transfection, and cell viability assay HEK 293T cells were cultured in DMEM with 10% heat-inactivated FBS, 100 U of penicillin and 100 lg/ml streptomycin in 5% CO2 at 37 °C. Two microgram pCMV-Tag2A(-) (Stratagene, La Jolla, CA) or its constructs containing caspase-12 genes were transfected into 5 105 of cells by calcium phosphate method [23] .Viability of cells transfected with caspase-12 for 24 h was assessed by mitochondrial dehydrogenase activity using the colorimetric measurements of mitochondrial viability (MTT) assay [17].
tified as a catalytic residue by comparison with other caspases. The result indicates that Dpro1 protein is active and auto-proteolytic process occurred in the bacterial cells. Since the purification of Dpro2 was unsuccessful, we tested a reverse form in which the small subunit precedes the large subunit (Fig. 1A) and may be constitutively active without the processing as shown in reverse forms of caspase-3 and -6 [24]. rev-Dpro2 protein (see Fig. 1A for primary structures) was recovered in the soluble fractions and was readily purified as Dpro1 (Fig. 1B, lanes 3 and 4). We also constructed rev-Dpro1 for comparison with Dpro1. It is noted that rev-Dpro1 was not processed (Fig. 1B, lane 3), although it was active (see below) and contains the potential cleavage site (ATAD) of wild-type caspase-12 (see Fig. 1A). DEVD, substrate sequence of caspase-3, was inserted between the potential cleavage site in rev-Dpro2 to test whether caspase-3 could cleave the site (Fig. 1A). However, it was not cleaved by the enzyme, either. One speculation for the observations is that the three dimensional structure surrounding the cleavage site might be important for the cleavage as well as the primary sequence. Finally, Dpro1 D318E/D320E was also constructed to test whether the cleavage is necessary for the activity of Dpro1 (Fig. 1B, lane 5, and see below). Effects of pH and ionic strength on caspase-12 activity Using purified Dpro1, optimal buffer condition for the activity was determined. High concentration (>75 mM) of NaCl suppressed the enzyme activity (Fig. 2). KCl similarly suppressed the activity (data not shown), implying the ionic strength was the controlling factor rather than the ion type. The physiological concentration (140 mM) of ion, especially K+, was suggested to prevent activation of caspases in healthy cells [17]. It might be suggested that activation of caspase-12 is similarly suppressed in the non-apoptotic condition. Dpro1 was more active at acidic condition than at the physiological pH (Fig. 2) like other caspases [25,26]. It might be in relation to pH drop observed during apoptotic process [27].
Immunoblot analysis Samples were separated by SDS–PAGE and transferred onto a nitrocellulose membrane. The membrane was immunoprobed with indicated primary antibodies and then, horseradish peroxidase-conjugated secondary antibodies (Pierce). The blots were visualized using ECL plus reagent (Amersham Pharmacia Biotech, Piscataway, NJ).
Procaspase-12 cleavage by auto-proteolysis and caspase-7, and mapping the cleavage sites Caspase-12 was shown to cleave only its own, but not other caspases [13,15]. We confirmed the results with purified Dpro1
Caspase-12 assay with cell extract Cell extracts, prepared from human neuroblastoma SK-N-BE (2) as described before [17], were incubated with 1 lg of Dpro1 in buffer A at 30° C for 2 h and analyzed by immunoblot assay. Results Construction and purification of caspase-12 variants Previously, caspase-12 constructs equivalent to the proteins processed by caspase-3/7 and m-calpain were expressed in E. coli, but they were not purified [7,13]. In the current study, we determined to use purified caspase-12 in the following experiments. Dpro1(G95-N419) and Dpro2 (T159-N419) are caspase-12 variants equivalent to those processed by caspase-7 and m-calpain, respectively (Fig. 1A). Dpro1 was highly expressed in E. coli and recovered in the soluble fractions, while full length of caspase-12 (procaspase12) and Dpro2 was mainly detected in the inclusion bodies (data not shown). Dpro1 protein was processed into the large and small subunits (Fig. 1B, lanes 1), whereas the processing was not observed in Dpro1 C298S containing Ser in place of Cys298 (Fig. 1B, lane 2), iden-
Fig. 2. Effects of ionic strength and pH on caspase-12 activity. (A) [35S]-Met-Dpro1 C298S was incubated in the absence (c) or the presence of Dpro1 (200 ng) at the indicated NaCl concentrations or pH. The arrows in the right point to cleaved products. (B) Relative amounts of cleavage products estimated from A are shown.
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73
enzyme (Fig. 3A). Next, cleavage sites in procaspase-12 were examined. Previously, a single cleavage site was identified at C-terminus of Asp318 (Asp319 in rat) [7,13]. Therefore, it was expected that a mutation of the residue (D318E in the current study) would abolish the cleavage, because caspase activity strictly depends on the Asp residue. However, the variant protein was still cleaved by Dpro1, and only D318E/D320E double mutant was resistant to the cleavage (Fig. 3B), suggesting Asp320 as an additional cleavage site. Dpro1 D318E/D320E protein was found to be expressed in E. coli as a single peptide without the cleavage (Fig. 1B, line 5), supporting the above result. It is noted that Dpro1 D318E mutant protein was less efficiently cleaved by Dpro1, probably due to an influence of the mutation on the latter cleavage site. We and others demonstrated that caspase-3 and -7, but not -1, -2, -4, -5, -6, -8 and -9 cleaved procaspase-12 [12,13,15]. It was also shown that caspase-3 cleaved procaspase-12 at carboxyl termini of Asp24 and Asp94 [15]. In the current study, the cleavage sites were mapped with caspase-7 that shares its substrate specificity with caspase-3. As expected, the two sites were also cleaved by caspase-7 (Fig. 3C). In summary, procaspase-12 can be cleaved at four sites; carboxyl termini of Asp318 and Asp320 auto-proteolytically and Asp24 and Asp94 by caspase-3 and -7.
Comparison of activities of caspase-12 variants m-calpain was also reported to generate active caspase-12 enzyme [7]. To examine the activity of caspase-12 cleaved by m-calpain, we constructed caspase-12 variant equivalent to that produced by the enzyme (Dpro2, T159-N419). As mentioned above, Dpro2 was found to be rarely soluble (data not shown). Thus, a reverse form of Dpro2 (rev-Dpro2) was used for the next experiments instead of Dpro2. Enzyme activities of Dpro1, rev-Dpro1 and rev-Dpro2 were measured using Dpro1 C298S mutant protein mixed with the same
71
protein labeled by [35S]-Met as a substrate. Cleavage rates appear to be lower for rev-Dpro2 at concentrations of 0.12 and 0.3 lM Dpro1 C298S, while it proceeded at a similar rate at concentration of 1.2 lM, which might be due to lower KM for the variant (Fig. 4A). Under the experimental condition, however, the saturation of the enzyme activity was never attained so that only kcat/KM value was available. The three variants have comparable kcat/KM values of 108, 158 and 111 M 1 s 1, respectively (Fig. 4A). The similarity in the catalytic parameters indicates that the reverse forms conserve the active site of wild-type protein. Direct comparison of the values with those obtained before [13] was difficult because of the different substrates and enzymes (natural vs. synthetic and purified vs. cell lysates). On the other hand, rough comparison of the values with that of caspase-3 (1.0 106 M 1 s 1) [28] indicates that catalytic efficiency of caspase-12 was four orders of lower magnitude. Low kcat/KM values invariably observed in the three active caspase-12 variant enzymes could be due to either low specificity or inefficient catalytic activity. In spite of the extensive investigation to search putative substrates for the enzyme, it was not successful [13], probably supporting high degree of specificity. This may imply that function of the caspase is highly confined to cleavage of its corresponding proenzyme as suggested before [13]. Next, non-cleavable Dpro1 D318E/D320E was examined whether it is catalytically active or not. It showed much lower magnitude of activity than the above three variants (Fig. 4B). Determination of catalytic parameters of the variant was not practical because of the low activity, but rough estimation from the product as shown Fig. 4B suggests that the activity was at least 10-fold lower than other active variants. The results indicate that the cleavage of the interdomain link is crucial for activation of caspase-12 like other caspases [28].
Death of cells transfected with variable caspase-12 constructs To examine cell death induced by the caspase-12 variants, we transfected the expression vectors of corresponding variants into 293T cells. Unexpectedly, Dpro1 and Dpro2 did not cause any detectable death (Fig. 5A). One notable observation was that they seemed not to be processed, although they were expressed in comparable levels (Fig. 5B). The result is inconsistent with the previous study [7], in which transfection of the Dpro2-equivalent into cells resulted in a significant cell death. Reason for the discrepancy
Fig. 3. Mapping of cleavage sites of procaspase-12 by Dpro1 and caspase-7. (A) Each [35S]-Met-caspase was incubated with 1 lg of Dpro1 and cleavage was examined as described in Materials and methods. (B) Mapping of cleavage sites of procaspase-12 by Dpro1. [35S]-Met-procaspase-12 variants were incubated with Dpro1 (1 lg) and the cleavage was examined. WT, wild-type procaspase-12. (C) Mapping of cleavage sites of procaspase-12 by caspase-7 (casp-7). The same sets of the variants shown in B were incubated with caspase-7 (100 ng) and the cleavage was examined. The arrows in the right indicate cleaved products and relative molecular weights (kDa) are shown in the left.
Fig. 4. Comparison of activity of caspase-12 variants. (A) Activities of Dpro1, revDpro1, rev-Dpro2 (1 lg of each) were measured using purified Dpro1 C298S mutant protein mixed with the same protein labeled by [35S]-Met as a substrate. The concentration of the non-labeled substrate was varied as indicated and thus, amount of the product should be different though the signal intensity appears similar. (B) Activities of caspase-12 variants were measured using purified Dpro1 C298S mutant protein labeled by [35S]-Met (without the non-labeled protein) as a substrate. Lane 1, control without enzyme; lane 2, 0.2 lg Dpro1; lane 3, 0.2 lg revDpro1; lane 4, 0.2 lg or rev-Dpro2; lane 5, 0.2 lg Dpro1 D318E/D320E; lane 6, 0.5 lg Dpro1 D318E/D320E. The arrows in the right indicate cleaved products.
72
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73
Fig. 5. Cell death in cells transfected with variable caspase-12 constructs. Cell viability by MTT assay (A) and processing of proteins by immunoblot assay (B) were assessed in 293T cells transiently transfected with vector (1), procaspase-12 (2), Dpro1 (3), Dpro2 (4), rev-Dpro1 (5) and rev-Dpro2 (6). The data in A are average ± standard deviation of three independent results. In B expressed caspase12 variants tagged with the flag are marked by arrows. C indicates positive control from apoptotic cells and relative molecular weights (kDa) are shown in the left.
remains to be determined, but it might be suggested that the autoprocessing did not occur readily under the intracellular condition such as high concentration of ion (for example, intracellular concentration of K+ is 150 mM) and physiological pH (7.4), because relatively low ion concentration and low pH were necessary for optimal caspase-12 activity (Fig. 2). More importantly, caspase12 with non-cleaved interdomain was barely active (Fig. 4B), implying that removing the prodomain may not be enough for the activation in cells. One the other hand, rev-Dpro1 and rev-Dpro2, constitutively active forms without the processing, were significantly more active in inducing cell death (13% and 25%) than others (Fig. 5A). Overexpression of the fully active proteins might overcome the high concentration of ions and high pH in cells to transfer the apoptotic signal. We subsequently analyzed caspase-3 activation (substrate of caspase-8 and -9), Bid (substrate of caspase-8) cleavage and DFF45 (substrate of caspase-3) cleavage in cells transfected with the expression vectors, but no detectable changes were observed even in the cells transfected with the death-inducing vectors (Fig. 5B). Moreover, any increase of caspase activities, measured with YVAD-AMC, DEVD-AMC, VEID-AMC, IETDAMC, and LEHD-AMC, were not found in the cell extract (data not shown). It was also tested whether Dpro1 directly added to cell extracts could process caspase-3, -7, -8, -9, DFF45 and Bid, but none were detected (data not shown). From these results, we concluded that activated caspase-12 is unlikely to transfer its signal to pathways related with receptor- or mitochondria-mediated apoptosis.
On the other hand, the prodomain of procaspase-12 appears to be removed by other proteases such as m-calpain, caspase-3 and -7 (Fig. 3B and C). Non-cleavable variant of caspase-12 (Dpro1 D318E/ D320E), equivalent to prodomain-free caspase-12 after digestion by caspase-3/-7, exhibited weak activity, indicating that the removal is not enough for full activation (Fig. 4B). Previously, it was demonstrated that the removal of prodomain could lead to full activation of an initiator caspase. In the study, cleavage between the prodomain and the large subunit of procaspase-8 was shown to produce p30 which contains both the large (p18) and small (p10) protease subunits [31]. It was suggested that p30, further processed to p10 and p18 by active caspase-8, could sensitize cells toward death receptor-induced apoptosis. Referring to the report on caspase-8, it might be suggested that caspase-12 is partially activated by the removal of the prodomain by other proteases and then, fully activated by auto-proteolysis. In this regard, procaspase-12 seems to be unique in that the major activation is attained by auto-proteolysis in the linker sites like the initiator caspases, while other proteases are involved in the removal of the prodomain. One speculation on implication of the distinct activation mechanism is that the removal of the prodomain might prime the activation and an additional signal might be necessary for full activation of caspase-12. By this way, additional regulatory site could be provided and probably, the cell could be sensitized to other signal. The consequences of caspase-12 activation remain to be elucidated. It was previously suggested that procaspase-9 was a substrate of caspase-12 [16]. However, the following reports including ours do not support their observation (Fig. 3A) [13,15]. Calcium– calpain–caspase-12–caspase-3 cascade was also suggested to play a role in ER-mediated apoptosis, but there has been no evidence that caspase-3 is a substrate of caspase-12 (Fig. 3A) [14]. The absence of known substrates of caspase-12 has made it difficult to study signal transduction pathway involving caspase-12. In this regard, it is the prime importance to identify substrates of the caspase for investigation of the enzyme-related process. On the other hand, caspase-12 may function in a distinct way as suggested in the inflammatory caspase activation pathway [11]. In the study, caspase-12 has been shown to modulate the activation pathway by a direct dominant negative effect on caspase-1. Although the catalytic activity was found to be dispensable for the effect, it was also suggested that the auto-proteolysis may release its inhibition on caspase-1 held in the early stages of complex formation. In this case, removing the prodomain by caspase or calpain might prime the auto-proteolytic activation. In summary, signal for activation of caspase-12 can be transferred via caspase-3/7 or m-calpain which removes the prodomain. It might be suggested that caspase-12 may function in a distinct way rather than involvement in receptor- or mitochondria-mediated apoptotic signal pathway, because no proteins participating in the pathway including caspase-8, -9, -3, Bid and DFF45 were found as substrates. Finally, the purified caspase-12 will be useful for following research such as development of synthetic substrate and inhibitors for caspase-12.
Discussion Initiator caspases such as caspase-8 are activated by auto-proteolytic cleavage in the linker sites between the large and small domains [29], whereas the linker sites of executioner caspases such as caspase-3 and -7 are cut by the initiators [30] and the prodomains may be deleted by auto-proteolysis [30]. We demonstrated that cleavage in the linkers is critical for the catalytic activation of caspase-12 (Fig. 4B). The cleavage is mediated by auto-proteolysis, and no other proteases including caspases tested by us and others were able to cleave the linkers [13,15], implying that caspase-12 follows the activation mechanism of the initiators.
Acknowledgments This work was supported, in part, by Research fund from Chosun University 2007 and 2008 to I.S.P. We thank Dr. Martine Vanhoucke in BCCM/LMBP plasmid collection for provision of caspase-12 gene. References [1] T.S. Zheng, R.A. Flavell, Nature 400 (1999) 410–411. [2] G.S. Salvesen, V.M. Dixit, Cell 91 (1997) 443–446.
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73 [3] N.A. Thornberry, Y. Lazebnik, Science 281 (1998) 1312–1316. [4] F.L. Scott, B. Stec, C. Pop, M.K. Dobaczewska, J.J. Lee, E. Monosov, H. Robinson, G.S. Salvesen, R. Schwarzenbacher, S.J. Riedl, Nature 457 (2009) 1019–1022. [5] X. Jiang, H.E. Kim, H. Shu, Y. Zhao, H. Zhang, J. Kofron, J. Donnelly, D. Burns, S.C. Ng, S. Rosenberg, X. Wang, Science 299 (2003) 223–226. [6] T. Nakagawa, H. Zhu, N. Morishima, E. Li, J. Xu, B.A. Yankner, J. Yuan, Nature 403 (2000) 98–103. [7] T. Nakagawa, J. Yuan, J. Cell Biol. 150 (2000) 887–894. [8] A.A. Welihinda, W. Tirasophon, R.J. Kaufman, Gene Expr. 7 (1999) 293–300. [9] S. Hu, S.J. Snipas, C. Vincenz, G. Salvesen, V.M. Dixit, J. Biol. Chem. 273 (1998) 29648–29653. [10] V. Bitko, S. Barik, J. Cell Biochem. 80 (2001) 441–454. [11] R.S. Hotchkiss, D.W. Nicholson, Apoptosis and caspases regulate death and inflammation in sepsis, Nat. Rev. Immunol. 6 (2006) 813–822. [12] R.V. Rao, E. Hermel, S. Castro-Obregon, G. del Rio, L.M. Ellerby, H.M. Ellerby, D.E. Bredesen, J. Biol. Chem. 276 (2001) 33869–33874. [13] S. Roy, J.R. Sharom, C. Houde, T.P. Loisel, J.P. Vaillancourt, W. Shao, M. Saleh, D.W. Nicholson, Proc. Natl. Acad. Sci. USA 105 (2008) 4133–4138. [14] A. Masud, A. Mohapatra, S.A. Lakhani, A. Ferrandino, R. Hakem, R.A. Flavell, J. Biol. Chem. 282 (2007) 14132–14139. [15] W. Lee, D.H. Kim, J.H. Boo, Y.H. Kim, I.S. Park, I. Mook-Jung, Apoptosis 10 (2005) 407–415. [16] N. Morishima, K. Nakanishi, H. Takenouchi, T. Shibata, Y. Yasuhiko, J. Biol. Chem. 277 (2002) 34287–34294.
73
[17] P. Karki, C. Seong, J.E. Kim, K. Hur, S.Y. Shin, J.S. Lee, B. Cho, I.S. Park, P. Karki, G.R. Dahal, I.S. Park, Cell Death Differ. 14 (2007) 2068–2075. [18] H.R. Stennicke, Q.L. Deveraux, E.W. Humke, J.C. Reed, V.M. Dixit, G.S. Salvesen, J. Biol. Chem. 274 (1999) 8359–8362. [19] Z. Xia, C.M. Kam, C. Huang, J.C. Powers, R.J. Mandle, R.L. Stevens, J. Lieberman, Biochem. Biophys. Res. Commun. 243 (1998) 384–389. [20] X. Liu, C.N. Kim, J. Yang, R. Jemmerson, X. Wang, Cell 86 (1996) 147–157. [21] I.S. Park, C.T. Walsh, J. Biol. Chem. 272 (1997) 9210–9214. [22] I.A. Lessard, V.L. Healy, I.S. Park, C.T. Walsh, Biochemistry 38 (1999) 14006– 14022. [23] G. Nunez, M.A. Benedict, Y. Hu, N. Inohara, Oncogene 17 (1998) 3237–3245. [24] S.M. Srinivasula, M. Ahmad, M. MacFarlane, Z. Luo, Z. Huang, T. FernandesAlnemri, E.S. Alnemri, J. Biol. Chem. 273 (1998) 10107–10111. [25] N.A. Thornberry, T.A. Rano, E.P. Peterson, D.M. Rasper, T. Timkey, M. GarciaCalvo, V.M. Houtzager, P.A. Nordstrom, S. Roy, J.P. Vaillancourt, K.T. Chapman, D.W. Nicholson, J. Biol. Chem. 272 (1997) 17907–17911. [26] M. Garcia-Calvo, E.P. Peterson, D.M. Rasper, J.P. Vaillancourt, R. Zamboni, D.W. Nicholson, N.A. Thornberry, Cell Death Differ. 6 (1999) 362–369. [27] M. Hengartner, Science 281 (1998) 1298–1299. [28] P. Karki, G.R. Dahal, I.S. Park, Biochem. Biophys. Res. Commun. 356 (2007) 1056–1061. [29] J. Zhang, D. Cado, A. Chen, N.H. Kabra, A. Winoto, Nature 392 (1998) 296–300. [30] H. Liu, D.W. Chang, X. Yang, J. Biol. Chem. 280 (2005) 11578–11582. [31] J.C. Hoffmann, A. Pappa, P.H. Krammer, I.N. Lavrik, Mol. Cell Biol. 29 (2009) 4431–4440.
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi
Purification of catalytically active caspase-12 and its biochemical characterization Hyun-Jung Lee a,1, Sung Haeng Lee b,1, Sung-Hee Park a, Md. Golam Sharoar c, Song Yub Shin b,c, Jung Sup Lee d, Byungyun Cho e, Il-Seon Park b,c,* a
Department of Biology, Ewha Womans University, Seoul, Republic of Korea Department of Cellular and Molecular Medicine, School of Medicine, Chosun University, Gwangju 501-759, Republic of Korea c Department of Bio-Materials Engineering, Chosun University, Gwangju, Republic of Korea d Department of Biotechnology, College of Natural Sciences, Chosun University, Gwangju, Republic of Korea e Department of Chemistry and Medicinal Chemistry, Yonsei University, Wonju, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 5 April 2010 and in revised form 7 July 2010 Available online 18 July 2010 Keywords: Caspase-12 Caspase-7 m-Calpain Apoptosis Purification
a b s t r a c t Caspase-12, mainly detected in endoplasmic reticulum (ER), has been suggested to play a role in ER-mediated apoptosis and inflammatory caspase activation pathway. Cleavage of the prodomain by caspase-3/7 at the carboxyl terminus of Asp94 or m-calpain at the carboxyl terminus of Lys158 was reported to be a part of caspase-12-involved apoptosis. We biochemically characterized the prodomain-free forms of caspase-12 and the equivalent enzymes; Dpro1(G95-D419), rev-Dpro1[(T319-N419)-(G95-D318), a reverse form of Dpro1] and rev-Dpro2[(T319-N419)-(T159-D318)]. The three variants showed comparable activities which were dependent on salt concentration and pH. Auto-proteolytic cleavage was observed at two sites (carboxyl termini of Asp318 and Asp320) in Dpro1. Constitutively active forms of caspase-12 (rev-Dpro1 and rev-Dpro2) could induce cell death in cells transfected with the corresponding expression vectors, but no cleavage of caspase-3, DFF45 or Bid was observed, indicating caspase-12 may mediate a distinct apoptotic pathway rather than caspase-8 or -9-mediated cell death. Ó 2010 Elsevier Inc. All rights reserved.
Introduction Caspases are crucial enzymes for apoptosis that is essential for maintaining homeostasis in multicellular organisms [1–3]. Among the 15 known caspases, caspase-8 is known to be involved in receptor-mediated apoptosis in which the enzyme is activated upon formation of death-inducing signaling complex (DISC)2 with Fas or TNF receptor [4]. Caspase-9, on the other hand, forms a protein complex called apoptosome with Apaf-1, cytochrome c and dATP, which plays a critical role in mitochondria-mediated apoptosis [5]. As a result of the both processes, the executioner caspase-3 and -7 are activated to cleave their substrates such as DFF45 [4,5]. The third class of apoptotic pathway has been reported to be induced by stimuli to endoplasmic reticulum (ER) [6]. Caspase-12 that shares a high homology with caspase-1, -4, -5 and -11 [6–9] is a main constituent of the ER-mediated apoptosis. Cytotoxic agents that induce the activation of the enzyme include amyloid b (Ab) [6], ERdamaging agents such as thapsigargin (inhibitor of the intracellular calcium–ATP transport) and brefeldin A (inhibitor of ER–Golgi trans* Corresponding author at: Department of Cellular and Molecular Medicine, School of Medicine, Chosun University, Gwangju 501-759, Republic of Korea. Fax: +82 62 227 8345. E-mail address: [email protected] (I.-S. Park). 1 These authors contributed equally to this work. 2 Abbreviations used: ER, endoplasmic reticulum; DISC, death-inducing signaling complex; DMEM, Dulbecco’s Modified Eagle Media; FBS, fetal bovine serum. 0003-9861/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2010.07.013
port) [7], and respiratory syncytial virus [10]. Moreover, it is also related to the cytokine processing group-I caspases [11]. Previously, caspase-7 was shown to cleave procaspase-12 at the carboxyl termini of Asp94 and Asp341 [12]. The report also suggested that Asp341 was the self-cleavage site. A recent report, however, indicated a single auto-proteolytic cleavage after Asp318 [13]. Caspase-3 which shares the substrate specificity with caspase-7 also processed procaspase-12 after Asp94 [14,15]. Moreover, m-calpain was also shown to cleave procaspase-12 to generate the active enzyme upon ER-stress by agents such as A23187 (calcium ionophore), thapsigargin and brefeldin A [6,7], wherein the cleavage occurred at the carboxyl termini of Lys158 [7]. On the other hand, substrates of caspase-12 have not been reported, although it was proposed to activate caspase-9 in a cytochrome c-independent manner [16]. The current study was performed to provide biochemical background for the caspase-12-involved cellular processes. One obstacle for the investigation was the absence of purified caspase-12. In the previous studies, crude preparation of caspase-12 in bacterial cell lysates was often used rather than purified enzyme [7,13]. Our initial effort to purify active caspase-12 was not successful because the protein was primarily localized in the inclusion bodies and the solubilization did not produce active enzyme, which might be reason why the enzyme in Escherichia coli cell lysate was used in the previous study [13]. Thus, we initially developed an optimal purification procedure for active caspase-12 and its variants. Through
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73
the subsequent characterization using the purified enzymes, we show that activity of caspase-12 depends on salt concentration and pH like other caspases [17] and the auto-proteolytic cleavage between the large and small domains is essential for the activation. Furthermore, we suggest that caspase-12 may function in a distinct cell death pathway rather than being involved in the pathway mediated by caspase-8 or -9. Materials and methods Materials Dulbecco’s Modified Eagle Media (DMEM), fetal bovine serum (FBS), penicillin and streptomycin were purchased from Life Technologies, Inc. (Gaithersburg, MD). Caspase-3 and DFF45 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and Transduction Laboratories (Lexington, KY), respectively. Polyclonal antibody against Bid was generated by injecting the purified protein into mouse. All other materials were purchased from Sigma. Construction of recombinant caspase-12 To purify active caspase-12, we tested 5 different recombinant caspase-12 constructs including 3 reverse forms (rev-) in which the small subunit (T319-D419) is followed by the large subunit. The recombinant enzymes were the full length (A2-N419), Dpro1 (G95-N419), Dpro2 (T159-N419), rev-full length [(T319-N419)(A2-D318)], rev-Dpro1 [(T319-N419)-(G95-D318)] and rev-Dpro2 [(T319-N419)-(T159-D318)] (see Fig. 1 except rev-full length). For the construction of caspase-12 full length, Dpro1 and Dpro2, polymerase chain reaction (PCR) were performed using pCAGGSmCASP12 plasmid DNA [kindly provided by Mrs. Martine Vanhoucke, BCCM/LMBP plasmid collection (http://www.belspo.be/ bccm/lmbp.html)] as a template with the following primers: full length, TACGGATCCGGCGGCCAGGAGGACACAT (sense), GGCCTCG
69
AGATTCCCGGGAAAAAGGTA (antisense); Dpro1, CGCGGATCCAGG ACCTCAGAAGATATGT (sense), GGCCTCGAGATTCCCGGGAAAAAG GTA (antisense); Dpro2, CGCGGATCCAACAGAAAGGGCAAAAGAG (sense), GGCCTCGAGATTCCCGGGAAAAAGGTA (antisense). The PCR products were cloned in pET28b(-) or pET21b(-). For the construction of the rev-full length and rev-Dpro2, the small subunit was amplified with the primers, GGAATTCCATATGACAGATGAG GAACGTGTG (sense), TACGGATCCATT CCCGGGAAAAAGGTA (antisense) and the PCR product was cloned in the NdeI and BamHI sites of pET21b(-). Next, the large subunits of rev-full length (Ala2Asp318) and rev-Dpro2 (Thr159-Asp318) were amplified with the following primers: (Ala2-Asp318), TACGGATCCGATGAAGTT GATGGTGCGGCCAGGAGG (sense), CCGCTCGAGATCAGCAGTGGCT ATCCC (antisense); (Thr159-Asp318), TACGGATCCGATGAAGTTGAT GGTACAGAAAGGGCAAAAGAG (sense), CCGCTCGAGATCAGCAGTG GCTATCCC (antisense). The PCR products were inserted after the small subunit into the BamHI and XhoI sites. The rev-Dpro1 was generated in a similar way. The small subunit was amplified with the following primers: GGAATTCCATATGACAGATGAGGAACGTGTG (sense), CGCGGATCC AAACCATCAGCAGTGGCATTCCCGGGAAAAAG GTA (antisense). The PCR product was cloned in pET21b(-). (Gly95-Asp318) was amplified with CGCGGATCCAGGACCTCAGAA GATATGT (sense), (Ala2-Asp318) antisense. The PCR product was cloned after the small subunit. Several site-directed mutants of procaspase-12 (D24E/D94E, D94E, D115E, D318E, D320E and D318E/D320E) and Dpro1 C298S were constructed using overlap PCR techniques [18,19]. The PCR products were cloned in pET28b (-) or pET21b(-). Purification of recombinant caspase-12 and caspase-7 The Dpro1, rev-Dpro1, rev-Dpro2, Dpro1 D318E/D320E and Dpro1 C298S cloned in pET21b(-) were overexpressed in E. coli BL21(DE3)pLysS with 0.4 mM isopropyl-b-D-thiogalactopyranoside at 30 °C for 3 h. The proteins were purified using a Ni-NTA column
Fig. 1. (A) Schematic diagrams of caspase-12 and its variant constructs used in this study. The putative cleavage sites of procaspase-12 by calpain (x) and other caspases (* and **, see Fig. 3 for details) are indicated. It is noted that rev-Dpro2 was constructed to have DEVD instead of ATAD in rev-Dpro1 to check cleavage by caspase-3. The numbers of amino acid residues were according to mouse caspase-12. (B) Purification of caspase-12 variants. Proteins were subjected to SDS–PAGE and stained with Coomassie Blue. Lane 1, HiTrap Q chromatographic fraction of Dpro1; lane 2, Ni–NTA chromatographic fraction of the Dpro1 C298S mutant; lane 3, Ni–NTA chromatographic fraction of the rev-Dpro1; lane 4, Ni–NTA chromatographic fraction of the rev-Dpro2; lane 5, HiTrap Q chromatographic fraction of the Dpro1 D318E/D320E. Relative molecular weights (kDa) were indicated in the left. The arrows in the right indicate purified caspase-12 protein or its subunits.
70
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73
(Qiagen, Chatsworth, CA) and a Hi-TrapQ column (Amersham Pharmacia Biotech, Piscataway, NJ) as described before [20]. About 1 mg of each was obtained from 1 L culture. Caspase-7 was purified essentially in a same way as caspase-12. Assay of caspase activity Substrates were prepared by in vitro transcribing and translating each plasmid DNA in the presence of 1 ll [35S]-Met (1175.0 Ci/ mmole, 10.2 mCi/ml, NEN Life Science Products, Inc., Boston, MA) using a TNT kit (Promega, Madison, WI) as described in the manufacturer’s manual. Reactions were performed in buffer A (20 mM HEPES, pH 7.0, 20 mM NaCl, 1.5 mM MgCl2, 1 mM EDTA, 1 mM EGTA and 10 mM dithiothreitol) at 30° C for 2 h. The products were analyzed by 13.5% sodium dodecylsulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and autoradiography. To determine kcat/ KM of caspase-12, [35S]-labeled Dpro1 C298S protein mixed with non-labeled, purified protein was used as a substrate, in which concentration of the non-labeled substrate was varied. Amounts of cleaved protein products were determined by multiplying the fraction of the cleaved protein by amount of the added protein substrate. The kinetic parameter was determined by using the Michaelis–Menten equation [21,22]. Cell culture, transfection, and cell viability assay HEK 293T cells were cultured in DMEM with 10% heat-inactivated FBS, 100 U of penicillin and 100 lg/ml streptomycin in 5% CO2 at 37 °C. Two microgram pCMV-Tag2A(-) (Stratagene, La Jolla, CA) or its constructs containing caspase-12 genes were transfected into 5 105 of cells by calcium phosphate method [23] .Viability of cells transfected with caspase-12 for 24 h was assessed by mitochondrial dehydrogenase activity using the colorimetric measurements of mitochondrial viability (MTT) assay [17].
tified as a catalytic residue by comparison with other caspases. The result indicates that Dpro1 protein is active and auto-proteolytic process occurred in the bacterial cells. Since the purification of Dpro2 was unsuccessful, we tested a reverse form in which the small subunit precedes the large subunit (Fig. 1A) and may be constitutively active without the processing as shown in reverse forms of caspase-3 and -6 [24]. rev-Dpro2 protein (see Fig. 1A for primary structures) was recovered in the soluble fractions and was readily purified as Dpro1 (Fig. 1B, lanes 3 and 4). We also constructed rev-Dpro1 for comparison with Dpro1. It is noted that rev-Dpro1 was not processed (Fig. 1B, lane 3), although it was active (see below) and contains the potential cleavage site (ATAD) of wild-type caspase-12 (see Fig. 1A). DEVD, substrate sequence of caspase-3, was inserted between the potential cleavage site in rev-Dpro2 to test whether caspase-3 could cleave the site (Fig. 1A). However, it was not cleaved by the enzyme, either. One speculation for the observations is that the three dimensional structure surrounding the cleavage site might be important for the cleavage as well as the primary sequence. Finally, Dpro1 D318E/D320E was also constructed to test whether the cleavage is necessary for the activity of Dpro1 (Fig. 1B, lane 5, and see below). Effects of pH and ionic strength on caspase-12 activity Using purified Dpro1, optimal buffer condition for the activity was determined. High concentration (>75 mM) of NaCl suppressed the enzyme activity (Fig. 2). KCl similarly suppressed the activity (data not shown), implying the ionic strength was the controlling factor rather than the ion type. The physiological concentration (140 mM) of ion, especially K+, was suggested to prevent activation of caspases in healthy cells [17]. It might be suggested that activation of caspase-12 is similarly suppressed in the non-apoptotic condition. Dpro1 was more active at acidic condition than at the physiological pH (Fig. 2) like other caspases [25,26]. It might be in relation to pH drop observed during apoptotic process [27].
Immunoblot analysis Samples were separated by SDS–PAGE and transferred onto a nitrocellulose membrane. The membrane was immunoprobed with indicated primary antibodies and then, horseradish peroxidase-conjugated secondary antibodies (Pierce). The blots were visualized using ECL plus reagent (Amersham Pharmacia Biotech, Piscataway, NJ).
Procaspase-12 cleavage by auto-proteolysis and caspase-7, and mapping the cleavage sites Caspase-12 was shown to cleave only its own, but not other caspases [13,15]. We confirmed the results with purified Dpro1
Caspase-12 assay with cell extract Cell extracts, prepared from human neuroblastoma SK-N-BE (2) as described before [17], were incubated with 1 lg of Dpro1 in buffer A at 30° C for 2 h and analyzed by immunoblot assay. Results Construction and purification of caspase-12 variants Previously, caspase-12 constructs equivalent to the proteins processed by caspase-3/7 and m-calpain were expressed in E. coli, but they were not purified [7,13]. In the current study, we determined to use purified caspase-12 in the following experiments. Dpro1(G95-N419) and Dpro2 (T159-N419) are caspase-12 variants equivalent to those processed by caspase-7 and m-calpain, respectively (Fig. 1A). Dpro1 was highly expressed in E. coli and recovered in the soluble fractions, while full length of caspase-12 (procaspase12) and Dpro2 was mainly detected in the inclusion bodies (data not shown). Dpro1 protein was processed into the large and small subunits (Fig. 1B, lanes 1), whereas the processing was not observed in Dpro1 C298S containing Ser in place of Cys298 (Fig. 1B, lane 2), iden-
Fig. 2. Effects of ionic strength and pH on caspase-12 activity. (A) [35S]-Met-Dpro1 C298S was incubated in the absence (c) or the presence of Dpro1 (200 ng) at the indicated NaCl concentrations or pH. The arrows in the right point to cleaved products. (B) Relative amounts of cleavage products estimated from A are shown.
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73
enzyme (Fig. 3A). Next, cleavage sites in procaspase-12 were examined. Previously, a single cleavage site was identified at C-terminus of Asp318 (Asp319 in rat) [7,13]. Therefore, it was expected that a mutation of the residue (D318E in the current study) would abolish the cleavage, because caspase activity strictly depends on the Asp residue. However, the variant protein was still cleaved by Dpro1, and only D318E/D320E double mutant was resistant to the cleavage (Fig. 3B), suggesting Asp320 as an additional cleavage site. Dpro1 D318E/D320E protein was found to be expressed in E. coli as a single peptide without the cleavage (Fig. 1B, line 5), supporting the above result. It is noted that Dpro1 D318E mutant protein was less efficiently cleaved by Dpro1, probably due to an influence of the mutation on the latter cleavage site. We and others demonstrated that caspase-3 and -7, but not -1, -2, -4, -5, -6, -8 and -9 cleaved procaspase-12 [12,13,15]. It was also shown that caspase-3 cleaved procaspase-12 at carboxyl termini of Asp24 and Asp94 [15]. In the current study, the cleavage sites were mapped with caspase-7 that shares its substrate specificity with caspase-3. As expected, the two sites were also cleaved by caspase-7 (Fig. 3C). In summary, procaspase-12 can be cleaved at four sites; carboxyl termini of Asp318 and Asp320 auto-proteolytically and Asp24 and Asp94 by caspase-3 and -7.
Comparison of activities of caspase-12 variants m-calpain was also reported to generate active caspase-12 enzyme [7]. To examine the activity of caspase-12 cleaved by m-calpain, we constructed caspase-12 variant equivalent to that produced by the enzyme (Dpro2, T159-N419). As mentioned above, Dpro2 was found to be rarely soluble (data not shown). Thus, a reverse form of Dpro2 (rev-Dpro2) was used for the next experiments instead of Dpro2. Enzyme activities of Dpro1, rev-Dpro1 and rev-Dpro2 were measured using Dpro1 C298S mutant protein mixed with the same
71
protein labeled by [35S]-Met as a substrate. Cleavage rates appear to be lower for rev-Dpro2 at concentrations of 0.12 and 0.3 lM Dpro1 C298S, while it proceeded at a similar rate at concentration of 1.2 lM, which might be due to lower KM for the variant (Fig. 4A). Under the experimental condition, however, the saturation of the enzyme activity was never attained so that only kcat/KM value was available. The three variants have comparable kcat/KM values of 108, 158 and 111 M 1 s 1, respectively (Fig. 4A). The similarity in the catalytic parameters indicates that the reverse forms conserve the active site of wild-type protein. Direct comparison of the values with those obtained before [13] was difficult because of the different substrates and enzymes (natural vs. synthetic and purified vs. cell lysates). On the other hand, rough comparison of the values with that of caspase-3 (1.0 106 M 1 s 1) [28] indicates that catalytic efficiency of caspase-12 was four orders of lower magnitude. Low kcat/KM values invariably observed in the three active caspase-12 variant enzymes could be due to either low specificity or inefficient catalytic activity. In spite of the extensive investigation to search putative substrates for the enzyme, it was not successful [13], probably supporting high degree of specificity. This may imply that function of the caspase is highly confined to cleavage of its corresponding proenzyme as suggested before [13]. Next, non-cleavable Dpro1 D318E/D320E was examined whether it is catalytically active or not. It showed much lower magnitude of activity than the above three variants (Fig. 4B). Determination of catalytic parameters of the variant was not practical because of the low activity, but rough estimation from the product as shown Fig. 4B suggests that the activity was at least 10-fold lower than other active variants. The results indicate that the cleavage of the interdomain link is crucial for activation of caspase-12 like other caspases [28].
Death of cells transfected with variable caspase-12 constructs To examine cell death induced by the caspase-12 variants, we transfected the expression vectors of corresponding variants into 293T cells. Unexpectedly, Dpro1 and Dpro2 did not cause any detectable death (Fig. 5A). One notable observation was that they seemed not to be processed, although they were expressed in comparable levels (Fig. 5B). The result is inconsistent with the previous study [7], in which transfection of the Dpro2-equivalent into cells resulted in a significant cell death. Reason for the discrepancy
Fig. 3. Mapping of cleavage sites of procaspase-12 by Dpro1 and caspase-7. (A) Each [35S]-Met-caspase was incubated with 1 lg of Dpro1 and cleavage was examined as described in Materials and methods. (B) Mapping of cleavage sites of procaspase-12 by Dpro1. [35S]-Met-procaspase-12 variants were incubated with Dpro1 (1 lg) and the cleavage was examined. WT, wild-type procaspase-12. (C) Mapping of cleavage sites of procaspase-12 by caspase-7 (casp-7). The same sets of the variants shown in B were incubated with caspase-7 (100 ng) and the cleavage was examined. The arrows in the right indicate cleaved products and relative molecular weights (kDa) are shown in the left.
Fig. 4. Comparison of activity of caspase-12 variants. (A) Activities of Dpro1, revDpro1, rev-Dpro2 (1 lg of each) were measured using purified Dpro1 C298S mutant protein mixed with the same protein labeled by [35S]-Met as a substrate. The concentration of the non-labeled substrate was varied as indicated and thus, amount of the product should be different though the signal intensity appears similar. (B) Activities of caspase-12 variants were measured using purified Dpro1 C298S mutant protein labeled by [35S]-Met (without the non-labeled protein) as a substrate. Lane 1, control without enzyme; lane 2, 0.2 lg Dpro1; lane 3, 0.2 lg revDpro1; lane 4, 0.2 lg or rev-Dpro2; lane 5, 0.2 lg Dpro1 D318E/D320E; lane 6, 0.5 lg Dpro1 D318E/D320E. The arrows in the right indicate cleaved products.
72
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73
Fig. 5. Cell death in cells transfected with variable caspase-12 constructs. Cell viability by MTT assay (A) and processing of proteins by immunoblot assay (B) were assessed in 293T cells transiently transfected with vector (1), procaspase-12 (2), Dpro1 (3), Dpro2 (4), rev-Dpro1 (5) and rev-Dpro2 (6). The data in A are average ± standard deviation of three independent results. In B expressed caspase12 variants tagged with the flag are marked by arrows. C indicates positive control from apoptotic cells and relative molecular weights (kDa) are shown in the left.
remains to be determined, but it might be suggested that the autoprocessing did not occur readily under the intracellular condition such as high concentration of ion (for example, intracellular concentration of K+ is 150 mM) and physiological pH (7.4), because relatively low ion concentration and low pH were necessary for optimal caspase-12 activity (Fig. 2). More importantly, caspase12 with non-cleaved interdomain was barely active (Fig. 4B), implying that removing the prodomain may not be enough for the activation in cells. One the other hand, rev-Dpro1 and rev-Dpro2, constitutively active forms without the processing, were significantly more active in inducing cell death (13% and 25%) than others (Fig. 5A). Overexpression of the fully active proteins might overcome the high concentration of ions and high pH in cells to transfer the apoptotic signal. We subsequently analyzed caspase-3 activation (substrate of caspase-8 and -9), Bid (substrate of caspase-8) cleavage and DFF45 (substrate of caspase-3) cleavage in cells transfected with the expression vectors, but no detectable changes were observed even in the cells transfected with the death-inducing vectors (Fig. 5B). Moreover, any increase of caspase activities, measured with YVAD-AMC, DEVD-AMC, VEID-AMC, IETDAMC, and LEHD-AMC, were not found in the cell extract (data not shown). It was also tested whether Dpro1 directly added to cell extracts could process caspase-3, -7, -8, -9, DFF45 and Bid, but none were detected (data not shown). From these results, we concluded that activated caspase-12 is unlikely to transfer its signal to pathways related with receptor- or mitochondria-mediated apoptosis.
On the other hand, the prodomain of procaspase-12 appears to be removed by other proteases such as m-calpain, caspase-3 and -7 (Fig. 3B and C). Non-cleavable variant of caspase-12 (Dpro1 D318E/ D320E), equivalent to prodomain-free caspase-12 after digestion by caspase-3/-7, exhibited weak activity, indicating that the removal is not enough for full activation (Fig. 4B). Previously, it was demonstrated that the removal of prodomain could lead to full activation of an initiator caspase. In the study, cleavage between the prodomain and the large subunit of procaspase-8 was shown to produce p30 which contains both the large (p18) and small (p10) protease subunits [31]. It was suggested that p30, further processed to p10 and p18 by active caspase-8, could sensitize cells toward death receptor-induced apoptosis. Referring to the report on caspase-8, it might be suggested that caspase-12 is partially activated by the removal of the prodomain by other proteases and then, fully activated by auto-proteolysis. In this regard, procaspase-12 seems to be unique in that the major activation is attained by auto-proteolysis in the linker sites like the initiator caspases, while other proteases are involved in the removal of the prodomain. One speculation on implication of the distinct activation mechanism is that the removal of the prodomain might prime the activation and an additional signal might be necessary for full activation of caspase-12. By this way, additional regulatory site could be provided and probably, the cell could be sensitized to other signal. The consequences of caspase-12 activation remain to be elucidated. It was previously suggested that procaspase-9 was a substrate of caspase-12 [16]. However, the following reports including ours do not support their observation (Fig. 3A) [13,15]. Calcium– calpain–caspase-12–caspase-3 cascade was also suggested to play a role in ER-mediated apoptosis, but there has been no evidence that caspase-3 is a substrate of caspase-12 (Fig. 3A) [14]. The absence of known substrates of caspase-12 has made it difficult to study signal transduction pathway involving caspase-12. In this regard, it is the prime importance to identify substrates of the caspase for investigation of the enzyme-related process. On the other hand, caspase-12 may function in a distinct way as suggested in the inflammatory caspase activation pathway [11]. In the study, caspase-12 has been shown to modulate the activation pathway by a direct dominant negative effect on caspase-1. Although the catalytic activity was found to be dispensable for the effect, it was also suggested that the auto-proteolysis may release its inhibition on caspase-1 held in the early stages of complex formation. In this case, removing the prodomain by caspase or calpain might prime the auto-proteolytic activation. In summary, signal for activation of caspase-12 can be transferred via caspase-3/7 or m-calpain which removes the prodomain. It might be suggested that caspase-12 may function in a distinct way rather than involvement in receptor- or mitochondria-mediated apoptotic signal pathway, because no proteins participating in the pathway including caspase-8, -9, -3, Bid and DFF45 were found as substrates. Finally, the purified caspase-12 will be useful for following research such as development of synthetic substrate and inhibitors for caspase-12.
Discussion Initiator caspases such as caspase-8 are activated by auto-proteolytic cleavage in the linker sites between the large and small domains [29], whereas the linker sites of executioner caspases such as caspase-3 and -7 are cut by the initiators [30] and the prodomains may be deleted by auto-proteolysis [30]. We demonstrated that cleavage in the linkers is critical for the catalytic activation of caspase-12 (Fig. 4B). The cleavage is mediated by auto-proteolysis, and no other proteases including caspases tested by us and others were able to cleave the linkers [13,15], implying that caspase-12 follows the activation mechanism of the initiators.
Acknowledgments This work was supported, in part, by Research fund from Chosun University 2007 and 2008 to I.S.P. We thank Dr. Martine Vanhoucke in BCCM/LMBP plasmid collection for provision of caspase-12 gene. References [1] T.S. Zheng, R.A. Flavell, Nature 400 (1999) 410–411. [2] G.S. Salvesen, V.M. Dixit, Cell 91 (1997) 443–446.
H.-J. Lee et al. / Archives of Biochemistry and Biophysics 502 (2010) 68–73 [3] N.A. Thornberry, Y. Lazebnik, Science 281 (1998) 1312–1316. [4] F.L. Scott, B. Stec, C. Pop, M.K. Dobaczewska, J.J. Lee, E. Monosov, H. Robinson, G.S. Salvesen, R. Schwarzenbacher, S.J. Riedl, Nature 457 (2009) 1019–1022. [5] X. Jiang, H.E. Kim, H. Shu, Y. Zhao, H. Zhang, J. Kofron, J. Donnelly, D. Burns, S.C. Ng, S. Rosenberg, X. Wang, Science 299 (2003) 223–226. [6] T. Nakagawa, H. Zhu, N. Morishima, E. Li, J. Xu, B.A. Yankner, J. Yuan, Nature 403 (2000) 98–103. [7] T. Nakagawa, J. Yuan, J. Cell Biol. 150 (2000) 887–894. [8] A.A. Welihinda, W. Tirasophon, R.J. Kaufman, Gene Expr. 7 (1999) 293–300. [9] S. Hu, S.J. Snipas, C. Vincenz, G. Salvesen, V.M. Dixit, J. Biol. Chem. 273 (1998) 29648–29653. [10] V. Bitko, S. Barik, J. Cell Biochem. 80 (2001) 441–454. [11] R.S. Hotchkiss, D.W. Nicholson, Apoptosis and caspases regulate death and inflammation in sepsis, Nat. Rev. Immunol. 6 (2006) 813–822. [12] R.V. Rao, E. Hermel, S. Castro-Obregon, G. del Rio, L.M. Ellerby, H.M. Ellerby, D.E. Bredesen, J. Biol. Chem. 276 (2001) 33869–33874. [13] S. Roy, J.R. Sharom, C. Houde, T.P. Loisel, J.P. Vaillancourt, W. Shao, M. Saleh, D.W. Nicholson, Proc. Natl. Acad. Sci. USA 105 (2008) 4133–4138. [14] A. Masud, A. Mohapatra, S.A. Lakhani, A. Ferrandino, R. Hakem, R.A. Flavell, J. Biol. Chem. 282 (2007) 14132–14139. [15] W. Lee, D.H. Kim, J.H. Boo, Y.H. Kim, I.S. Park, I. Mook-Jung, Apoptosis 10 (2005) 407–415. [16] N. Morishima, K. Nakanishi, H. Takenouchi, T. Shibata, Y. Yasuhiko, J. Biol. Chem. 277 (2002) 34287–34294.
73
[17] P. Karki, C. Seong, J.E. Kim, K. Hur, S.Y. Shin, J.S. Lee, B. Cho, I.S. Park, P. Karki, G.R. Dahal, I.S. Park, Cell Death Differ. 14 (2007) 2068–2075. [18] H.R. Stennicke, Q.L. Deveraux, E.W. Humke, J.C. Reed, V.M. Dixit, G.S. Salvesen, J. Biol. Chem. 274 (1999) 8359–8362. [19] Z. Xia, C.M. Kam, C. Huang, J.C. Powers, R.J. Mandle, R.L. Stevens, J. Lieberman, Biochem. Biophys. Res. Commun. 243 (1998) 384–389. [20] X. Liu, C.N. Kim, J. Yang, R. Jemmerson, X. Wang, Cell 86 (1996) 147–157. [21] I.S. Park, C.T. Walsh, J. Biol. Chem. 272 (1997) 9210–9214. [22] I.A. Lessard, V.L. Healy, I.S. Park, C.T. Walsh, Biochemistry 38 (1999) 14006– 14022. [23] G. Nunez, M.A. Benedict, Y. Hu, N. Inohara, Oncogene 17 (1998) 3237–3245. [24] S.M. Srinivasula, M. Ahmad, M. MacFarlane, Z. Luo, Z. Huang, T. FernandesAlnemri, E.S. Alnemri, J. Biol. Chem. 273 (1998) 10107–10111. [25] N.A. Thornberry, T.A. Rano, E.P. Peterson, D.M. Rasper, T. Timkey, M. GarciaCalvo, V.M. Houtzager, P.A. Nordstrom, S. Roy, J.P. Vaillancourt, K.T. Chapman, D.W. Nicholson, J. Biol. Chem. 272 (1997) 17907–17911. [26] M. Garcia-Calvo, E.P. Peterson, D.M. Rasper, J.P. Vaillancourt, R. Zamboni, D.W. Nicholson, N.A. Thornberry, Cell Death Differ. 6 (1999) 362–369. [27] M. Hengartner, Science 281 (1998) 1298–1299. [28] P. Karki, G.R. Dahal, I.S. Park, Biochem. Biophys. Res. Commun. 356 (2007) 1056–1061. [29] J. Zhang, D. Cado, A. Chen, N.H. Kabra, A. Winoto, Nature 392 (1998) 296–300. [30] H. Liu, D.W. Chang, X. Yang, J. Biol. Chem. 280 (2005) 11578–11582. [31] J.C. Hoffmann, A. Pappa, P.H. Krammer, I.N. Lavrik, Mol. Cell Biol. 29 (2009) 4431–4440.