Advertisement

Biosynthesis and Enzymatic Characterization of Human SKI-1/S1P and the Processing of Its Inhibitory Prosegment*

Open AccessPublished:January 28, 2000DOI:https://doi.org/10.1074/jbc.275.4.2349
      Biochemical and enzymatic characterization of the novel human subtilase hSKI-1 was carried out in various cell lines. Within the endoplasmic reticulum of LoVo cells, proSKI-1 is converted to SKI-1 by processing of its prosegment into 26-, 24-, 14-, 10-, and 8-kDa products, some of which remain tightly associated with the enzyme. N-terminal sequencing and mass spectrometric analysis were used to map the cleavage sites of the most abundant fragments, which were confirmed by synthetic peptide processing. To characterize its in vitro enzymatic properties, we generated a secreted form of SKI-1. Our data demonstrate that SKI-1 is a Ca2+-dependent proteinase exhibiting optimal cleavage at pH 6.5. We present evidence that SKI-1 processes peptides mimicking the cleavage sites of the SKI-1 prosegment, pro-brain-derived neurotrophic factor, and the sterol regulatory element-binding protein SREBP-2. Among the candidate peptides encompassing sections of the SKI-1 prosegment, the RSLK137- andRRLL186-containing peptides were best cleaved by this enzyme. Mutagenesis of the latter peptide allowed us to develop an efficiently processed SKI-1 substrate and to assess the importance of several P and P′ residues. Finally, we demonstrate that, in vitro, recombinant prosegments of SKI-1 inhibit its activity with apparent inhibitor constants of 100–200 nm.
      aa
      amino acid(s)
      VV
      vaccinia virus
      PC
      precursor convertase
      ER
      endoplasmic reticulum
      TGN
      trans-Golgi network
      SREBP
      sterol regulatory element-binding protein
      BDNF
      brain-derived neurotrophic factor
      BTMD
      before transmembrane domain
      APMSF
      (4-amidinophenyl)methylsulfonyl fluoridephenylmethylsulfonyl fluoride
      PMSF
      phenylmethylsulfonyl fluoride
      TPCK
      N-tosyl-l-phenylalanine chloromethyl ketone
      TLCK
      N α-p-tosyl-l-lysine chloromethyl ketone
      SBTI
      soybean trypsin inhibitor
      PCR
      polymerase chain reaction
      RP
      reverse phase
      HPLC
      high performance liquid chromatography
      PAGE
      polyacrylamide gel electrophoresis
      Ab
      antibody
      WT
      wild type
      BFA
      brefeldin A
      s
      sense
      as
      antisense
      Fmoc
      N-(9-fluorenyl)methoxycarbonyl
      Tricine
      N-tris(hydroxymethyl)methylglycine
      MALDI-TOF
      matrix-assisted laser desorption/time of flight
      MES
      2-(N-morpholino)ethanesulfonic acid
      Abz
      O-aminobenzoic acid
      Y(NO2)
      3-nitrotyrosine
      Over the last 30 years (
      • Seidah N.G.
      • Day R.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Steiner D.F.
      ), our understanding of the complex cellular processing by limited proteolysis of inactive secretory precursors into active polypeptides and proteins has greatly expanded. It is now becoming clear that, following removal of the signal peptide, precursor cleavage can occur intracellularly, at the cell surface or within the extracellular milieu. The sites of cleavage are composed of either (i) single or pairs of basic residues (Lys or Arg) within the general motif (R/K)-(X)n-(K/R), where n = 0, 2, 4, or 6 and X is any amino acid (aa)1 except Cys, or (ii) hydrophobic (e.g. Leu, Phe, Val, or Met) and small aa such as Ala, Thr or Ser (
      • Seidah N.G.
      • Mbikay M.
      • Marcinkiewicz M.
      • Chrétien M.
      ). The former cleavage type occurs in many growth factors and their receptors, most polypeptide hormones and neuropeptide precursors, surface glycoproteins (including adhesion and viral envelope glycoproteins), as well as a host of other secretory proteins (
      • Seidah N.G.
      • Day R.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Steiner D.F.
      ). The latter type of cellular processing has been implicated in the generation of bioactive peptides (
      • Ling N.
      • Burgus R.
      • Guillemin R.
      ,
      • Burbach J.P.H.
      • Seidah N.G
      • Chrétien M.
      ,
      • Hudson P.
      • Haley J.
      • Cronk M.
      • Shine J.
      • Niall H.
      ), proteins (
      • Gupta S.K.
      • Hassel T.
      • Singh J.P.
      ), and transcription factors (
      • Duncan E.A.
      • Brown M.S.
      • Goldstein J.L.
      • Sakai J.
      ).
      Some of the proteinases involved in intracellular endoproteolytic events that result in cleavage at specific single or paired basic residues are members of a family of calcium-dependent serine proteinases related to the yeast subtilase kexin (
      • Seidah N.G.
      • Day R.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Steiner D.F.
      ,
      • Seidah N.G.
      • Mbikay M.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Siezen R.J.
      • Leunissen J.A.
      ). These dibasic- and monobasic-specific “precursor convertases” (PCs), of which seven mammalian members are presently known, comprise PC1 (PC3), PC2, furin (PACE), PC4, PC5 (PC6), PACE4, and PC7 (LPC, PC8). Each of these kexin-like subtilases contains a unique N-terminal prosegment that presumably functions both as an intramolecular chaperone and a proteinase inhibitor (
      • Seidah N.G.
      • Day R.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Steiner D.F.
      ,
      • Seidah N.G.
      • Mbikay M.
      • Marcinkiewicz M.
      • Chrétien M.
      ). Acting in concert, these enzymes determine the time and cellular location at which biologically active products are derived from inactive precursor proteins (
      • Seidah N.G.
      • Day R.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Steiner D.F.
      ,
      • Seidah N.G.
      • Mbikay M.
      • Marcinkiewicz M.
      • Chrétien M.
      ).
      Efforts to identify the proteinases responsible for the intracellular processing of precursors at hydrophobic or small aa have led to the recent cloning of a new subtilase called SKI-1 (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ) or S1P (
      • Sakai J.
      • Rawson R.B.
      • Espenshade P.J.
      • Cheng D.
      • Seegmiller A.C.
      • Goldstein J.L.
      • Brown M.S.
      ), whose aa sequence is highly conserved among human and rodent species. According to Siezen and Leunissen's classification (
      • Siezen R.J.
      • Leunissen J.A.
      ), this enzyme belongs to the pyrolysin branch of subtilases (compared with PCs, which are within the kexin branch). Tissue distribution analyses by both Northern blots and in situ hybridization reveal that SKI-1 mRNA is widely expressed (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ,
      • Nagase T.
      • Miyajima N.
      • Tanaka A.
      • Sazuka T.
      • Seki N.
      • Sato S.
      • Tabata S.
      • Ishikawa K.
      • Kawarabayasi Y.
      • Kotani H.
      ). We reported previously that human SKI-1 (hSKI-1) produces a 28-kDa product from the 32-kDa brain-derived neurotrophic factor precursor (proBDNF) via selective cleavage within the sequence R GL T ↓SL (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ). Independently, Sakai et al. demonstrated that hamster SKI-1/S1P is responsible for the site 1 cleavage of sterol regulatory element-binding proteins (SREBPs) (
      • Sakai J.
      • Rawson R.B.
      • Espenshade P.J.
      • Cheng D.
      • Seegmiller A.C.
      • Goldstein J.L.
      • Brown M.S.
      ), highlighting the critical role of SKI-1/S1P in the regulation of the synthesis and metabolism of cholesterol and fatty acids. In their model, SKI-1/S1P cleaves SREBP-2 at an R SV L ↓SF sequence within the lumen of the endoplasmic reticulum (ER). Mutational analyses demonstrated that the presence of Arg at the P4 substrate position is critical for cleavage, whereas the P1 Leu could be replaced by a number of other aa (
      • Duncan E.A.
      • Brown M.S.
      • Goldstein J.L.
      • Sakai J.
      ).
      In this work, we first present data regarding the cellular biosynthesis of membrane-bound hSKI-1 and its zymogen processing. Then, based on our previous discovery of a secreted (shed) form of hSKI-1 (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ), we produced a vaccinia virus (VV) recombinant of a soluble form of this enzyme, which, by analogy to rPC7 (
      • Munzer J.S.
      • Basak A.
      • Zhong M.
      • Mamarbachi A.
      • Hamelin J.
      • Savaria D.
      • Lazure C.
      • Benjannet S.
      • Chrétien M.
      • Seidah N.G.
      ), we called before the transmembrane domain SKI-1 (BTMD-SKI-1). This isoform, collected from cell media, was used to study the in vitro cleavage properties of this enzyme on a number of synthetic substrates. In addition, we present data on thein vitro inhibitory character of three prosegment constructs of SKI-1, which we obtained as bacterial recombinant proteins. Moreover, we examined the processing of hSKI-1 in LoVo cells infected with a VV recombinant as well as in a stable transfectant of HK293 cells (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ).

      DISCUSSION

      Limited proteolysis of inactive precursor proteins at sites marked by paired or multiple basic residues is a widespread process (
      • Seidah N.G.
      • Day R.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Steiner D.F.
      ). Less common is the recent finding that bioactive peptides or proteins can also be generated by limited proteolysis after either hydrophobic or small residues (
      • Seidah N.G.
      • Mbikay M.
      • Marcinkiewicz M.
      • Chrétien M.
      ). SKI-1 represents the first mammalian member of subtilisin-like processing enzymes with such substrate specificity (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ,
      • Sakai J.
      • Rawson R.B.
      • Espenshade P.J.
      • Cheng D.
      • Seegmiller A.C.
      • Goldstein J.L.
      • Brown M.S.
      ). It is a widely expressed enzyme (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ) that may play a crucial role in cholesterol and fatty acid metabolism (
      • Sakai J.
      • Rawson R.B.
      • Espenshade P.J.
      • Cheng D.
      • Seegmiller A.C.
      • Goldstein J.L.
      • Brown M.S.
      ). Due to its very recent discovery, information regarding its enzymatic properties, substrate specificity, and the function of its proregion have only begun to be addressed.
      Many peptidyl hydrolases, including subtilases, possess a prodomain that acts both as an intramolecular chaperone and a highly potent inhibitor of its associated protease (
      • Inouye M.
      ,
      • Gallagher T.
      • Gilliland G.
      • Wang L.
      • Bryan P.
      ). Activation of the enzyme typically requires release of the prosegment in an organelle-specific manner. For furin (
      • Anderson E.D.
      • Vanslyke J.K.
      • Thulin C.D.
      • Jean F.
      • Thomas G.
      ) the release occurs in the TGN, whereas for PC1 and PC2 (
      • Malide D.
      • Seidah N.G.
      • Chrétien M.
      • Bendayan M.
      ) it occurs in immature secretory granules. The data presented in this report demonstrate that SKI-1 is unique among the mammalian subtilases, since both the C-terminal and internal cleavages of its prosegment occur in the ER. Hence, this enzyme does not appear to require an acidic environment for activation, assuming, by analogy with other subtilases (
      • Seidah N.G.
      • Mbikay M.
      • Marcinkiewicz M.
      • Chrétien M.
      ), that prosegment release is the crucial step leading to zymogen activation. We propose the following sequence of events presumably leading to SKI-1 activation. 1) The signal peptide is removed in the ER by a signal peptidase cleavage at LVVLLC17 GKKHLG (Fig. 3C). 2) The prosegment is processed into a non-N-glycosylated polypeptide with an apparent molecular mass of ∼24–26-kDa (Fig. 2). 3) This prosegment is further processed into 14-, 10-, and 8-kDa intermediates (Fig. 2). While these multiple cleavages may be catalyzed by SKI-1 itself, the participation of other proteases cannot be excluded. The major cleavages leading to the formation of the ∼24- and ∼14-kDa products occur within 10 min, and the other secondary ones within 30 min (data not shown). Since treatment of cells with BFA did not significantly alter these processing events, they most likely occur in the ER (Fig.2). It is possible that the generation of prosegment fragments from the ∼24–26-kDa pro-form leads to a loss of inhibition in a fashion similar to that of subtilisin E (
      • Inouye M.
      ,
      • Gallagher T.
      • Gilliland G.
      • Wang L.
      • Bryan P.
      ). Indeed, our results demonstrate that, while the full-length prosegment is inhibitory, its ∼14-kDa product is not. Surprisingly, some pro-region-derived polypeptides are found associated with SKI-1 in cell culture media. Thus, in contrast to furin (
      • Anderson E.D.
      • Vanslyke J.K.
      • Thulin C.D.
      • Jean F.
      • Thomas G.
      ), the low pH and high Ca2+concentrations prevailing in the TGN do not lead to propeptide dissociation. High ionic concentrations (up to 1 m NaCl) such as those used in immunoprecipitation (Fig. 1 B) and metal chelation protein purification (Fig. 1 C) also do not disrupt the complex. It is only during RP-HPLC purification (Fig.3 A), in the presence of strong acids and organic solvents, that the prosegment peptides dissociate from SKI-1. These data suggest that hydrophobic interactions may be critical, as is the case for subtilisin (
      • Inouye M.
      ,
      • Gallagher T.
      • Gilliland G.
      • Wang L.
      • Bryan P.
      ).
      To distinguish the SKI-1 prosegment autoprocessing sites (C-terminal and internal) from several closely situated candidate sites, we employed a combination of mass spectrometry and synthetic peptide digestion. In the case of the C-terminal site, only one of three candidate peptides (III) was processed by SKI-1 (Table I), indicating that R RL L 186 RAIP is the most likely autoprocessing site. For the internal site, preliminary mass spectrometric data suggested three distinct cleavages occurring within the sequence PQRKVFRSLKYAESD142 (Fig.3 E). Two of the three possible sites (PQ R KVF133 ↓RSLKYAESD and PQR K VFR 134 SLKYAESD) appeared to satisfy the proposed SKI-1 recognition motif requiring a P4 basic residue (
      • Duncan E.A.
      • Brown M.S.
      • Goldstein J.L.
      • Sakai J.
      ). The third possibility (PQRKVF RSL136 ↓KYAESD) could be considered by assuming the cleavage actually occurred at PQRKVF R SLK 137 YAESD, followed by endogenous, basic carboxypeptidase removal of the C-terminal Lys residue (
      • Lei Y.
      • Xin X.
      • Morgan D.
      • Pintar J.E.
      • Fricker L.D.
      ). Assays carried out in vitrowith synthetic peptides corresponding to this region of the prosegment (peptides VIII and IX) produced the same cleavage products (data not shown), but only the PQRKVF R SLK 137 YAESD cleavage was unique to SKI-1. Thus, we propose that the aforementioned site is the most likely internal autoprocessing site, with the qualification that PQ R KVF133 ↓RSLKYAESD may occur to a lesser extent (see “Results” and Fig.4 B).
      Other information regarding the substrate preferences of SKI-1 was obtained by replacing the P3′ and P4′ Ile and Pro residues of the C-terminal cleavage site peptide (III) by Leu and Glu (peptides IV and V) to create a very well processed SKI-1 substrate. While it would appear that the presence of an acidic residue at P4′ significantly enhances the rate of substrate hydrolysis, it is also possible that the presence of Pro at P4′ hinders efficient substrate processing. The presence of similar acidic residues at the P3′ or P4′ position of the two confirmed substrates of SKI-1 (peptides I and II) as well as in the prosegment internal cleavage site RSLK137 YAES (Table I) lends support to the first argument. In addition to these residues, others also appear to play a role in SKI-1 substrate cleavage catalysis. The peptide pairs IV/V and X/XI both point to influences of positions N-terminal to the P4 residue. Interestingly, the efficiency of the truncated C-terminal peptide V is lower than that of peptide IV, whereas that of the truncated internal (quenched) peptide XI is higher. Taken together, these data indicate the importance of aa at both the P and P′ positions in SKI-1-mediated substrate hydrolysis.
      The data presented in Fig. 6 indicate that SKI-1 functions most efficiently near neutral pH and at 2–3 mmCa2+. This is in general agreement with the conditions that reportedly prevail in the ER (
      • Kendall J.M.
      • Badminton M.N.
      • Dormer R.L.
      • Campbell A.K.
      ,
      • Sambrook J.F.
      ). However, closer examination of the data reveal that the pH optimum of SREBP-2 cleavage (peptide II, Fig. 6A) is actually 6.5, an observation that we confirmed using our purified SKI-1 preparation (data not shown). This suggests that the processing of SREBP might occur outside of the ER, perhaps in the Golgi where pH values of ∼6.5 have recently been reported (
      • Llopis J
      • McCaffery J.M.
      • Miyawaki A
      • Farquhar M.G.
      • Tsien R.Y.
      ,
      • Kim J.H.
      • Johannes L.
      • Goud B.
      • Antony C.
      • Lingwood C.A.
      • Daneman R.
      • Grinstein S.
      ). Indeed, there is now cellular evidence suggesting that SREBP cleavage may occur in the Golgi rather than in the ER (
      • Nohturfft A
      • DeBose-Boyd R.A.
      • Scheek S.
      • Goldstein J.L.
      • Brown M.S.
      ,
      • Nohturfft A
      • Brown M.S.
      • Goldstein J.L.
      ). The pH optimum of SKI-1 appears to be dependent on the substrate employed; proBDNF (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ) and its related peptide (I), appear to be well cleaved even at pH 5.5, suggesting that it could cleave this (and possibly other substrates) in acidic endosome-like compartments where it was previously localized (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ). On the other hand, cleavage of the internal, autocatalytic, prosegment processing site PQRKVF R SLK 137 YAESD (Fig. 4 B) is optimal at pH 8 (data not shown), implying that this event, as we concluded from our biosynthesis assays, takes place most effectively in the ER. Overall, the pH and Ca2+profiles of SKI-1 resemble those of the constitutively secreted PCs (
      • Seidah N.G.
      • Day R.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Munzer J.S.
      • Basak A.
      • Zhong M.
      • Mamarbachi A.
      • Hamelin J.
      • Savaria D.
      • Lazure C.
      • Benjannet S.
      • Chrétien M.
      • Seidah N.G.
      ). The inhibitor profile of SKI-1 (Ref.
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      , Table III), showing that enzymatic activity is significantly inhibited by EDTA, EGTA, and only high concentrations of o-phenanthroline, tend to discount the likelihood that SKI-1 is a transition metal-dependent proteinase. In fact, SKI-1 activity is inhibited by low concentrations of certain transition metals, such as Cu2+ and Zn2+.
      Directed by the observation that peptides containing the primary processing site of the prosegment of PC1 are potent inhibitors of its activity, and that the C-terminal basic residues of furin and PC7 are essential for enzyme inhibition (
      • Boudreault A.
      • Gauthier D.
      • Lazure C.
      ,
      • Zhong M.
      • Munzer J.S.
      • Basak A.
      • Benjannet S.
      • Mowla S.J.
      • Decroly E.
      • Chrétien M.
      • Seidah N.G.
      ), we assessed the inhibitory potency of three SKI-1 recombinant propeptides (Fig. 7 A). All of these end at sequences near the RRLL186RA cleavage site. Interestingly, the three prosegments displayed comparable inhibitory potencies (Table V). Compared with proPC1 (
      • Boudreault A.
      • Gauthier D.
      • Lazure C.
      ), pro-furin and proPC7 (
      • Zhong M.
      • Munzer J.S.
      • Basak A.
      • Benjannet S.
      • Mowla S.J.
      • Decroly E.
      • Chrétien M.
      • Seidah N.G.
      ), the K i(app) values (Table V) are up to 250-fold higher. This suggests that the prosegment of SKI-1, although potentially inhibitory in vivo, may function more as a chaperone, catalyzing the productive folding of SKI-1. Indeed, since SKI-1 may be active in the ER (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ,
      • Sakai J.
      • Rawson R.B.
      • Espenshade P.J.
      • Cheng D.
      • Seegmiller A.C.
      • Goldstein J.L.
      • Brown M.S.
      ), whereas the PCs are not (
      • Munzer J.S.
      • Basak A.
      • Zhong M.
      • Mamarbachi A.
      • Hamelin J.
      • Savaria D.
      • Lazure C.
      • Benjannet S.
      • Chrétien M.
      • Seidah N.G.
      ,
      • Anderson E.D.
      • Vanslyke J.K.
      • Thulin C.D.
      • Jean F.
      • Thomas G.
      ), the lower inhibitory potency of the prosegment of SKI-1 may be adapted to the conditions prevailing in this cellular compartment. In the case of PCs, highly effective inhibition by the prosegment may be needed in order to ensure that these enzymes are activated only when they reach the TGN or secretory granules (
      • Seidah N.G.
      • Day R.
      • Marcinkiewicz M.
      • Chrétien M.
      ,
      • Steiner D.F.
      ,
      • Seidah N.G.
      • Mbikay M.
      • Marcinkiewicz M.
      • Chrétien M.
      ). The 14-kDa fragment, which represents the major secreted form of the prosegment, is tightly associated with SKI-1 (Fig.1 C), yet it is not inhibitory (data not shown). Accordingly, this segment may serve a chaperonin-like function similar to that reported for the N-terminal 150 aa of 7B2 toward proPC2 (
      • Muller L.
      • Zhu X.R.
      • Lindberg I.
      ,
      • Benjannet S.
      • Mamarbachi A.M.
      • Hamelin J.
      • Savaria D.
      • Munzer J.S.
      • Chrétien M.
      • Seidah N.G.
      ).
      In conclusion, the present work firmly establishes that SKI-1 is a Ca2+-dependent subtilase with a reasonably neutral pH optimum, depending on the substrate employed. We also demonstrate that SKI-1 can cleave substrates C-terminal to Thr, Leu, and Lys residues, thus providing direct, in vitro evidence that it is a candidate converting enzyme responsible for the generation of 28-kDa proBDNF (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ) and SREBP-2 processing at site 1 (
      • Sakai J.
      • Rawson R.B.
      • Espenshade P.J.
      • Cheng D.
      • Seegmiller A.C.
      • Goldstein J.L.
      • Brown M.S.
      ). For efficient cleavage, it appears that substrates should contain a basic residue at P4 and an aliphatic one at P2 (Table I). Furthermore, aa at the P3′ and P4′ positions seem to exert an important discriminatory effect. The best substrate tested so far is the quenched fluorogenic substrate Abz- RSLK 234 YAESDY(NO2)A, thereby providing a convenient and sensitive assay for SKI-1 activity. The present data demonstrate that only the full-length SKI-1 prosegment is inhibitory. Thus, overexpression of this prosegment in cell lines may provide a novel method for inhibiting the cellular activity of this enzyme in a fashion similar to that of over-expressed profurin and proPC7 (
      • Zhong M.
      • Munzer J.S.
      • Basak A.
      • Benjannet S.
      • Mowla S.J.
      • Decroly E.
      • Chrétien M.
      • Seidah N.G.
      ). Finally, it is anticipated that precursor substrates other than the sterol-regulating SREBPs (
      • Duncan E.A.
      • Brown M.S.
      • Goldstein J.L.
      • Sakai J.
      ) and the neurotrophin proBDNF (
      • Seidah N.G.
      • Mowla S.J.
      • Hamelin J.
      • Mamarbachi A.M.
      • Benjannet S.
      • Toure B.B.
      • Basak A.
      • Munzer J.S.
      • Marcinkiewicz J.
      • Zhong M.
      • Barale J.C.
      • Lazure C.
      • Murphy R.A.
      • Chrétien M.
      • Marcinkiewicz M.
      ) will be identified, thereby extending the spectrum of activity of this unique and versatile enzyme.

      Addendum

      While this manuscript was under review, two articles describing the processing, purification and in vitro activity of hamster SKI-1/S1P were published (
      • Espenshade P.J.
      • Cheng D.
      • Goldstein J.L.
      • Brown M.S.
      ,
      • Cheng D.
      • Espenshade P.J.
      • Slaughter C.A.
      • Jaen J.C.
      • Brown M.S.
      • Goldstein J.L.
      ). On most points, our results are in close agreement with those recently published. Thus, these authors characterized the processing of the SKI-1/S1P prosegment, proposing that the ER is the major site of autocatalytic activation of SKI-1 at the same cleavage sites as we present here. They also went on to purify a soluble form of the enzyme, showing that it correctly processes SREBP-2-derived peptides as well as a 16 residue peptide spanning the internal prosegment cleavage site. In addition, they find that cleavage of fluorogenic RSLK-MCA peptide derived from the same sequence is optimal at ∼3 mmCa2+ at slightly alkaline pH. Discrepancies such as the lack of detectable shed SKI-1/S1P, multiple secreted prosegment forms, and a different signal peptidase site can most likely be attributed to the different cell types and species employed in the two studies.

      Acknowledgments

      We thank A. M. Mamarbachi, J. Hamelin, O. Théberge, A. Lemieux, and D. Gauthier for their technical help throughout this study. We are grateful to M. Zhong, A. Blanc, and E. Bergeron for the initial constructions of the prosegments and/or for helpful advice and discussions. The secretarial assistance of Sylvie Emond is greatly appreciated.

      REFERENCES

        • Seidah N.G.
        • Day R.
        • Marcinkiewicz M.
        • Chrétien M.
        Ann. N. Y. Acad. Sci. 1998; 839: 9-24
        • Steiner D.F.
        Curr. Opin. Chem. Biol. 1998; 2: 31-39
        • Seidah N.G.
        • Mbikay M.
        • Marcinkiewicz M.
        • Chrétien M.
        Hook V.Y.H Proteolytic and Cellular Mechanisms in Prohormone and Neuropeptide Precursor Processing. R. G. Landes Co., Georgetown, TX1998: 49-76
        • Ling N.
        • Burgus R.
        • Guillemin R.
        Proc. Natl. Acad. Sci. U. S. A. 1976; 73: 3042-3046
        • Burbach J.P.H.
        • Seidah N.G
        • Chrétien M.
        Eur. J. Biochem. 1986; 156: 137-142
        • Hudson P.
        • Haley J.
        • Cronk M.
        • Shine J.
        • Niall H.
        Nature. 1981; 291: 127-131
        • Gupta S.K.
        • Hassel T.
        • Singh J.P.
        Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 7799-7803
        • Duncan E.A.
        • Brown M.S.
        • Goldstein J.L.
        • Sakai J.
        J. Biol. Chem. 1997; 272: 12778-12785
        • Siezen R.J.
        • Leunissen J.A.
        Protein Sci. 1997; 6: 501-523
        • Seidah N.G.
        • Mowla S.J.
        • Hamelin J.
        • Mamarbachi A.M.
        • Benjannet S.
        • Toure B.B.
        • Basak A.
        • Munzer J.S.
        • Marcinkiewicz J.
        • Zhong M.
        • Barale J.C.
        • Lazure C.
        • Murphy R.A.
        • Chrétien M.
        • Marcinkiewicz M.
        Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 1321-1326
        • Sakai J.
        • Rawson R.B.
        • Espenshade P.J.
        • Cheng D.
        • Seegmiller A.C.
        • Goldstein J.L.
        • Brown M.S.
        Mol. Cell. 1998; 2: 505-514
        • Nagase T.
        • Miyajima N.
        • Tanaka A.
        • Sazuka T.
        • Seki N.
        • Sato S.
        • Tabata S.
        • Ishikawa K.
        • Kawarabayasi Y.
        • Kotani H.
        DNA Res. 1995; 2: 37-43
        • Munzer J.S.
        • Basak A.
        • Zhong M.
        • Mamarbachi A.
        • Hamelin J.
        • Savaria D.
        • Lazure C.
        • Benjannet S.
        • Chrétien M.
        • Seidah N.G.
        J. Biol. Chem. 1997; 272: 19672-19681
        • Rovère C.
        • Luis J.
        • Lissitzky J.-C.
        • Basak A.
        • Marvaldi J.
        • Chrétien M.
        • Seidah N.G.
        J. Biol. Chem. 1999; 274: 12461-12467
        • Basak A.
        • Boudreault A.
        • Chen A.
        • Chrétien M.
        • Seidah N.G.
        • Lazure C.
        J. Pept. Sci. 1995; 1: 385-395
        • Basak A.
        • Ernst B.
        • Brewer D.
        • Seidah N.G.
        • Munzer J.S.
        • Lazure C.
        • Lajoie G.A.
        J. Pept. Res. 1997; 49: 596-603
        • Jean F.
        • Boudreault A.
        • Basak A.
        • Seidah N.G.
        • Lazure C.
        J. Biol. Chem. 1995; 270: 19225-19231
        • Hooper N.M.
        • Karran E.H.
        • Turner A.J.
        Biochem. J. 1997; 321: 265-279
        • Lippincott-Schwartz J.
        • Yuan L.
        • Tipper C.
        • Amherdt M.
        • Orci L.
        • Klausner R.D.
        Cell. 1991; 67: 601-616
        • Gram H.
        • Ramage P.
        • Memmert K.
        • Gamse R.
        • Kocher P.
        Bio/Technology. 1994; 12: 1017-1023
        • Rittenhose J.
        • Marcus F.
        Anal. Biochem. 1984; 138: 442-448
        • Mizuno K.
        • Nakamura T.
        • Ohshima T.
        • Tanaka S.
        • Matsuo H.
        Biochem. Biophys. Res. Commun. 1989; 159: 305-311
        • Lei Y.
        • Xin X.
        • Morgan D.
        • Pintar J.E.
        • Fricker L.D.
        DNA Cell Biol. 1999; 18: 175-185
        • Inouye M.
        Enzyme. 1991; 45: 314-321
        • Gallagher T.
        • Gilliland G.
        • Wang L.
        • Bryan P.
        Structure. 1995; 3: 907-914
        • Anderson E.D.
        • Vanslyke J.K.
        • Thulin C.D.
        • Jean F.
        • Thomas G.
        EMBO J. 1997; 16: 1508-1518
        • Malide D.
        • Seidah N.G.
        • Chrétien M.
        • Bendayan M.
        J. Histochem. Cytochem. 1995; 43: 11-19
        • Kendall J.M.
        • Badminton M.N.
        • Dormer R.L.
        • Campbell A.K.
        Anal. Biochem. 1994; 221: 173-181
        • Sambrook J.F.
        Cell. 1990; 61: 197-199
        • Llopis J
        • McCaffery J.M.
        • Miyawaki A
        • Farquhar M.G.
        • Tsien R.Y.
        Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 6803-6808
        • Kim J.H.
        • Johannes L.
        • Goud B.
        • Antony C.
        • Lingwood C.A.
        • Daneman R.
        • Grinstein S.
        Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 2997-3002
        • Nohturfft A
        • DeBose-Boyd R.A.
        • Scheek S.
        • Goldstein J.L.
        • Brown M.S.
        Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 11235-11240
        • Nohturfft A
        • Brown M.S.
        • Goldstein J.L.
        Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 12848-12853
        • Boudreault A.
        • Gauthier D.
        • Lazure C.
        J. Biol. Chem. 1998; 273: 31574-31580
        • Zhong M.
        • Munzer J.S.
        • Basak A.
        • Benjannet S.
        • Mowla S.J.
        • Decroly E.
        • Chrétien M.
        • Seidah N.G.
        J. Biol. Chem. 1999; 274: 33913-33920
        • Muller L.
        • Zhu X.R.
        • Lindberg I.
        J. Cell Biol. 1997; 139: 625-638
        • Benjannet S.
        • Mamarbachi A.M.
        • Hamelin J.
        • Savaria D.
        • Munzer J.S.
        • Chrétien M.
        • Seidah N.G.
        FEBS Lett. 1998; 428: 37-42
        • Espenshade P.J.
        • Cheng D.
        • Goldstein J.L.
        • Brown M.S.
        J. Biol. Chem. 1999; 274: 22795-22804
        • Cheng D.
        • Espenshade P.J.
        • Slaughter C.A.
        • Jaen J.C.
        • Brown M.S.
        • Goldstein J.L.
        J. Biol. Chem. 1999; 274: 22805-22812