- Email: [email protected]
Cell-surface expression of a new splice variant of the mouse signal peptide peptidase
Biochimica et Biophysica Acta 1759 (2006) 159 – 165 http://www.elsevier.com/locate/bba
Cell-surface expression of a new splice variant of the mouse signal peptide peptidase Jens Urny, Irm Hermans-Borgmeyer, H. Chica Schaller ⁎ Zentrum für Molekulare Neurobiologie, Universität Hamburg, Martinistr. 52, 20246 Hamburg, Germany Received 2 December 2005; received in revised form 17 February 2006; accepted 28 February 2006 Available online 31 March 2006
Abstract The signal peptide peptidase (SPP) is an intramembrane-cleaving aspartyl protease that acts on type II transmembrane proteins. SPP substrates include signal peptides after they have been cleaved from a preprotein, hence the name. The known SPP isoform, which we renamed SPPα, contains an endoplasmic reticulum retention signal at the carboxy terminus. We found a new splice variant, SPPβ, with an additional in-frame exon inserted between exons 11 and 12 of SPPα. Insertion of the new exon led to a complete change in the amino-acid sequence of the carboxy tail. A stop codon within this new exon resulted in silencing of exon 12 and eliminated the endoplasmic reticulum retention signal. The new SPP isoform predominantly localised to the cell surface in contrast to the more restricted localisation of SPPα in the endoplasmic reticulum. Differential expression in mouse tissues and in subcellular compartments suggests new functions for SPP in addition to cleaving signal peptides. © 2006 Elsevier B.V. All rights reserved. Keywords: Signal peptide peptidase; SPP; Cell-surface localisation; ER-retention signal; Intramembrane cleavage; Aspartyl protease
1. Introduction Cleavage of proteins within transmembrane domains plays an ever increasing role in cellular communication, development and pathophysiology [1–3]. The most outstanding examples are presenilins which, as active components of the γ-secretase system, on one hand, control development by influencing Notch signalling and on the other hand cell death in Alzheimer's disease by cleaving the amyloid precursor protein to release the neurotoxic fragment Aβ1–42 [4]. Another such intramembranecleaving protease is the signal peptide peptidase (SPP), which derived its name by acting on some signal peptides after prior cleavage by the signal peptidase [5]. Functions for SPP comprise immune surveillance, processing of the core domain of Hepatitis C virus and endoplasmic reticulum (ER) quality control to remove misfolded membrane proteins [2,6]. SPP, four SPP-like proteases and presenilins belong to the family of multipass intramembrane-cleaving aspartyl proteases, which share limited sequence homology, but are distinguished by a similar overall topology with the active aspartyl site motifs ⁎ Corresponding author. Fax: +49 40 42803 5101. E-mail address: [email protected] (H.C. Schaller). 0167-4781/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.bbaexp.2006.02.007
YD and GXGD located in adjacent transmembrane helices [7]. The loop between these active-site containing transmembrane domains of SPP and SPP-like proteases extends into the ER lumen or extracellular space and that of presenilins into the cytoplasm [8,9]. This inverted organisation of the active motifs is interpreted to imply reverse orientation of substrates, such that SPP and SPP-like proteases cleave type II and presenilins type I transmembrane proteins. Presenilins undergo proteolysis and require partners for activation, SPP and SPP-like proteases seem to act alone as uncleaved monomers or homodimers [4,10]. For the respective proteolytic functions, the aspartates are absolutely required, as evidenced by mutation and by common inhibitors targeting the active site [11,12]. For SPP, the number of predicted transmembrane domains varies between seven and nine. Unequivocal is the topology of the glycosylated amino terminus as extending into the lumen [2,10] and the carboxy terminus into the cytosol [13] demanding an unequal number of transmembrane helices for the complete structure. SPP was found to be expressed in many human and mouse tissues including the brain [13]. Its function as signal peptide peptidase suggests an exclusive intracellular localisation in the ER [13], probably due to the presence of the carboxyterminal KKEK motif as ER-retention signal [14].
160
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
We describe in this paper the identification and expression of a splice variant of SPP, SPPβ. Instead of the known carboxy terminus of SPPα, SPPβ contains a new extended sequence without the KKEK ER-retention signal. Accordingly, its localisation is not restricted to the ER, but allows trafficking to the cell-surface suggesting functions for SPPβ different from those of SPPα. Differential expression of the two splice variants in brain and various organs of the mouse supports this notion.
from exon 1 and nucleotides 1256–1279 from exon new served as forward and reverse primers, respectively. All nucleotides are derived from mouse SPP (accession number DQ168449). For in situ analysis sense and antisense probes were labeled with [α-35S]UTP. Probes, specific for SPPβ, comprised nucleotides 1172–1362, and a probe common for SPPα and SPPβ covered nucleotides 1370–1559 of mouse SPPβ (accession number DQ168449). Cryosections of 10 μm from adult mouse brain and different mouse organs were processed as described previously (13).
3. Results 2. Materials and methods
3.1. Identification and cloning of SPPβ 2.1. Cloning and expression of SPPβ Mouse SPPβ was found by sequence analysis of an EST clone from mouse liver (accession number AW259871) and human SPPβ by sequencing an EST from human retina (accession number AA019983), both kindly provided by RZBD (Deutsches Ressourcenzentrum für Genforschung GmbH, Berlin, Germany). For heterologous expression full-length SPPα and SPPβ of mouse were inserted with and without FLAG tag at the amino terminus into the eukaryotic expression vector pcDNA3 (Invitrogen, Karlsruhe, Germany).
2.2. Subcellular localisation SPPα and SPPβ were transiently expressed in COS-7 and CHO-K1 cells. To monitor cell-surface expression living cells were incubated 48 h after transfection with first antibodies for 30 min on ice before fixation with 4% paraformaldehyde. For intracellular localisation cells were fixed with 4% paraformaldehyde, permeabilised with 0.05% Triton X-100, and then reacted with first antibodies. As primary antiserum, specific for the carboxy terminus of SPPα, C18 was used which we had produced in rabbits against a recombinant protein containing the last 25 carboxy-terminal amino acids of SPPα [13]. The antiserum obtained after the fifth booster, C18.5, was diluted 1:3000 both for western blotting and immunocytochemistry. The monoclonal antibody against FLAG (M2) was from Sigma-Aldrich (Taufkirchen, Germany) and applied at a dilution of 1:200. As secondary antibodies served Alexa Fluor 488 goat antirabbit (Invitrogen) and Cy3 anti-mouse (Amersham Biosciences, Freiburg, Germany), both used at a dilution of 1:2000.
2.3. Biotinylation and western blotting CHO-K1 cells were transiently transfected by electroporation with SPPα and SPPβ constructs containing FLAG tags at the amino terminus. 48 h after transfection cells were treated for 30 min at room temperature with 1 mM S-NHSbiotin (Perbio Science, Bonn, Germany), which is membrane-impermeable. The cell pellets were homogenised with a pestle in buffer containing 0.28 M sucrose, 25 mM K3PO4, 5 mM EDTA and protease-inhibitor cocktail (Roche Diagnostics, Mannheim, Germany). The nuclear fraction was pelleted by centrifugation at 1000×g for 5 min at 4 °C. The membrane fraction was pelleted from the supernatant by centrifugation at 150000×g for 1.5 h at 4 °C and solubilised by shaking for 2 h at 4 °C in 1% Triton X-100, 0.5% NP40, 150 mM NaCl. 7 mM EDTA, 1 mM EGTA, 10 mM Tris–HCl, pH 7.4, and protease-inhibitor cocktail. After centrifugation at 150000×g for 1 h at 4 °C the supernatant was diluted 5fold to reduce the detergent concentration and incubated with anti-FLAG M2 agarose (Sigma-Aldrich) overnight at 4 °C, centrifuged, washed and applied to SDS-PAGE containing 8 M urea. Biotinylated proteins were detected with an avidin-peroxidase conjugate (Bio-Rad, München, Germany) using West Dura Super Signal as peroxidase substrate (Perbio Science).
2.4. Tissue distribution and in situ analysis A mouse multiple-tissue cDNA panel (Clontech, Heidelberg, Germany) served as template to amplify SPPα- and SPPβ-specific sequences by PCR. To amplify both isoforms simultaneously, we used nucleotides 318–341 from exon 1 and nucleotides 994–1018 from exon 11 as forward and nucleotides 1429–1452 from exon 12 as reverse primers. To amplify SPPβ specifically, nucleotides 318–341
In search for a full-length cDNA of human SPP twenty ESTs were sequenced. One EST from human retina (accession number AA019983) differed by containing an additional exon of 200 nucleotides between exon 11 and exon 12 of the original sequence (Fig. 1A). Likewise, an EST from mouse liver (accession number AW259871) also contained this exon. The mouse EST comprised 1652 nucleotides, included start and stop codon of full-length SPPβ, and harboured additional sequences from the 5′- and 3′-untranslated regions (submitted to GenBank under the accession number DQ168449). A comparison between mouse SPPα (Fig. 1B) and SPPβ (Fig. 1C) shows that all features typical for intramembrane-cleaving aspartyl proteases are preserved (highlighted in grey in Fig. 1B), and that insertion of the new exon behind the last putative transmembrane domain gives rise to a carboxy-terminal tail with a completely different amino-acid sequence (Fig. 1C). The new carboxy tail of SPPβ also differs in size coding for 49 amino acids compared to 33 in SPPα. The outstanding new feature of the splice variant is that it lacks the carboxy-terminal ER-retention signal KKEK of SPPα (Fig. 1B). The derived amino-acid sequences of the new carboxy tail for SPPβ of mouse and human differ by a single amino-acid exchange (Fig. 1D). The human sequence was submitted to GenBank under the accession number DQ168450. Exon–intron boundaries were verified by comparison with the human and mouse genome sequences. 3.2. Subcellular localisation of SPPα and SPPβ The absence of the ER-retention signal KKEK in SPPβ indicated a possible difference in subcellular localisation of the two splice variants. To study this SPPα and SPPβ sequences were integrated into eukaryotic expression vectors provided with and without FLAG tags at the extracellular amino-terminal domains, respectively. For visualisation of protein expression served monoclonal antibodies against the FLAG epitope and a polyclonal antiserum, produced against the last 25 carboxyterminal amino acids of human SPPα [13]. Transient cotransfection of SPPα without and SPPβ with FLAG tag allowed specific recognition of each splice variant (Fig. 2A). A striking difference was found in subcellular localisation of the two proteins, SPPβ (green) being very strongly expressed at the cell surface, whereas SPPα (red) showed a predominant internal staining pattern indicative of ER localisation. This confirms earlier findings demonstrating co-staining of SPPα with ER markers [3,13]. Cell-surface expression of SPPβ on living cells
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
161
Fig. 1. Genomic organisation, nucleotide and deduced amino-acid sequence of SPPα and SPPβ. (A) Exon and intron boundaries were determined by blasting new and old SPP sequences against human and mouse genome databases. The new additional exon of SPPβ (shaded in grey) is located both in mouse and human between exons 11 and 12 of SPPα. (B) Up to the splice at the end of exon 11, the sequence of SPPα and SPPβ is identical, shown in panels B and C for the mouse. The splice site, where the new exon of SPPβ is integrated, is marked by a black arrow head. For the common sequence of SPPα and SPPβ, the start codon is printed in bold, the ER-retention signal is underlined and the carboxy-terminal stop codon of SPPα is indicated by an asterisk. The motifs typical for aspartyl proteases of the SPP and presenilin family are shaded in grey. (C) Start and stop of the new exon of SPPβ are marked by black arrow heads. The nucleotides of the new exon are highlighted by bold letters. (D) Mouse and human amino-acid sequences of SPPβ are compared showing only a single amino-acid exchange.
162
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
present as monomer and dimer and as additional band of intermediate size. Reaction with avidin showed that only SPPβ was present at the cell surface, and that cell-surface expression was restricted to the monomeric and intermediate forms of SPPβ. 3.3. Differential expression of SPPα and SPPβ mRNAs in mouse tissues
Fig. 2. Subcellular localisation of SPPα and SPPβ. (A) Confocal micrographs of COS-7 cells overexpressing FLAG-tagged SPPβ together with untagged SPPα are shown as overlays. 48 h after transfection cells were fixed, permeabilised and the proteins visualised by antibodies against the FLAG epitope expressed by SPPβ (shown in green) and by the SPPα-specific antiserum against the carboxy terminus (shown in red). Cell-surface localisation is clearly detectable for SPPβ, but not for SPPα. (B) Living, non-permeabilised CHO-K1 cells overexpressing N-terminally FLAG-tagged SPPβ were reacted for 30 min on ice with antibodies against the FLAG tag before fixation.
was demonstrated by transiently expressing SPPβ with an amino-terminal FLAG-tag in CHO-K1 cells and immunostaining without permeabilisation. The FLAG immunoreactivity was nicely visible at the outer cell membrane (Fig. 2B). To confirm these results membrane fractions of CHO-K1 cells transiently transfected with FLAG-tagged SPPα and SPPβ constructs were analysed by western blotting after cell-surface labelling with membrane-impermeable biotin. After solubilisation and immunoprecipitation with anti-FLAG agarose samples were separated by SDS-PAGE containing 8 M urea. The immunoblots were reacted with anti-FLAG antibodies (Fig. 3, left panel) or with an avidin-peroxidase conjugate (Fig. 3, middle panel). We found that SPPα appears as monomer and dimer of the expected sizes of 50 and 100 kDa. Likewise, SPPβ was also
Fig. 3. Western blot analysis of SPPα and SPPβ heterologously expressed in CHO-K1 cells. CHO-K1 cells transfected with SPPα and SPPβ constructs containing a FLAG tag at the amino terminus were cell-surface biotinylated with membrane-impermeable biotin 48 h after transfection. The solubilised membrane fraction was immunoprecipitated with anti-FLAG agarose and separated on urea containing SDS-PAGE. The immunoblots were reacted with anti-FLAG antibodies (left panel) or with an avidin–peroxidase conjugate (middle panel). Ponceau staining served as loading control (right panel).
To determine the expression of mouse SPPα and SPPβ mRNAs in different tissues, a mouse multiple-tissue cDNA panel was probed by PCR with gene-specific primers derived from exon 1 and exon 11 as forward and exon 12 as reverse primer, with the idea to amplify simultaneously both SPPα and SPPβ. In addition, primers over exon 1 and exon new were used to amplify SPPβ specifically (Fig. 1). The amplification reaction yielded the expected products differing in molecular mass by 200 nucleotides, which are present in SPPβ and not in SPPα (Fig. 4). In most of the tissues analysed, SPPα was more strongly expressed than SPPβ. Thus spleen, lung, liver, kidney of adult mouse and embryos of various ages showed strong and predominant expression of SPPα, in heart and testis the expression levels were comparable, and only in brain and skeletal muscle SPPβ expression seemed to prevail over that of SPPα (Fig. 4). To investigate the differential expression in more detail, in situ hybridisations were performed on tissue sections of adult mice. An RNA transcript covering 189 nucleotides of the new exon served as SPPβ-specific probe. Since all exons of SPPα were also present in SPPβ, a specific probe for SPPα could not be constructed (Fig. 1). As control of similar size a fragment of 190
Fig. 4. Expression of SPPα and SPPβ mRNAs in different mouse tissues. A mouse multiple-tissue cDNA panel was analysed by PCR with gene-specific primers from exon 1 as forward and from exon 12 and exon new as reverse primers. In addition primers from exon 11 and exon 12 were used as control to amplify a smaller product. Bands corresponding to SPPα and SPPβ are indicated by arrows. Amplification of GAPDH served as loading control.
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
163
nucleotides from exon 12 was used, which should react with the RNA of both isoforms. Due to the shortness of the probes the signal to noise ratio was suboptimal. Expression analysis in peripheral organs showed that the common probe reacted strongly with liver and kidney (Fig. 5A and C), whereas no signal was detectable with the SPPβ-specific probe (Fig. 5B and D), indicating that SPPα is the prevailing isoform in these tissues. No differences in mRNA levels were found in testis (Fig 5E and F) which, in agreement with the results of Fig. 4, is probably due to a relatively equal presentation of both SPP-splice variants in testis. The same probes were used to study the expression of SPP mRNAs in sagittal sections of adult mouse brain (Fig. 6). Most areas of the brain showed comparable hybridisation with both probes, which suggests either equal representation of both or only of SPPβ. In accordance with earlier findings elevated levels were observed in the hippocampus (Fig. 6A and B, E and F), the piriform cortex (Fig. 6A and B), the arcuate nucleus (Fig. 6A and
Fig. 6. Differential expression of SPPα and SPPβ mRNA in the adult mouse brain. (A, B) Neighbouring sagittal sections of adult mouse brain at the level of the hippocampus were hybridised in panels A, C and E with the common probe and in panels B, D, and F with the SPPβ-specific probe (same probes as in Fig. 5). An overview of SPPα and SPPβ mRNA expression is shown in panels A and B and higher magnifications of the hindbrain in panels C and D and of the hippocampal area in panles E and F. Arc, arcuate hypothalamic nucleus; bnm, basal nucleus Meynert; chp, choriod plexus; hi, hippocampus; lc, locus coeruleus; pir, piriform cortex; pn, pontine nuclei.
Fig. 5. Differential expression of SPPα and SPPβ mRNA in different organs of the adult mouse. Darkfield photomicrographs of sections from liver (A, B), kidney (C, D) and testis (E, F) were probed with an RNA transcript from exon 12, recognising SPPα and SPPβ (A, C, E) or with a transcript specific for SPPβ (B, D, F).
B), the locus coeruleus (Fig. 6A–D) and the Purkinje cells of the cerebellum (Fig. 6A–D). Absence of any signal for SPPβ in the choroid plexus indicated exclusive expression of SPPα in this location (Fig. 6A–F).
164
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
4. Discussion In search for full-length SPP we discovered that some of the human and mouse ESTs contained an additional exon, inserted between exons 11 and 12 of the known sequence. Using primers from neighbouring exons to amplify cDNAs we found that the new isoform, which we named SPPβ, contained 200 nucleotides more than the known SPP, which we renamed SPPα. Insertion of the new exon resulted in a frame shift in exon 12 which, together with a premature stop codon, totally prevented translation of exon 12 and removed the ER-retention signal. The resulting SPPβ protein thus contained a new, distinct carboxy tail consisting of 49 amino acids compared to 33 completely different amino acids in SPPα. The small change in molecular mass was not detectable in western blots, where SPPβ like SPPα appeared as monomer and dimer, and as an additional band of intermediate size. Cell-surface biotinylation experiments showed that only the monomeric and intermediate forms of SPPβ were present at the cell surface. To assess frequency and presentation of the two isoforms, a multiple-tissue cDNA panel of mouse was probed with primer pairs amplifying either both or only the new isoform. The resulting amplification products of expected sizes revealed predominance of SPPα over SPPβ. A close look at the exon– intron junctions clearly favoured SPPα expression. Since no additional bands were obtained, we assume that no other variants are present in mouse. In the NCBI database, four isoforms were described for human SPP and none for mouse. One human isoform represented a truncated version comprising exons one to three of SPP. Two isoforms corresponded to SPPα and SPPβ, as described by us, and a new isoform was detected as result of a splice site located in front of the stop codon in the SPPβ-specific exon yielding an RNA of intermediate size between SPPα and SPPβ. Since we detected only two bands and no intermediate or shorter versions by RNA analysis, we conclude that the new human variants are absent or very rare in the mouse. In the mouse, we observed a clear predominance of SPPα over SPPβ, both in the embryo and in adult tissues such as spleen, lung, liver and kidney. An almost equal, but weaker representation was found in testis and heart, whereas in brain and skeletal muscle SPPβ prevailed. In situ hybridisation confirmed these results. This may indicate that SPPβ has an important role in the brain. We also investigated the subcellular localisation of the two isoforms. As expected from the missing ER-retention signal in SPPβ, marked differences were found. While SPPα was most strongly expressed in the ER, SPPβ appeared at the cell surface. SPPα contains with KKEK a consensus sequence for ERretention (KKXX–COOH), and the carboxy tail extends into the cytoplasm [14]. By using an antiserum against the last 25 amino acids of the carboxy tail of SPPα we had shown earlier that after digitonin treatment, which only permeabilises the outer cell membrane, indeed the carboxy tail is accessible for immune staining [13]. As control served disulphide-isomerase, which due to its carboxy-terminal KDEL-COOH ER-retention sequence [15], was located in the lumen of the ER [13]. To confirm the differential intracellular localisation of SPPα and SPPβ,
both were provided with an amino-terminal FLAG tag. This allowed analysis of its localisation at the outer cell membrane. We found that both in permeabilised and non-permeabilised cells most of the SPPβ protein was present at the cell surface. Cellsurface biotinylation, where only SPPβ and not SPPα was labeled, supports this notion. We interpret this to mean that SPPβ preferentially cleaves type II transmembrane proteins at the cell surface, whereas SPPα may have a more restricted intracellular function. The fact that SPP is present in early development of various organisms including mouse [3,13] and nematodes [16] makes possible targets for cleavage a very interesting topic for future research. Acknowledgments We thank Dr. S. Hoffmeister-Ullerich for providing mouseorgan sections, Simon Hempel for help with the figures, and Kathrin Hilke-Steen for preparing the manuscript.
References [1] R. Kopan, M.X. Ilagan, Gamma-secretase: proteasome of the membrane? Nat. Rev. Mol. Cell. Biol. 5 (2004) 499–504. [2] A. Weihofen, B. Martoglio, Intramembrane-cleaving proteases: controlled liberation of proteins and bioactive peptides, Trends Cell Biol. 13 (2) (2003) 71–78. [3] D.J. Casso, S. Tanda, B. Biehs, B. Martoglio, T.B. Kornberg, Drosophila signal peptide peptidase is an essential protease for larval development, Genetics 170 (2005) 139–148. [4] J.H. Chyung, D.M. Raper, D.J. Selkoe, Gamma-secretase exists on the plasma membrane as an intact complex that accepts substrates and effects intramembrane cleavage, J. Biol. Chem. 280 (2005) 4383–4392. [5] A. Weihofen, K. Binns, M.K. Lemberg, K. Ashman, B. Martoglio, Identification of signal peptide peptidase, a presenilin-type aspartic protease, Science 296 (2002) 2215–2218. [6] S.G. Crawshaw, B. Martoglio, S.L. Meacock, S. High, A misassembled transmembrane domain of a polytopic protein associates with signal peptide peptidase, Biochem. J. 384 (2004) 9–17. [7] A. Henricson, L. Kall, E.L. Sonnhammer, A novel transmembrane topology of presenilin based on reconciling experimental and computational evidence, FEBS J. 272 (2005) 2727–2733. [8] E. Friedmann, M.K. Lemberg, A. Weihofen, K.K. Dev, U. Dengler, G. Rovelli, B. Martoglio, Consensus analysis of signal peptide peptidase and homologous human aspartic proteases reveals opposite topology of catalytic domains compared with presenilins, J. Biol. Chem. 279 (2004) 50790–50798. [9] A.C. Nyborg, K. Jansen, T.B. Ladd, A. Fauq, T.E. Golde, A signal peptide peptidase (SPP) reporter activity assay based on the cleavage of type II membrane protein substrates provides further evidence for an inverted orientation of the SPP active site relative to presenilin, J. Biol. Chem. 279 (2004) 43148–43156. [10] A.C. Nyborg, A.Y. Kornilova, K. Jansen, T.B. Ladd, M.S. Wolfe, T.E. Golde, Signal peptide peptidase forms a homodimer that is labeled by an active site-directed gamma-secretase inhibitor, J. Biol. Chem. 279 (2004) 15153–15160. [11] P. Krawitz, C. Haffner, R. Fluhrer, H. Steiner, B. Schmid, C. Haass, Differential localization and identification of a critical aspartate suggest non-redundant proteolytic functions of the presenilin in homologues SPPL2b and SPPL3, J. Biol. Chem. 280 (2005) 39515–39523. [12] A. Weihofen, M.K. Lemberg, E. Friedmann, H. Rueeger, A. Schmitz, P. Paganetti, G. Rovelli, B. Martoglio, Targeting presenilin-type aspartic protease signal peptide peptidase with gamma-secretase inhibitors, J. Biol. Chem. 278 (2003) 16528–16533.
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165 [13] J. Urny, I. Hermans-Borgmeyer, G. Gercken, H.C. Schaller, Expression of the presenilin-like signal peptide peptidase (SPP) in mouse adult brain and during development, Gene Expr. Patterns 3 (2003) 685–691. [14] N. Zerangue, M.J. Malan, S.R. Fried, P.F. Dazin, Y.N. Jan, L.Y. Jan, B. Schwappach, Analysis of endoplasmic reticulum trafficking signals by combinatorial screening in mammalian cells, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 2431–2436.
165
[15] R. Noiva, W.J. Lennarz, Protein disulfide isomerase. A multifunctional protein resident in the lumen of the endoplasmic reticulum, J. Biol. Chem. 267 (1992) 3553–3556. [16] A.P. Grigorenko, Y.K. Moliaka, M.C. Soto, C.C. Mello, E.I. Rogaev, The Caenorhabditis elegans IMPAS gene, imp-2, is essential for development and is functionally distinct from related presenilins, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 14955–14960.
Cell-surface expression of a new splice variant of the mouse signal peptide peptidase Jens Urny, Irm Hermans-Borgmeyer, H. Chica Schaller ⁎ Zentrum für Molekulare Neurobiologie, Universität Hamburg, Martinistr. 52, 20246 Hamburg, Germany Received 2 December 2005; received in revised form 17 February 2006; accepted 28 February 2006 Available online 31 March 2006
Abstract The signal peptide peptidase (SPP) is an intramembrane-cleaving aspartyl protease that acts on type II transmembrane proteins. SPP substrates include signal peptides after they have been cleaved from a preprotein, hence the name. The known SPP isoform, which we renamed SPPα, contains an endoplasmic reticulum retention signal at the carboxy terminus. We found a new splice variant, SPPβ, with an additional in-frame exon inserted between exons 11 and 12 of SPPα. Insertion of the new exon led to a complete change in the amino-acid sequence of the carboxy tail. A stop codon within this new exon resulted in silencing of exon 12 and eliminated the endoplasmic reticulum retention signal. The new SPP isoform predominantly localised to the cell surface in contrast to the more restricted localisation of SPPα in the endoplasmic reticulum. Differential expression in mouse tissues and in subcellular compartments suggests new functions for SPP in addition to cleaving signal peptides. © 2006 Elsevier B.V. All rights reserved. Keywords: Signal peptide peptidase; SPP; Cell-surface localisation; ER-retention signal; Intramembrane cleavage; Aspartyl protease
1. Introduction Cleavage of proteins within transmembrane domains plays an ever increasing role in cellular communication, development and pathophysiology [1–3]. The most outstanding examples are presenilins which, as active components of the γ-secretase system, on one hand, control development by influencing Notch signalling and on the other hand cell death in Alzheimer's disease by cleaving the amyloid precursor protein to release the neurotoxic fragment Aβ1–42 [4]. Another such intramembranecleaving protease is the signal peptide peptidase (SPP), which derived its name by acting on some signal peptides after prior cleavage by the signal peptidase [5]. Functions for SPP comprise immune surveillance, processing of the core domain of Hepatitis C virus and endoplasmic reticulum (ER) quality control to remove misfolded membrane proteins [2,6]. SPP, four SPP-like proteases and presenilins belong to the family of multipass intramembrane-cleaving aspartyl proteases, which share limited sequence homology, but are distinguished by a similar overall topology with the active aspartyl site motifs ⁎ Corresponding author. Fax: +49 40 42803 5101. E-mail address: [email protected] (H.C. Schaller). 0167-4781/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.bbaexp.2006.02.007
YD and GXGD located in adjacent transmembrane helices [7]. The loop between these active-site containing transmembrane domains of SPP and SPP-like proteases extends into the ER lumen or extracellular space and that of presenilins into the cytoplasm [8,9]. This inverted organisation of the active motifs is interpreted to imply reverse orientation of substrates, such that SPP and SPP-like proteases cleave type II and presenilins type I transmembrane proteins. Presenilins undergo proteolysis and require partners for activation, SPP and SPP-like proteases seem to act alone as uncleaved monomers or homodimers [4,10]. For the respective proteolytic functions, the aspartates are absolutely required, as evidenced by mutation and by common inhibitors targeting the active site [11,12]. For SPP, the number of predicted transmembrane domains varies between seven and nine. Unequivocal is the topology of the glycosylated amino terminus as extending into the lumen [2,10] and the carboxy terminus into the cytosol [13] demanding an unequal number of transmembrane helices for the complete structure. SPP was found to be expressed in many human and mouse tissues including the brain [13]. Its function as signal peptide peptidase suggests an exclusive intracellular localisation in the ER [13], probably due to the presence of the carboxyterminal KKEK motif as ER-retention signal [14].
160
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
We describe in this paper the identification and expression of a splice variant of SPP, SPPβ. Instead of the known carboxy terminus of SPPα, SPPβ contains a new extended sequence without the KKEK ER-retention signal. Accordingly, its localisation is not restricted to the ER, but allows trafficking to the cell-surface suggesting functions for SPPβ different from those of SPPα. Differential expression of the two splice variants in brain and various organs of the mouse supports this notion.
from exon 1 and nucleotides 1256–1279 from exon new served as forward and reverse primers, respectively. All nucleotides are derived from mouse SPP (accession number DQ168449). For in situ analysis sense and antisense probes were labeled with [α-35S]UTP. Probes, specific for SPPβ, comprised nucleotides 1172–1362, and a probe common for SPPα and SPPβ covered nucleotides 1370–1559 of mouse SPPβ (accession number DQ168449). Cryosections of 10 μm from adult mouse brain and different mouse organs were processed as described previously (13).
3. Results 2. Materials and methods
3.1. Identification and cloning of SPPβ 2.1. Cloning and expression of SPPβ Mouse SPPβ was found by sequence analysis of an EST clone from mouse liver (accession number AW259871) and human SPPβ by sequencing an EST from human retina (accession number AA019983), both kindly provided by RZBD (Deutsches Ressourcenzentrum für Genforschung GmbH, Berlin, Germany). For heterologous expression full-length SPPα and SPPβ of mouse were inserted with and without FLAG tag at the amino terminus into the eukaryotic expression vector pcDNA3 (Invitrogen, Karlsruhe, Germany).
2.2. Subcellular localisation SPPα and SPPβ were transiently expressed in COS-7 and CHO-K1 cells. To monitor cell-surface expression living cells were incubated 48 h after transfection with first antibodies for 30 min on ice before fixation with 4% paraformaldehyde. For intracellular localisation cells were fixed with 4% paraformaldehyde, permeabilised with 0.05% Triton X-100, and then reacted with first antibodies. As primary antiserum, specific for the carboxy terminus of SPPα, C18 was used which we had produced in rabbits against a recombinant protein containing the last 25 carboxy-terminal amino acids of SPPα [13]. The antiserum obtained after the fifth booster, C18.5, was diluted 1:3000 both for western blotting and immunocytochemistry. The monoclonal antibody against FLAG (M2) was from Sigma-Aldrich (Taufkirchen, Germany) and applied at a dilution of 1:200. As secondary antibodies served Alexa Fluor 488 goat antirabbit (Invitrogen) and Cy3 anti-mouse (Amersham Biosciences, Freiburg, Germany), both used at a dilution of 1:2000.
2.3. Biotinylation and western blotting CHO-K1 cells were transiently transfected by electroporation with SPPα and SPPβ constructs containing FLAG tags at the amino terminus. 48 h after transfection cells were treated for 30 min at room temperature with 1 mM S-NHSbiotin (Perbio Science, Bonn, Germany), which is membrane-impermeable. The cell pellets were homogenised with a pestle in buffer containing 0.28 M sucrose, 25 mM K3PO4, 5 mM EDTA and protease-inhibitor cocktail (Roche Diagnostics, Mannheim, Germany). The nuclear fraction was pelleted by centrifugation at 1000×g for 5 min at 4 °C. The membrane fraction was pelleted from the supernatant by centrifugation at 150000×g for 1.5 h at 4 °C and solubilised by shaking for 2 h at 4 °C in 1% Triton X-100, 0.5% NP40, 150 mM NaCl. 7 mM EDTA, 1 mM EGTA, 10 mM Tris–HCl, pH 7.4, and protease-inhibitor cocktail. After centrifugation at 150000×g for 1 h at 4 °C the supernatant was diluted 5fold to reduce the detergent concentration and incubated with anti-FLAG M2 agarose (Sigma-Aldrich) overnight at 4 °C, centrifuged, washed and applied to SDS-PAGE containing 8 M urea. Biotinylated proteins were detected with an avidin-peroxidase conjugate (Bio-Rad, München, Germany) using West Dura Super Signal as peroxidase substrate (Perbio Science).
2.4. Tissue distribution and in situ analysis A mouse multiple-tissue cDNA panel (Clontech, Heidelberg, Germany) served as template to amplify SPPα- and SPPβ-specific sequences by PCR. To amplify both isoforms simultaneously, we used nucleotides 318–341 from exon 1 and nucleotides 994–1018 from exon 11 as forward and nucleotides 1429–1452 from exon 12 as reverse primers. To amplify SPPβ specifically, nucleotides 318–341
In search for a full-length cDNA of human SPP twenty ESTs were sequenced. One EST from human retina (accession number AA019983) differed by containing an additional exon of 200 nucleotides between exon 11 and exon 12 of the original sequence (Fig. 1A). Likewise, an EST from mouse liver (accession number AW259871) also contained this exon. The mouse EST comprised 1652 nucleotides, included start and stop codon of full-length SPPβ, and harboured additional sequences from the 5′- and 3′-untranslated regions (submitted to GenBank under the accession number DQ168449). A comparison between mouse SPPα (Fig. 1B) and SPPβ (Fig. 1C) shows that all features typical for intramembrane-cleaving aspartyl proteases are preserved (highlighted in grey in Fig. 1B), and that insertion of the new exon behind the last putative transmembrane domain gives rise to a carboxy-terminal tail with a completely different amino-acid sequence (Fig. 1C). The new carboxy tail of SPPβ also differs in size coding for 49 amino acids compared to 33 in SPPα. The outstanding new feature of the splice variant is that it lacks the carboxy-terminal ER-retention signal KKEK of SPPα (Fig. 1B). The derived amino-acid sequences of the new carboxy tail for SPPβ of mouse and human differ by a single amino-acid exchange (Fig. 1D). The human sequence was submitted to GenBank under the accession number DQ168450. Exon–intron boundaries were verified by comparison with the human and mouse genome sequences. 3.2. Subcellular localisation of SPPα and SPPβ The absence of the ER-retention signal KKEK in SPPβ indicated a possible difference in subcellular localisation of the two splice variants. To study this SPPα and SPPβ sequences were integrated into eukaryotic expression vectors provided with and without FLAG tags at the extracellular amino-terminal domains, respectively. For visualisation of protein expression served monoclonal antibodies against the FLAG epitope and a polyclonal antiserum, produced against the last 25 carboxyterminal amino acids of human SPPα [13]. Transient cotransfection of SPPα without and SPPβ with FLAG tag allowed specific recognition of each splice variant (Fig. 2A). A striking difference was found in subcellular localisation of the two proteins, SPPβ (green) being very strongly expressed at the cell surface, whereas SPPα (red) showed a predominant internal staining pattern indicative of ER localisation. This confirms earlier findings demonstrating co-staining of SPPα with ER markers [3,13]. Cell-surface expression of SPPβ on living cells
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
161
Fig. 1. Genomic organisation, nucleotide and deduced amino-acid sequence of SPPα and SPPβ. (A) Exon and intron boundaries were determined by blasting new and old SPP sequences against human and mouse genome databases. The new additional exon of SPPβ (shaded in grey) is located both in mouse and human between exons 11 and 12 of SPPα. (B) Up to the splice at the end of exon 11, the sequence of SPPα and SPPβ is identical, shown in panels B and C for the mouse. The splice site, where the new exon of SPPβ is integrated, is marked by a black arrow head. For the common sequence of SPPα and SPPβ, the start codon is printed in bold, the ER-retention signal is underlined and the carboxy-terminal stop codon of SPPα is indicated by an asterisk. The motifs typical for aspartyl proteases of the SPP and presenilin family are shaded in grey. (C) Start and stop of the new exon of SPPβ are marked by black arrow heads. The nucleotides of the new exon are highlighted by bold letters. (D) Mouse and human amino-acid sequences of SPPβ are compared showing only a single amino-acid exchange.
162
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
present as monomer and dimer and as additional band of intermediate size. Reaction with avidin showed that only SPPβ was present at the cell surface, and that cell-surface expression was restricted to the monomeric and intermediate forms of SPPβ. 3.3. Differential expression of SPPα and SPPβ mRNAs in mouse tissues
Fig. 2. Subcellular localisation of SPPα and SPPβ. (A) Confocal micrographs of COS-7 cells overexpressing FLAG-tagged SPPβ together with untagged SPPα are shown as overlays. 48 h after transfection cells were fixed, permeabilised and the proteins visualised by antibodies against the FLAG epitope expressed by SPPβ (shown in green) and by the SPPα-specific antiserum against the carboxy terminus (shown in red). Cell-surface localisation is clearly detectable for SPPβ, but not for SPPα. (B) Living, non-permeabilised CHO-K1 cells overexpressing N-terminally FLAG-tagged SPPβ were reacted for 30 min on ice with antibodies against the FLAG tag before fixation.
was demonstrated by transiently expressing SPPβ with an amino-terminal FLAG-tag in CHO-K1 cells and immunostaining without permeabilisation. The FLAG immunoreactivity was nicely visible at the outer cell membrane (Fig. 2B). To confirm these results membrane fractions of CHO-K1 cells transiently transfected with FLAG-tagged SPPα and SPPβ constructs were analysed by western blotting after cell-surface labelling with membrane-impermeable biotin. After solubilisation and immunoprecipitation with anti-FLAG agarose samples were separated by SDS-PAGE containing 8 M urea. The immunoblots were reacted with anti-FLAG antibodies (Fig. 3, left panel) or with an avidin-peroxidase conjugate (Fig. 3, middle panel). We found that SPPα appears as monomer and dimer of the expected sizes of 50 and 100 kDa. Likewise, SPPβ was also
Fig. 3. Western blot analysis of SPPα and SPPβ heterologously expressed in CHO-K1 cells. CHO-K1 cells transfected with SPPα and SPPβ constructs containing a FLAG tag at the amino terminus were cell-surface biotinylated with membrane-impermeable biotin 48 h after transfection. The solubilised membrane fraction was immunoprecipitated with anti-FLAG agarose and separated on urea containing SDS-PAGE. The immunoblots were reacted with anti-FLAG antibodies (left panel) or with an avidin–peroxidase conjugate (middle panel). Ponceau staining served as loading control (right panel).
To determine the expression of mouse SPPα and SPPβ mRNAs in different tissues, a mouse multiple-tissue cDNA panel was probed by PCR with gene-specific primers derived from exon 1 and exon 11 as forward and exon 12 as reverse primer, with the idea to amplify simultaneously both SPPα and SPPβ. In addition, primers over exon 1 and exon new were used to amplify SPPβ specifically (Fig. 1). The amplification reaction yielded the expected products differing in molecular mass by 200 nucleotides, which are present in SPPβ and not in SPPα (Fig. 4). In most of the tissues analysed, SPPα was more strongly expressed than SPPβ. Thus spleen, lung, liver, kidney of adult mouse and embryos of various ages showed strong and predominant expression of SPPα, in heart and testis the expression levels were comparable, and only in brain and skeletal muscle SPPβ expression seemed to prevail over that of SPPα (Fig. 4). To investigate the differential expression in more detail, in situ hybridisations were performed on tissue sections of adult mice. An RNA transcript covering 189 nucleotides of the new exon served as SPPβ-specific probe. Since all exons of SPPα were also present in SPPβ, a specific probe for SPPα could not be constructed (Fig. 1). As control of similar size a fragment of 190
Fig. 4. Expression of SPPα and SPPβ mRNAs in different mouse tissues. A mouse multiple-tissue cDNA panel was analysed by PCR with gene-specific primers from exon 1 as forward and from exon 12 and exon new as reverse primers. In addition primers from exon 11 and exon 12 were used as control to amplify a smaller product. Bands corresponding to SPPα and SPPβ are indicated by arrows. Amplification of GAPDH served as loading control.
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
163
nucleotides from exon 12 was used, which should react with the RNA of both isoforms. Due to the shortness of the probes the signal to noise ratio was suboptimal. Expression analysis in peripheral organs showed that the common probe reacted strongly with liver and kidney (Fig. 5A and C), whereas no signal was detectable with the SPPβ-specific probe (Fig. 5B and D), indicating that SPPα is the prevailing isoform in these tissues. No differences in mRNA levels were found in testis (Fig 5E and F) which, in agreement with the results of Fig. 4, is probably due to a relatively equal presentation of both SPP-splice variants in testis. The same probes were used to study the expression of SPP mRNAs in sagittal sections of adult mouse brain (Fig. 6). Most areas of the brain showed comparable hybridisation with both probes, which suggests either equal representation of both or only of SPPβ. In accordance with earlier findings elevated levels were observed in the hippocampus (Fig. 6A and B, E and F), the piriform cortex (Fig. 6A and B), the arcuate nucleus (Fig. 6A and
Fig. 6. Differential expression of SPPα and SPPβ mRNA in the adult mouse brain. (A, B) Neighbouring sagittal sections of adult mouse brain at the level of the hippocampus were hybridised in panels A, C and E with the common probe and in panels B, D, and F with the SPPβ-specific probe (same probes as in Fig. 5). An overview of SPPα and SPPβ mRNA expression is shown in panels A and B and higher magnifications of the hindbrain in panels C and D and of the hippocampal area in panles E and F. Arc, arcuate hypothalamic nucleus; bnm, basal nucleus Meynert; chp, choriod plexus; hi, hippocampus; lc, locus coeruleus; pir, piriform cortex; pn, pontine nuclei.
Fig. 5. Differential expression of SPPα and SPPβ mRNA in different organs of the adult mouse. Darkfield photomicrographs of sections from liver (A, B), kidney (C, D) and testis (E, F) were probed with an RNA transcript from exon 12, recognising SPPα and SPPβ (A, C, E) or with a transcript specific for SPPβ (B, D, F).
B), the locus coeruleus (Fig. 6A–D) and the Purkinje cells of the cerebellum (Fig. 6A–D). Absence of any signal for SPPβ in the choroid plexus indicated exclusive expression of SPPα in this location (Fig. 6A–F).
164
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165
4. Discussion In search for full-length SPP we discovered that some of the human and mouse ESTs contained an additional exon, inserted between exons 11 and 12 of the known sequence. Using primers from neighbouring exons to amplify cDNAs we found that the new isoform, which we named SPPβ, contained 200 nucleotides more than the known SPP, which we renamed SPPα. Insertion of the new exon resulted in a frame shift in exon 12 which, together with a premature stop codon, totally prevented translation of exon 12 and removed the ER-retention signal. The resulting SPPβ protein thus contained a new, distinct carboxy tail consisting of 49 amino acids compared to 33 completely different amino acids in SPPα. The small change in molecular mass was not detectable in western blots, where SPPβ like SPPα appeared as monomer and dimer, and as an additional band of intermediate size. Cell-surface biotinylation experiments showed that only the monomeric and intermediate forms of SPPβ were present at the cell surface. To assess frequency and presentation of the two isoforms, a multiple-tissue cDNA panel of mouse was probed with primer pairs amplifying either both or only the new isoform. The resulting amplification products of expected sizes revealed predominance of SPPα over SPPβ. A close look at the exon– intron junctions clearly favoured SPPα expression. Since no additional bands were obtained, we assume that no other variants are present in mouse. In the NCBI database, four isoforms were described for human SPP and none for mouse. One human isoform represented a truncated version comprising exons one to three of SPP. Two isoforms corresponded to SPPα and SPPβ, as described by us, and a new isoform was detected as result of a splice site located in front of the stop codon in the SPPβ-specific exon yielding an RNA of intermediate size between SPPα and SPPβ. Since we detected only two bands and no intermediate or shorter versions by RNA analysis, we conclude that the new human variants are absent or very rare in the mouse. In the mouse, we observed a clear predominance of SPPα over SPPβ, both in the embryo and in adult tissues such as spleen, lung, liver and kidney. An almost equal, but weaker representation was found in testis and heart, whereas in brain and skeletal muscle SPPβ prevailed. In situ hybridisation confirmed these results. This may indicate that SPPβ has an important role in the brain. We also investigated the subcellular localisation of the two isoforms. As expected from the missing ER-retention signal in SPPβ, marked differences were found. While SPPα was most strongly expressed in the ER, SPPβ appeared at the cell surface. SPPα contains with KKEK a consensus sequence for ERretention (KKXX–COOH), and the carboxy tail extends into the cytoplasm [14]. By using an antiserum against the last 25 amino acids of the carboxy tail of SPPα we had shown earlier that after digitonin treatment, which only permeabilises the outer cell membrane, indeed the carboxy tail is accessible for immune staining [13]. As control served disulphide-isomerase, which due to its carboxy-terminal KDEL-COOH ER-retention sequence [15], was located in the lumen of the ER [13]. To confirm the differential intracellular localisation of SPPα and SPPβ,
both were provided with an amino-terminal FLAG tag. This allowed analysis of its localisation at the outer cell membrane. We found that both in permeabilised and non-permeabilised cells most of the SPPβ protein was present at the cell surface. Cellsurface biotinylation, where only SPPβ and not SPPα was labeled, supports this notion. We interpret this to mean that SPPβ preferentially cleaves type II transmembrane proteins at the cell surface, whereas SPPα may have a more restricted intracellular function. The fact that SPP is present in early development of various organisms including mouse [3,13] and nematodes [16] makes possible targets for cleavage a very interesting topic for future research. Acknowledgments We thank Dr. S. Hoffmeister-Ullerich for providing mouseorgan sections, Simon Hempel for help with the figures, and Kathrin Hilke-Steen for preparing the manuscript.
References [1] R. Kopan, M.X. Ilagan, Gamma-secretase: proteasome of the membrane? Nat. Rev. Mol. Cell. Biol. 5 (2004) 499–504. [2] A. Weihofen, B. Martoglio, Intramembrane-cleaving proteases: controlled liberation of proteins and bioactive peptides, Trends Cell Biol. 13 (2) (2003) 71–78. [3] D.J. Casso, S. Tanda, B. Biehs, B. Martoglio, T.B. Kornberg, Drosophila signal peptide peptidase is an essential protease for larval development, Genetics 170 (2005) 139–148. [4] J.H. Chyung, D.M. Raper, D.J. Selkoe, Gamma-secretase exists on the plasma membrane as an intact complex that accepts substrates and effects intramembrane cleavage, J. Biol. Chem. 280 (2005) 4383–4392. [5] A. Weihofen, K. Binns, M.K. Lemberg, K. Ashman, B. Martoglio, Identification of signal peptide peptidase, a presenilin-type aspartic protease, Science 296 (2002) 2215–2218. [6] S.G. Crawshaw, B. Martoglio, S.L. Meacock, S. High, A misassembled transmembrane domain of a polytopic protein associates with signal peptide peptidase, Biochem. J. 384 (2004) 9–17. [7] A. Henricson, L. Kall, E.L. Sonnhammer, A novel transmembrane topology of presenilin based on reconciling experimental and computational evidence, FEBS J. 272 (2005) 2727–2733. [8] E. Friedmann, M.K. Lemberg, A. Weihofen, K.K. Dev, U. Dengler, G. Rovelli, B. Martoglio, Consensus analysis of signal peptide peptidase and homologous human aspartic proteases reveals opposite topology of catalytic domains compared with presenilins, J. Biol. Chem. 279 (2004) 50790–50798. [9] A.C. Nyborg, K. Jansen, T.B. Ladd, A. Fauq, T.E. Golde, A signal peptide peptidase (SPP) reporter activity assay based on the cleavage of type II membrane protein substrates provides further evidence for an inverted orientation of the SPP active site relative to presenilin, J. Biol. Chem. 279 (2004) 43148–43156. [10] A.C. Nyborg, A.Y. Kornilova, K. Jansen, T.B. Ladd, M.S. Wolfe, T.E. Golde, Signal peptide peptidase forms a homodimer that is labeled by an active site-directed gamma-secretase inhibitor, J. Biol. Chem. 279 (2004) 15153–15160. [11] P. Krawitz, C. Haffner, R. Fluhrer, H. Steiner, B. Schmid, C. Haass, Differential localization and identification of a critical aspartate suggest non-redundant proteolytic functions of the presenilin in homologues SPPL2b and SPPL3, J. Biol. Chem. 280 (2005) 39515–39523. [12] A. Weihofen, M.K. Lemberg, E. Friedmann, H. Rueeger, A. Schmitz, P. Paganetti, G. Rovelli, B. Martoglio, Targeting presenilin-type aspartic protease signal peptide peptidase with gamma-secretase inhibitors, J. Biol. Chem. 278 (2003) 16528–16533.
J. Urny et al. / Biochimica et Biophysica Acta 1759 (2006) 159–165 [13] J. Urny, I. Hermans-Borgmeyer, G. Gercken, H.C. Schaller, Expression of the presenilin-like signal peptide peptidase (SPP) in mouse adult brain and during development, Gene Expr. Patterns 3 (2003) 685–691. [14] N. Zerangue, M.J. Malan, S.R. Fried, P.F. Dazin, Y.N. Jan, L.Y. Jan, B. Schwappach, Analysis of endoplasmic reticulum trafficking signals by combinatorial screening in mammalian cells, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 2431–2436.
165
[15] R. Noiva, W.J. Lennarz, Protein disulfide isomerase. A multifunctional protein resident in the lumen of the endoplasmic reticulum, J. Biol. Chem. 267 (1992) 3553–3556. [16] A.P. Grigorenko, Y.K. Moliaka, M.C. Soto, C.C. Mello, E.I. Rogaev, The Caenorhabditis elegans IMPAS gene, imp-2, is essential for development and is functionally distinct from related presenilins, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 14955–14960.