PEDS Advance Access originally published online on March 9, 2007
Protein Engineering Design and Selection 2007 20(4):163-170; doi:10.1093/protein/gzm007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Engineering of 
1-antitrypsin variants selective for subtilisin-like proprotein convertases PACE4 and PC6: Importance of the P2' residue in stable complex formation of the serpin with proprotein convertase
1 Department of Biological Science and Technology, The University of Tokushima Graduate School, 2-1 Minamijosanjima, Tokushima 770-8506, Japan
2 To whom correspondence should be addressed. E-mail: tsuji{at}bio.tokushima-u.ac.jp
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionAcknowledgementReferences |
|---|
Furin and PACE4, members of the subtilisin-like proprotein convertase (SPC) family, have been implicated in the metastatic progression of certain tumors in addition to the activation of viral coat proteins and bacterial toxins, indicating that these enzymes are potential targets for therapeutic agents.
1-Antitrypsin Portland is an engineered 1-antitrypsin designed as a furin-specific inhibitor and has been used as a tool in the functional analysis of furin. In this work, we engineered rat 1-antitrypsin to create a PACE4-specific inhibitor. Substituting Arg-Arg-Arg-Arg for Ala-Val-Pro-Met352 at P4-P1 and Ala for Leu354 at P2' created a potent PACE4- and PC6-specific inhibitor. This variant (RRRRSA) formed an SDS- and heat-stable serpin/proteinase complex with PACE4 or PC6 and inhibited both enzyme activities. The RRRRSA variant was efficiently cleaved by furin without formation of the stable complex. This is the first report of a highly selective protein-based inhibitor of PACE4 and PC6. This inhibitor will be useful in delineating the roles of PACE4 and PC6 localized in the extracellular matrix.
Keywords: 1-antitrypsin/furin/PACE4/PC6/proprotein convertase
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementReferences |
|---|
The subtilisin-like proprotein convertases (SPCs) are a family of Ca2 +-dependent serine endoproteases that play a key role in the proteolytic activation of proproteins through cleavage at RXKR, RXRR, RXXR, RR and KR sequences. To date, seven mammalian SPCs including furin, PC1 (also known as PC3), PC2, PACE4, PC4, PC6 (also known as PC5) and PC8 (also known as LPC or PC7) have been identified (Nakayama, 1997
; Thomas, 2002). Furin and PC8 are located in the Golgi compartment as membrane-bound enzymes. This subgroup shows ubiquitous distribution, whereas PC1, PC2 and PC4 are located in secretory vesicles of endocrine cells and play an important role in the maturation of propeptide hormones. PACE4 and PC6 are secretory enzymes. We have documented the highly regulated expression profile of PACE4, which differs from furin, during cell differentiation (Tsuji et al., 1999a; Yoshida et al., 2001; Bando et al., 2002; Koide et al., 2003). PACE4 and PC6 are unique SPCs that anchor heparan sulfate proteoglycans on the extracellular matrix (ECM) (Tsuji et al., 2003). Precise biochemical knowledge pertaining to the distinct role of PACE4 and furin remains unknown.
1-Antitrypsin Portland (1-PDX) is an engineered antitrypsin variant designed as a furin inhibitor (Anderson et al., 1993). This variant contains a minimal consensus sequence (Arg-X-X-Arg) for efficient cleavage by furin in its reactive site loop. PC6 and PACE4 are also inhibited by this variant (Jean et al., 1998; Tsuji et al., 1999b). 1-Antitrypsin is a member of the serpin family. Serpins employ an extraordinary mechanism of protease inhibition driven by rapid and marked conformational change that results in destruction of the covalently linked protease (Huntington, 2006). Intracellular expression of 1-PDX cDNA or the exogenous addition of 1-PDX protein blocks various pathological events including tumor growth, virus replication and the activation of bacterial toxin through inhibition of SPCs (Jean et al., 1998, 2000; Khatib et al., 2001; Mahloogi et al., 2002; Bassi et al., 2005). These observations indicate the potential therapeutic value of specific inhibitors of PACE4, furin and PC6. Compared with other inhibitors such as EDTA and peptidyl chloromethylketone compounds, 1-PDX is selective for SPCs including furin, PACE4 and PC6, which function in the constitutive secretory pathway (Jean et al., 1998; Tsuji et al., 1999b). More selective inhibition of furin and PACE4 in lieu of 1-PDX is essential in order to identify their respective biological roles. We previously demonstrated that a novel 1-antitrypsin variant, AVRR (AVPM352/AVRR), containing dibasic sequence within its reactive site loop, inhibited furin and PC6, but not PACE4 (Tsuji et al., 2002).
In this work, we prepared rat 1-antitrypsin variants that contained RRRR352 in P4-P1 in addition to a different amino acid at P2'. The selectivity of these variants was determined by investigating SDS-stable serpin/proteinase complex formation and the inhibition of proteinase activity. Stable complex formation of variants with furin changed markedly when the P2' position of the RRRR352 variants was mutated from leucine to alanine, phenylalanine, tryptophan or glutamate. The 1-antitrypsin variant RRRRSA (AVPMSL354/RRRRSA) efficiently formed SDS- and heat-stable complexes with PACE4 or PC6, whereas this variant failed to form a stable complex with furin, and inhibited PACE4, but not furin, activity.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementReferences |
|---|
Materials
Pyroglutamyl-Arg-Thr-Lys-Arg-4-methylcoumaryl-7-amide (pyr-RTKR-MCA) was purchased from the Peptide Institute (Osaka, Japan). Rabbit anti-human 1-antitrypsin antiserum was from ZYMED (San Francisco, CA, USA). Rabbit anti-human laminin -5 (H-160) antibody was from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Chinese hamster ovary (CHO) cells stably expressing mouse soluble furin were a gift from Prof. Kazuhisa Nakayama (Kyoto University, Kyoto, Japan). Easy TagTM Express Protein Labeling Mix [35S] was from NEN Life Science. The other reagents used were of the highest grade available.
Construction of expression plasmids
Human PACE4A (A-I isoform) was subcloned into the mammalian expression vector, pALTER-MAX (Promega Corp., Madison, WI, USA), as previously described (Nagahama et al., 1998). Expression vectors for truncated mouse furin and PC6A were kindly provided by Prof. K. Nakayama. Rat wild-type 1-antitrypsin (AVPMSL354, P1 site is Met352), 1-PDX (AVPMSL354/RVPRSL) and AVRR (AVPMSL354/AVRRSL) were subcloned into the pcDNA3 vector (Invitrogen, Carlsbad, CA, USA) as previously described (Tsuji et al., 2002). Novel variants of rat 1-antitrypsin were generated by a two-step PCR as previously described (Tsuji et al., 2002). The sense mutagenic primers used for introducing the mutations are shown in Table I. First the 3'-terminal cDNA fragment containing the mutated reactive site loop was generated by PCR using a sense mutagenic primer and EcoRI-linked antisense primer (5'-TTGAATTCGTGATTAACGTGTGGGATCTATC-3', stop codon underlined). The amplified fragment was used as an antisense primer for a second PCR. The full-length variant cDNA was generated by a second PCR using BamHI-linked sense primer (5'-GGATCCTGAAAATGGCACCCTCCATCTCACGG-3', initiation codon underlined) and the first PCR product as the antisense primer. AVRR (AVPMSL354/AVRRSL) and RVRR (AVPMSL354/RVRRSL) 1-antitrypsin variant cDNA was used as the template to create ARRR (AVPMSL354 /ARRRSL) and RRRRSL (AVPMSL354/RRRRSL) variants, respectively. cDNAs coding RRRRSA (AVPMSL354 /RRRRSA), RRRRSF (AVPMSL354/RRRRSF), RRRRSW (AVPMSL354/RRRRSW), RRRRSK (AVPMSL354/RRRRSK), RRRRSE (AVPMSL354/RRRRSE) and RRRRSQ (AVPMSL354/RRRRSQ) variants were prepared using RRRRSL variant cDNA as the template. Amplified cDNA was digested with restriction enzymes (BamHI and EcoRI) and subcloned into the BamHI and EcoRI site of the pcDNA3 vector. The complete insert sequence was confirmed using an ALF DNA sequencer (Pharmacia, Uppsala, Sweden).
|
Formation of enzyme-inhibitor complex ex vivo
The ability of the 1-antitrypsin variant to form an SDS- and heat-stable complex with PACE4, furin or PC6 was analyzed by transfecting human embryonic kidney (HEK) 293-EBNA, Cos-7 and CHO cells stably expressing furin. HEK293-EBNA and Cos-7 cells were grown at 37°C under 5% CO2 in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and antibiotics. CHO cells stably expressing soluble furin were cultured in Eagle's minimal essential medium supplemented with 10% fetal calf serum, antibiotics and 0.3 µM methotrexate. Expression plasmids for SPC (1µg) and 1-antitrypsin variants (1µg) were cotransfected into cultured cells in a 35 mm dish and radiolabeled with [35S] methionine/cysteine as previously described (Tsuji et al., 2002). As a negative control, wild-type 1-antitrypsin was also cotransfected with the expression vector for SPC. The conditioned medium was then immunoprecipitated using anti-PACE4 homoB, anti-subtilisin-like catalytic domain (SCD) or anti-1-antitrypsin antibodies and subsequently analyzed by 7.5% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (Laemmli, 1970) followed by fluorography as previously reported (Mori et al., 1997). Anti-PACE4 homoB antibody was used for the immunoprecipitation of PACE4. Anti-SCD antibody was used for furin and PC6. Radioactive bands in the gel were analyzed using a BAS-1500 bioimaging analyzer (Fuji Film, Tokyo, Japan). Experiments were performed at least three times to ensure reproducibility.
SPC proteinase activity was assayed using pyr-RTKR-MCA in the presence of leupeptin, chymostatin, E-64 (trans-epoxysuccinyl-L-leucylamido-guanidino-butane), pepstatin and bestatin as previously described (Tsuji et al., 1999b). The 7-amino-4-methylcoumarin liberated was determined fluometrically with a Hitachi F-2000 spectrofluorometer (380 nm excitation, 460 nm emission). The protein content was determined by the method of Bradford using Bio-Rad protein assay reagent with bovine serum albumin as a standard (Bradford, 1976).
Inhibition of furin by 1-antitrypsin variant ex vivo
CHO cells (80% confluent) stably expressing soluble furin in a 60 mm dish were transfected with expression plasmid (2.5 µg) for 1-antitrypsin variants, 1-PDX, AVRR, RRRRSL or RRRRSA. Following transfection for 48 h, cells were rinsed twice with phosphate-buffered saline and incubated in serum-free Opti-MEM (Gibco BRLs, Rockville, MD, USA). After 48 h, the conditioned medium was collected, concentrated (8-fold) by ultrafiltration and furin activity was assayed as mentioned earlier.
Preparation of recombinant furin and 1-antitrypsin variants
Confluent CHO cells stably expressing soluble furin lacking the transmembrane domain were rinsed twice with phosphate-buffered saline and incubated in serum-free Opti-MEM. After 24 h, the conditioned medium was collected, concentrated by ultrafiltration and stored at – 80°C until use. HEK293-EBNA cells were transfected with expression plasmid for the 1-PDX or RRRRSA variant. Following transfection for 48 h, cells were rinsed twice with phosphate-buffered saline and incubated in serum-free Opt-MEM. After 24 h, the conditioned medium was collected and concentrated by ultrafiltration. Western blot analysis of 1-antitrypsin variant was performed as previously described (Yoshida et al., 2001).
Inhibition of enzyme activity of ECM-associated PACE4 or PC6 by 1-antitrypsin variant ex vivo
Expression plasmids for SPC (PACE4 or PC6) and 1-antitrypsin variant (1-PDX, RRRRSL or RRRRSA) were cotransfected into Cos-7 cells. The ECM fraction remaining attached to the culture dishes was prepared by the method of Blenis and Hawkes (1983) with modifications as follows. After 48 h, the conditioned medium was removed and the cells were harvested by scraping. The insoluble cross-linked ECM fraction remaining associated with the culture dish surface was washed five times with phosphate-buffered saline containing 1 mM EDTA. This solidified ECM layer was used for the determination of ECM-associated PACE4 or PC6. We previously showed that secreted PACE4 and PC6 were bound to this ECM fraction on the culture dish (Tsuji et al., 2003). Approximately 5 µg protein was recovered from an ECM-coated dish (35 mm diameter) following extraction by boiling in SDS buffer (50 mM Tris–HCl pH 7.5 containing 1% SDS). ECM-associated PACE4 and PC6 activity was assayed following incubation of the ECM-coated dish (35 mm diameter) with substrate reaction mixture (0.6 ml). As a negative control, the activity in the presence of 10 mM EDTA was analyzed and subtracted from the activity obtained in the absence of EDTA. Western blot of laminin -5 was performed using anti-human laminin -5 (H-160) antibody as described previously (Yoshida et al., 2001).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementReferences |
|---|
Mutagenesis between P4-P2' residue of
1-antitrypsin
We had previously shown that the 1-PDX variant possesses relatively broad selectivity toward furin, PACE4 and PC6, and the novel variants AVRR (AVPM352/AVRR) and RVRR (AVPM352/RVRR) had distinct selectivity toward these enzymes (Tsuji et al., 2002). The AVRR variant inhibited furin and PC6, but not PACE4. Although the RVRR variant effectively inhibited furin and PACE4, a 600-fold higher concentration relative to 1-PDX was required to inhibit PC6. These results suggested that further development of 1-antitrypsin variant selectivity would be possible by mutagenesis of the reactive site loop.
In an effort to generate an 1-antitrypsin variant with increased reactivity toward PACE4 as opposed to 1-PDX, the novel variants ARRR and RRRRSL were prepared and their selectivity was determined by SDS-stable proteinase/inhibitor complex formation ex vivo. As shown in Fig. 1A, the RRRRSL variant efficiently formed a complex with PACE4 and PC6, like the case with 1-PDX. Just as with the AVRR variant, complex formation of the ARRR variant with PACE4 was negligible, confirming that Arg at the P4 position is essential for the inhibition of PACE4. PC6 equally formed complexes with 1-PDX, AVRR, ARRR and RRRRSL variants. On the other hand, the efficiency of SDS-stable complex formation of the RRRRSL variant with furin was lower compared to that with 1-PDX (Fig. 1B). These results indicate that the RRRRSL variant is more selective for PACE4 and PC6, rather than for furin. Further mutations were introduced to the RRRRSL variant and the effect on selectivity for furin, PACE4 and PC6 was examined. The serpin family shares a homologous C-terminal reactive site as shown in Fig. 1C. The P1 and upstream residues determine the specificity of serpins. Although the P1' and P2' sites are usually Ser/Thr and a hydrophobic residue, respectively, the P1' residue of plasmin inhibitor and leukocyte elastase inhibitor is Met. The length of the side chain of the hydrophobic P2' residue can also vary; in the case of plasmin inhibitor, the P2' residue is Ser. All amino acids except proline were tolerated at the P1' position of PAI-1, and P1' mutation had only minimal effect on its inhibitory activity (Sherman et al., 1992). On the other hand, the specificity of 1-antitrypsin could be altered by mutation at the P2'-P3' position (Filion et al., 2004). The P1'-P3' position (Ser-Ile-Pro) of human 1-antitrypsin was replaced with the corresponding heparin cofactor sequence (Ser-Thr-Gln) at the P1'-P3' position. Heparin cofactor II is a serpin specific for thrombin (Parker and Tollefsen, 1985). This variant can form an SDS-stable complex with thrombin, unlike wild-type 1-antitrypsin. These results suggested that the carboxyl terminal side of the proteinase cleavage site of 1-antitrypsin is also important for its specificity.
|
1-Antitrypsin variants containing different combinations of arginine residues at the P1, P2, P4 and P6 positions were evaluated for their inhibitory activity toward furin (Dufour et al., 2001). However, the effect of mutation at the P1' and P2' positions of 1-antitrypsin variant on its selectivity toward furin, PACE4 and PC6 remained unexamined. Studies on cleavage specificity indicates that furin prefers polar/acidic P1' and hydrophobic P2' residues (Nakayama, 1997). Ser at P1' of the RRRRSL variant was initially replaced with Glu (RRRREL variant) and SDS-stable complex formation ex vivo with furin, PACE4 and PC6 was examined. As shown in Fig. 1D, the efficiency of stable complex formation of the RRRREL variant with PACE4 and PC6 was lower compared with 1-PDX. In the case of furin, the efficiency remained more or less unchanged by this mutation. We then introduced further mutations at the P2' position of the RRRRSL variant in an effort to improve its selectivity. At P2', Leu was replaced by Ala (RRRRSA), Phe (RRRRSF), Trp (RRRRSW), Lys (RRRRSK), Glu (RRRRSE) or Gln (RRRRSQ), and complex formation ex vivo was examined. The susceptibility of these variants to proteolysis and their secretion level in HEK293-EBNA and Cos-7 cells was examined. Cells were transfected with the expression plasmid for wild-type or variant 1-antitrypsin without cotransfection with SPC. The conditioned medium was immunoprecipitated with anti-1-antitrypsin antibody (Fig. 2). The variants were secreted in uncleaved form (55 kDa) in both cell types with similar expression levels. In particular, a small amount of cleaved 1-antitrypsin variants (51 kDa) were detected in Cos-7 cells.
|
Complex formation of PACE4, PC6 and furin with the RRRR variant ex vivo
To confirm cleavage at the P1 site of the variant, the conditioned medium from cells transfected with expression vectors for furin and the variant was immunoprecipitated with anti-1-antitrypsin antibody. As shown in Fig. 3A, wild-type 1-antitrypsin failed to form a complex and was observed as a 55 kDa band. The band corresponding to the furin/1-antitrypsin variant complex and the cleaved form of the variant (51 kDa) were observed in the case of 1-PDX. The ratio of complex/cleaved 1-PDX was 1.0. Similarly, the RRRRSL variant was also secreted as a cleaved form and the ratio of complex/cleaved variant was 0.6. In contrast, bands corresponding to complexes were barely detected when RRRRSA, RRRRSF, RRRRSW or RRRRSE variant was coexpressed with furin, even though all variants were secreted as the cleaved form. On the other hand, the RRRRSK and RRRRSQ variants formed complexes and were secreted as the cleaved form. To confirm the difference in selectivity between RRRRSL and RRRRSA variants, the variants were expressed in CHO cells stably expressing soluble furin, and the conditioned medium subsequently immunoprecipitated using anti-1-antitrypsin antibody (Fig. 3B). A band corresponding to the stable complex of the RRRRSA variant with furin was not detected, even though the RRRRSL variant formed an SDS-stable complex.
|
These variants were tested for their ability to form complexes with PACE4 and PC6 ex vivo (Fig. 4). Although PACE4 formed SDS-stable complexes equally with all variants, the cleavage specificity toward these variants differed. The efficiency of cleavage of the variant (cleaved/uncleaved) by PACE4 was 32% (
1-PDX), 39% (RRRRSL), 73% (RRRRSA), 58% (RRRRSF), 55% (RRRRSW), 25% (RRRRSK), 46% (RRRRSE) and 35% (RRRRSQ). Of the variants tested, PACE4 was found to cleave RRRRSA variant best. Similarly, the efficiency of the complex formation of PC6 with these variants did not vary significantly. The reactivity of 1-PDX, RRRRSQ and RRRRSK toward PC6 was slightly higher than the other variants. The cleavage specificity toward these variants clearly differed, as in the case with PACE4. The RRRRSA, RRRRSW and RRRRSE variants were completely cleaved by PC6, whereas 1-PDX, RRRRSL and RRRRSK were cleaved to a lesser extent.
|
Sensitivity of furin to the various
1-antitrypsin variants ex vivo and in vitro
To examine the difference in selectivity of the 1-antitrypsin variants toward furin, the effect of these variants on furin activity ex vivo and in vitro was examined. The expression vector containing wild-type, 1-PDX, AVRR, RRRRSL or RRRRSA variants of 1-antitrypsin was transfected into CHO cells expressing stably soluble furin, and secreted furin activity was examined using pyr-RTKR-MCA. 1-Antitrypsin secreted into the conditioned medium was detected by Western blot analysis (Fig. 5A Inset). Wild-type 1-antitrypsin was secreted as the uncleaved form. In contrast, the majority of 1-PDX, RRRRSL and RRRRSA variants secreted were of the cleaved form. The AVRR variant secreted was detected in the uncleaved and cleaved forms. Furin activity in the conditioned medium was markedly inhibited with expression of 1-PDX, AVRR or RRRRSL variants, although the wild-type had no effect on the activity (Fig. 5A). Furin activity was not inhibited by the RRRRSA variant. The inhibitory effect of these variants was consistent with their ability to form SDS-stable complexes with furin. The inhibitory effect of the variants in vitro was examined using recombinant variants and furin. The conditioned medium from HEK293-EBNA cells transiently expressing 1-antitrypsin variants was mixed with conditioned medium from CHO cells stably expressing soluble furin and proteinase activity subsequently determined using pyr-RTKR-MCA. Western blot analysis indicated that the concentration of uncleaved form of wild-type 1-antitrypsin, 1-PDX and RRRRSA variants in the conditioned media tested in this experiment was about the same (Fig. 5B Inset). 1-PDX efficiently inhibited furin activity and the RRRRSA variant showed no inhibitory activity (Fig. 5B). Thus, the RRRRSA variant had neither inhibitory activity nor SDS-stable complex formation activity with respect to furin.
|
Inhibition of ECM-associated PACE4 and PC6 by the RRRRSA variant ex vivo
The inhibitory activity of the RRRRSA variant toward ECM-associated PACE4 and PC6 was examined. We previously showed that PACE4 and PC6 are heparin-binding proteins and are localized in the ECM (Tsuji et. al., 2003). ECM-associated PACE4 and PC6 were prepared from Cos-7 cells transiently expressing PACE4 or PC6 and 1-antitrypsin variant. As shown in Fig. 6A, laminin -5 fragments and its smaller degradation product (67' kDa) were detected in the solubilized ECM fraction, indicating that ECM proteins were recovered in this fraction. The RRRRSA variant efficiently inhibited both ECM-associated PACE4 and PC6, just as in the case with 1-PDX (Fig. 6B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementReferences |
|---|
Serpins can act as either inhibitors or substrates of proteinases (Bird et al., 1998
; Silverman et al., 2001; Huntington, 2006). With inhibition (inhibitory pathway), the proteinase initially forms a non-covalent Michaelis-like complex through interactions with residues flanking the scissile bond (P1-P1'). Attack of the active site serine on the scissile bond leads to formation of a covalent ester involving the active site serine of the proteinase and the backbone carbonyl of the P1 residue. Efficient serpin inhibitors form an essentially irreversible complex with the proteinase, and cleavage of the serpin by target proteinases at the reactive site loop (P1-P1') occurs slowly. In contrast, a substrate-like serpin is rapidly cleaved by the proteinase, resulting in dissociation of the complex and release of active proteinase and inactive serpin (substrate pathway). In an effort to create more selective inhibitors than 1-PDX, we prepared novel variants carrying mutations between P4 and P2'. As mentioned, the nature of any serpin, whether it acts as an inhibitor or substrate of a proteinase, is dependent on the susceptibility of cleavage at the reactive site loop by the proteinase. The RRRR sequence, introduced at the P4-P1 position, was more efficiently cleaved by furin, PACE4 and PC6 than was the RVPR sequence. Further, Leu at the P2' position was replaced by Ala, Phe, Trp, Lys, Glu or Gln.
Recently the crystal structures of furin and kexin, the homologue of furin in yeast, were determined (Henrich et al., 2003; Holyoak et al., 2003). Henrich et al. modeled the structure of the other SPC on the basis of these experimental structures of furin and kexin, and examined the structure of the substrate binding region of SPC (Henrich et al., 2005). The S6-S1 substrate binding region of furin, PACE4 and PC6 is characterized by a cluster consisting of an extremely large number of negatively charged residues. SPCs differ slightly from each other in terms of the number and distribution of these charges. Positively charged S1' is bordered by side chains of strictly conserved residues His194 and His364 and of residue 193, Arg193 in furin, and Lys193 in PACE4 and PC6. Therefore, P1'-Asp and Glu side chains could favorably interact with the His364 and Arg193 side chain in furin. Indeed, Ser, Tyr, Asp and Glu are frequently found at the P1' position of the processing site of proprotein. Amino acids with a hydrophobic aliphatic side chain are unsuitable (Nakayama, 1997). The S2' subsite of furin consists of a small depression extending toward the exposed indole moiety of Trp328. In all SPCs, the S2' subsite is well equipped to accommodate hydrophobic medium-sized residues. Thus, the geometry of the substrate binding regions of furin, PACE4 and PC6 are thought to be very similar.
When the S2' position of the RRRR (P4-P1) variant was mutated, the ability of 1-antitrypsin variants to form SDS-stable complexes with each SPC varied significantly. The Ala, Phe, Trp and Glu-P2'mutant of the RRRR variant did not form SDS-stable complexes with furin, unlike PACE4 and PC6, whereas Lys and Gln-P2' mutants formed complexes with furin, PACE4 and PC6 (Figs. 3 and 4). No inhibitory activity was found of the RRRRSA variant toward furin ex vivo or in vitro (Fig. 5). These results and the structural data of furin suggested that Ala, Phe, Trp and Glu at the P2' position were able to interact with the S2' site better than Leu. Furthermore, cleavage of RRRRSL, RRRRSK or RRRRSQ sequences by furin might be slower than cleavage of RRRRSA/F/W/E sequences. Similarly, cleavage of RRRRSA, RRRRSF, RRRRSW and RRRRSE sequences by PACE4 and PC6 were shown to be more efficient than cleavage of RRRRSK and RRRRSQ sequences (Fig. 4). The difference in selectivity toward the Ala, Phe, Trp and Glu-P2' mutants of furin, PACE4 and PC6 is thought to result from the cleavage velocity of the mutated reactive site loop by these enzymes. Since furin can cleave RRRRSA, RRRRSF, RRRRSW and RRRRSQ sequences much faster than PACE4 and PC6, the variant cleaved by furin (inactive serpin) might be released from the acyl-enzyme intermediate without SDS-stable acyl-complex formation. Similarly, the human 1-antitrypsin variant (1-AT-EK3, RERIRR358) with arginine at the P1, P2, P4 and P6 positions was shown to be cleaved more rapidly by furin than 1-PDX, although only a small amount of AT-EK3 formed an SDS-stable complex with the enzyme (Dufour et al., 2001). Thus, for the inhibitory activity, the sequence of the reactive site loop of the 1-antitrypsin variant does not necessarily reflect the substrate preference of furin. Our results highlighted the significance of the P2' residue of the 1-antitrypsin variant in SDS- and heat-stable acyl-complex formation with furin. In contrast, these variants (RRRRSA, RRRRSF, RRRRSW and RRRRSQ) act as both inhibitors and substrates of PACE4 or PC6.
Taken together, these findings clearly indicate that the RRRRSA variant is a highly selective inhibitor of PACE4 and PC6. However, it still remains uncertain whether the RRRRSA variant can specifically inhibit PACE4 or PC6 in the presence of furin when the RRRRSA variant is expressed following transfection of a suitable plasmid vector. Furin might cleave the RRRRSA variant rapidly in the trans-Golgi network (TGN) prior to inhibiting PACE4 or PC6 given that the RRRRSA variant is a good substrate of furin. Furthermore, the RRRRSA variant might effectively compete with endogenous substrate proteins for occupancy of the furin catalytic site as a competitive inhibitor, just as in the case of antitrypsin variant AT-EK3 (Dufour et al., 2001).
The marked advantage of the RRRRSA variant is its specific inhibition of PACE4 and PC6 on the ECM, with no inhibition of intracellular furin, following exogenous addition of this variant. Although furin is mainly localized to TGN, intracellular furin activity can be inhibited by the addition of 1-PDX to the culture medium. Since furin cycles to the cell surface and is retrieved to the TGN via endosomal compartments, 1-PDX can bind to furin sorted to the cell surface, resulting in decreased furin activity in the Golgi compartment and plasma membrane (Jean et al., 2000; Thomas, 2002). The RRRRSA variant did not form an SDS-stable complex with furin. Therefore, unlike the case with 1-PDX, intracellular furin activity is expected to be unaffected following the exogenously added RRRRSA variant, although the variant might competitively inhibit plasma membrane-bound furin inhibitor. In contrast, given that PACE4 and PC6 are secretory enzymes and localized in the ECM (Tsuji et al., 2003; Nour et al., 2005), the RRRRSA variant can effectively inhibit PACE4 and PC6 on the cell surface by acting as a suicide substrate as shown in Fig. 6.
To date, various synthetic SPC inhibitors such as peptidyl chloroalkylketones (Garten et al., 1994), the propeptide segment of SPC (Zhong et al., 1999) and polyarginine (Cameron et al., 2000) have been developed. However, PACE4- and PC6-specific inhibitors have not been reported. Use of the RRRRSA variant could be a powerful tool in delineating the specific role of PACE4 and PC6. PACE4 was reported to play a key role in the metastatic progression of certain tumors (Mahloogi et al., 2002; Bassi et al., 2005). Our findings provide useful insights in the further development of therapeutic agents for the treatment of diseases caused by PACE4 overexpression.
|
Footnotes |
|---|
Edited by Dick Janssen
|
Acknowledgement |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementReferences |
|---|
We thank Prof. K. Nakayama for the gifts of PC6A, the truncated soluble furin expression vectors and CHO cells expressing soluble furin. This work was supported by Knowledge Cluster Initiative from Ministry of Education, Science and Technology.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementReferences |
|---|
Anderson E.D., Thomas L., Hayflick J.S., Thomas G. J. Biol. Chem. (1993) 268:24887–24891.
Bando M., Matsuoka A., Tsuji A., Matsuda Y. J. Biochem. (2002) 132:127–134.
Bassi D.E., De Cicco R.L., Cenna J., Litwin S., Cukierman E., Klein-Szanto A.J.P. Cancer Res. (2005) 65:7310–7319.
Bird C.H., Sutton V.R., Sun J., Hirst C.E., Novak A., Kumar S., Trapani J.A., Bird P.I. Mol. Cell Biol. (1998) 18:6387–6398.
Blenis J., Hawkes S.P. Proc. Natl Acad. Sci. USA (1983) 80:770–774.
Bradford M.M. Anal. Biochem. (1976) 72:248–254.[CrossRef][Web of Science][Medline]
Cameron A., Appel J., Houghten R.A., Lindberg I. J. Biol. Chem. (2000) 275:36741–36749.
Dufour E.K., Denault J-B., Bissonnette L., Hopkins P.C.R., Lavigne P., Leduc R. J. Biol. Chem. (2001) 276:38971–38979.
Filion M., Bhakta V., Nguyen L.H., Liaw P.S., Sheffield W.P. Biochemistry (2004) 43:14864–14872.[CrossRef][Medline]
Garten W., Hallenberger S., Ortmann D., Schafer W., Vey M., Angliker H., Shaw E., Klenk H.D. Biochimie (1994) 76:217–225.[Medline]
Henrich S., Cameron A., Bourenkov G.P., Kiefersauer R., Huber R., Lindberg I., Bode W., Than M.E. Nat. Struct. Biol. (2003) 10:520–526.[CrossRef][Web of Science][Medline]
Henrich S., Lindberg I., Bode W., Than M.E. J. Mol. Biol. (2005) 345:211–227.[CrossRef][Web of Science][Medline]
Holyoak T., Wilson M.A., Fenn T.D., Kettner C.A., Petsko G.A., Fuller R.S., Ringe D. Biochemistry (2003) 42:6709–6718.[CrossRef][Medline]
Huntington J.A. Trends Biochem. Sci. (2006) 31:427–435.[CrossRef][Medline]
Jean F., Stella K., Thomas L., Liu G., Xiang Y., Reason A.J., Thomas G. Proc. Natl Acad. Sci. USA (1998) 95:7293–7298.
Jean F., Thomas L., Molloy S.S., Liu G., Jarvis M.A., Nelson J.A., Thomas G. Proc. Natl Acad. Sci. USA (2000) 97:2864–2869.
Khatib A-M., Siegfried G., Prat A., Louis J., Chrétien M., Metrakos P., Seidah N.G. J. Biol. Chem. (2001) 276:30686–30693.
Koide S., Yoshida I., Tsuji A., Matsuda Y. J. Biochem. (2003) 134:433–440.
Laemmli U.K. Nature (1970) 227:680–685.[CrossRef][Medline]
Mahloogi H., Bassi D.E., Klein-Szanto A.J.P. Carcinogenesis (2002) 23:565–572.
Mori K., Kii S., Tsuji A., Nagahama M., Imamaki A., Hayashi K., Akamatsu T., Nagamune H., Matsuda Y. J. Biochem. (1997) 121:941–948.
Nagahama M., Taniguchi T., Hashimoto E., Imamaki A., Mori K., Tsuji A., Matsuda Y. FEBS Lett. (1998) 434:155–159.[CrossRef][Web of Science][Medline]
Nakayama K. Biochem. J. (1997) 327:625–635.[Web of Science][Medline]
Nour N., Mayer G., Mort J.S., Salvas A., Mbikay M., Morrison C.J., Overall C.M., Seidah N.G. Mol. Biol. Cell. (2005) 16:5215–5226.
Parker K.A., Tollefsen D.M. J. Biol. Chem. (1985) 260:3501–3505.
Schechter I., Berger A. Biochem. Biophys. Res. Commun. (1967) 27:157–162.[CrossRef][Web of Science][Medline]
Sherman P.M., Lawrence D.A., Yang A.Y., Vandenberg E.T., Paielli D., Olson S.T., Shore J.D., Ginsburg D. J. Biol. Chem. (1992) 267:7588–7595.
Silverman G.A., et al. J. Biol. Chem. (2001) 276:33293–33296.
Thomas G. Nat. Rev. Mol. Cell Biol. (2002) 3:753–766.[CrossRef][Web of Science][Medline]
Tsuji A., Yoshida S., Hasegawa S., Bando M., Yoshida I., Koide S., Mori K., Matsuda Y. J. Biochem. (1999a) 126:494–502.
Tsuji A., Hashimoto E., Ikoma T., Taniguchi T., Mori K., Nagahama M., Matsuda Y. J. Biochem. (1999b) 126:591–603.
Tsuji A., Ikoma T., Hashimoto E., Matsuda Y. Protein Eng. (2002) 15:123–130.
Tsuji A., Sakurai K., Kiyokage E., Yamazaki T., Koide S., Toida K., Ishimura K., Matsuda Y. Biochim. Biophys. Acta (2003) 1645:95–104.[Medline]
Yoshida I., Koide S., Hasagawa S., Nakagawara A., Tsuji A., Matsuda Y. Biochem. J. (2001) 360:683–689.[CrossRef][Web of Science][Medline]
Zhong M., Munzer J.S., Basak A., Benjannet S., Mowla S., Decroly E., Chrétien M., Seidah N.G. J. Biol. Chem. (1999) 274:33913–33920.
Received April 27, 2006; revised January 15, 2007; accepted January 18, 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||