www.elsevier.com/locate/ymcne Mol. Cell. Neurosci. 29 (2005) 65 – 73
Presenilin endoproteolysis is an intramolecular cleavage Anne L. Brunkan, Maribel Martinez, Emily S. Walker, and Alison M. GoateT Departments of Psychiatry, Neurology, and Genetics, Washington University School of Medicine, St. Louis, MO 63110, USA Received 28 October 2004; revised 28 December 2004; accepted 31 December 2004 Available online 23 February 2005 Mutations in the presenilin genes (PS) account for most cases of familial Alzheimer’s disease. PS contain the active site of the ;secretase complex that cleaves within the transmembrane domain of Bamyloid precursor protein (APP). Full-length PS undergoes regulated endoproteolysis to produce fragments that comprise the active form of PS. The bpresenilinaseQ responsible for endoproteolysis is unknown but may be the same presenilin-dependent ;-secretase activity that cleaves APP. To investigate the mechanism of endoproteolysis, we examined sequence specificity at the cleavage site and tested whether PS dimers are important for endoproteolysis as well as ;-secretase activity. No single point mutation, or a double mutation M292D/V293K, was able to completely abolish endoproteolysis and all mutants supported ;secretase activity. When wtPS1 was co-expressed with either M292D/ V293K or D257A, it was unable to restore normal endoproteolysis to either mutant. Lack of transcleavage by wtPS1 suggests that PS1 endoproteolysis occurs via intramolecular cleavage and does not require dimerization. D 2005 Elsevier Inc. All rights reserved.
Introduction Alzheimer’s disease is the most common cause of progressive neurodegeneration and is characterized pathologically by the accumulation of senile plaques and neurofibrillary tangles in the brain. The major component of senile plaques, amyloid-h peptide (Ah), is produced by proteolytic processing of the h-amyloid precursor protein (APP). APP is first cleaved by either a-secretase (e.g., TACE, ADAM-10) or h-secretase (BACE) to release a large ectodomain and generate membrane-embedded C-terminal fragments of 83 (C83) or 99 (C99) residues (Esch et al., 1990; Seubert
Abbreviations: Ah, amyloid-h peptide; APP, h-amyloid precursor protein; CTF, C-terminal fragment; ELISA, enzyme-linked immunosorbent assay; FL, full-length; HEK, human embryonic kdney 293; IP, immunoprecipitation; MEF, mouse embryonic fibroblasts; NDE, Notch DE; NCT, Nicastrin; NICD, Notch intracellular domain; NTF, N-terminal fragment; PS, presenilin; TM, transmembrane; wt, wild-type. T Corresponding author. Fax: +1 314 747 2983. E-mail address:
[email protected] (A.M. Goate). Available online on ScienceDirect (www.sciencedirect.com). 1044-7431/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.mcn.2004.12.012
et al., 1993). C83 and C99 are subsequently cleaved near the intracellular side of the membrane (Murphy et al., 1999) by gsecretase to release an intracellular C-terminal fragment (CTFg) that enters the nucleus and interacts with transcription factors (Cao and Sudhof, 2001; Cupers et al., 2001; Gao and Pimplikar, 2001; Kimberly et al., 2001). g-secretase also cleaves midway through the membrane domain to generate the p3 (from C83) and Ah (from C99) fragments (Busciglio et al., 1993; Haass and Selkoe, 1993; Haass et al., 1992; Shoji et al., 1992). Several lines of evidence suggest that these CTFg and p3/Ah fragments are not produced by a single cleavage event (Hecimovic et al., 2004). Ah species produced in this manner range in length from 37 to 43 residues (Wang et al., 1996). The major Ah species is comprised of 40 residues (Ah40); however, the longer 42–43 residue species (Ah42, Ah43) are associated with Ah deposition (Jarrett et al., 1993). Notch, a protein that plays an important role in cell fate determination, undergoes processing similar to that of APP. Notch is first cleaved by furin at the S1 site during transport to the plasma membrane (Blaumueller et al., 1997; Logeat et al., 1998). When activated by ligand at the cell surface, Notch then undergoes cleavage by a metalloprotease (such as TACE (Brou et al., 2000)) at the S2 site to release the large ectodomain (Mumm et al., 2000). The remaining membrane-embedded fragment is a substrate for gsecretase cleavage at the S3 site near the intracellular edge of the membrane. Proteolysis at the S3 site releases the Notch intracellular domain (NICD) that travels to the nucleus to modulate transcription (Schroeter et al., 1998). Mutations in APP and in the Presenilin 1 and 2 genes (PS1 and PS2) account for approximately 50% of inherited early-onset familial Alzheimer’s disease and increase the production of the highly fibrillogenic Ah42 species (Scheuner et al., 1996). PS is hypothesized to contain the active site of g-secretase, a multiprotein complex that also contains Nicastrin (NCT), APH-1, and PEN-2 (Edbauer et al., 2003; Francis et al., 2002; Goutte et al., 2002; Kimberly et al., 2003; Steiner et al., 2002; Yu et al., 2000b). g-secretase has characteristics of an aspartyl protease and mutation of either of two aspartic acid residues in PS1 (D257 and D385) inactivates g-secretase (Kimberly et al., 2000; Steiner et al., 1999b; Wolfe et al., 1999a,b). PS1 is a putative 8-transmembrane (TM) protein that undergoes endoproteolytic processing to generate a 6TM N-terminal fragment (NTF) and a 2-TM C-terminal fragment (CTF) (Thinakaran et al., 1996). The NTF and CTF remain
66
A.L. Brunkan et al. / Mol. Cell. Neurosci. 29 (2005) 65–73
associated to form the active PS molecule. The active site aspartate residues are each located in one fragment, D257 in the NTF (TM6) and D385 in the CTF (TM7). Mutation of either of these aspartic acid residues not only abrogates g-secretase activity but also abolishes PS1 endoproteolysis (Kimberly et al., 2000; Steiner et al., 1999b; Wolfe et al., 1999a; Yu et al., 2000a). It has been proposed that full-length PS1 (FL-PS1) is a zymogen that is activated by autoproteolysis to form the active NTF/CTF heterodimer (Li et al., 2000; Wolfe et al., 1999a; Yu et al., 2000a). The NTF and CTF are present in a 1:1 ratio and protein levels remain constant, suggesting that endoproteolysis is a highly regulated event (Thinakaran et al., 1996). Indeed, overexpression of PS1 leads to an accumulation of FL-PS1 rather than an increase in the NTF/CTF levels (Thinakaran et al., 1997). FL-PS1 is rapidly degraded and is found in low molecular weight inactive g-secretase complexes, while the NTF/CTF remain stable for over 24 h and are present in high molecular weight active g-secretase complexes (Capell et al., 1998; Ratovitski et al., 1997; Thinakaran et al., 1997; Yu et al., 1998; Zhang et al., 1998). The major PS species in vivo is the cleaved NTF/CTF heterodimer (Thinakaran et al., 1996), and it is these fragments that bind to transition state analogue inhibitors of g-secretase (Li et al., 2000), again suggesting that the cleaved NTF/CTF PS complex is the active form of the molecule. Finally, although the bpresenilinaseQ that cleaves the FL-PS1 molecule is unknown, it has the characteristics of an aspartyl protease and some, but not all g-secretase inhibitors block PS1 endoproteolysis (Beher et al., 2001; Campbell et al., 2002, 2003), supporting the hypothesis that PS endoproteolysis is an autocatalytic event (Wolfe et al., 1999a) such that the same g-secretase active site within PS that cleaves APP and Notch may also cleave and activate PS itself. It has been suggested that g-secretase activity might require PS1 dimerization. We and others have shown that two PS1 molecules can interact through co-immunoprecipitation (Schroeter et al., 2003), cross-linking between NTF:CTF and NTF:NTF (Schroeter et al., 2003), and yeast two-hybrid studies showing NTF:NTF and CTF:CTF homodimerization (Cervantes et al., 2001). Furthermore, the PS homologue SPP exists primarily in a homodimer and g-secretase inhibitors bind the dimeric form of SPP (Nyborg et al., 2004), supporting the hypothesis that dimerization may be important for g-secretase activity. If the same g-secretase activity functions as the bpresenilinaseQ that cleaves PS as well as the g-secretase that cleaves APP and Notch, dimerization would be important for presenilin endoproteolysis. This model would predict that autoproteolysis by PS could occur by two possible mechanisms: intermolecular cleavage, in which one PS molecule contains the active site that cleaves a second PS molecule, or intramolecular cleavage in which a single PS molecule contains the active site to cleave itself. This mechanism has not been directly addressed although some observations have been made with PS1 mutants such as the active site mutant D257A that does not undergo presenilinase cleavage. When expressed at high levels in cells that contain endogenous PS, the D257A transgene breplacesQ the endogenous PS, making it impossible to determine whether or not the endogenous wtPS would be able to endoproteolyse D257A. However, when D257A is expressed at low levels in these cells, the endogenous wtPS NTF is indistinguishable from possible D257A fragments. In this study, we further characterize the sequence specificity of PS1 endoproteolysis and demonstrate that endoproteolysis is not an intermolecular cleavage event in which two PS1 molecules cleave and activate each other. Rather our results indicate that dimeriza-
tion is not important for presenilinase activity and that endoproteolysis is an intramolecular event dependent upon the activity present within a single PS1 molecule.
Results Mutations at the site of PS1 endoproteolysis do not eliminate PS1 processing We introduced point mutations at the site of endoproteolysis to further characterize the sequence specificity at this site and to determine whether subtle mutations that block endoproteolysis would also block presenilin activity. The P1V residue at the Notch S3 site, which is cleaved by g-secretase, is a valine (V1744), as is the P1V residue at the APP S3-like site (V721). Mutation of Notch V1744 to lysine, leucine, or glycine diminishes production of NICD from the Notch DE precursor (Huppert et al., 2000; Schroeter et al., 1998), and mutation of APP V721 to lysine or alanine results in decreased steady-state levels of the CTFg produced from the C99 precursor, while V721G and V721L mutations have minimal effects (Hecimovic et al., 2004). Interestingly, the P1V residue at the PS cleavage site is also a valine (V293). Previous work (Steiner et al., 1999a) suggested that an M292D mutation at the P1 site of cleavage created a molecule that does not undergo endoproteolytic processing or affect PS1Vs amyloidogenic capability. We further analyzed the M292D mutation as well as another non-conservative mutation (M292E) at the P1 site. We also introduced conservative and nonconservative mutations at the P1V site similar to those previously tested in Notch and APP: V293G, V293L, V293A, and V293K. We first tested the ability of point mutations to block PS1 processing in primary fibroblasts from PS1/PS2 double knock-out mice (Herreman et al., 2000) (PS1/2KO cells) (Fig. 1A). These cells contain no endogenous PS1, allowing assessment of endoproteolysis without the need to distinguish between the transgene and endogenous fragments. Vector alone (-PS1) or PS1 constructs (PS1, D257A, or PS1 endoproteolysis mutants) were transiently co-transfected into the PS1/2KO cells with either Notch DE (NDE) or APP-C99 (C99) and lysates were examined for endoproteolysis or g-secretase production of Notch NICD (NICD) or APP-CTFg (CTFg) (Figs. 1B, C). The NDE and APP-C99 constructs each contain a 6-myc tag at the C-terminus for detection with commercial antibodies. In the absence of PS1 (lane 1), neither NICD nor CTFg was produced by these cells. As expected, the wtPS1 (PS1) molecule underwent endoproteolysis and was able to support both NICD and CTFg production, while the aspartic acid mutant D257A was not cleaved and was unable to rescue gsecretase activity. The P1 mutations M292D and M292E both substantially inhibit endoproteolysis but long exposures of another western blot reveal low levels of endoproteolysis for both mutants (Fig. 1A lower panel). The conservative P1V mutations V293G, V293L, and V293A all support wild-type levels of the PS1-NTF, while the non-conservative V293K mutation shows an intermediate endoproteolysis phenotype. The double mutant M292D/V293K was cleaved at low levels, similar to the P1 site mutants (Fig. 1A). All endoproteolysis mutants supported similar levels of CTFg production (Fig. 1C) and restored NICD production (Fig. 1B), however, the V293K and M292D/V293K mutations resulted in lower steady state levels of NICD. This reduction was confirmed using an antibody specific to NICD (data not shown). Thus, the
A.L. Brunkan et al. / Mol. Cell. Neurosci. 29 (2005) 65–73
67
Fig. 1. Non-conservative mutations at the site of PS1 endoproteolysis greatly reduce but do not eliminate endoproteolysis and do not interfere with endogenous PS1. (A–C) PS1/2KO MEFs were transiently transfected with vector ( PS1), or PS1 constructs as indicated and either NotchDE or APPC99-6MT. (A) Cells were lysed in co-IP buffer and were immunoprecipitated with polyclonal anti-PS1 N-terminus antibody 7N14. PS1 was detected by western blotting with monoclonal anti-PS1 N-terminus antibody NT1. Lower panel, long exposure of another western blot shows NTF produced by PS1, M292D, M292E, M292D/ V293K, and D257A. (B–C) Cell lysates were western blotted with monoclonal anti-myc antibody 9E10 to detect DE and NICD (B) or C99-6MT and CTFg (C). (D–E) HEK cells were transiently transfected with vector (endogenous) or PS1 constructs and APPsw and cells were lysed in HA buffer. (D) Lysates were immunoprecipitated with 7N14 and PS1 was detected by western blotting with NT1 antibody (upper). Cell lysates were directly western blotted with monoclonal anti-APP antibody 13G8 to detect holo-APP and the APPC99 and C83 fragments (lower). (E) Ah levels in conditioned media were measured by ELISA. Total Ah (upper) is shown as a percent of the wtPS1 values. Ah42 and Ah40 levels were normalized to the wtPS1 values and the resulting figures were used to calculate the Ah42/Ahtotal ratio (lower). Error bars represent SE and asterisks indicate a significant (one-way ANOVA with Bonferroni correction) difference from the wtPS1 value (*P b 0.005). Data are the average of three independent experiments.
M292D/E mutations, the V293K mutation, and the double mutation all cause a reduction in endoproteolysis but appear to support a functional g-secretase. Endoproteolysis mutations do not interfere with endogenous PS1 processing or function Some FAD mutations in PS1 and mutation of active site residues (D257, D385) have an inhibitory effect on endogenous wtPS1 in overexpression assays (Borchelt et al., 1996; Wolfe et al., 1999a). To test whether the endoproteolysis mutants would have a similar effect on endogenous PS, we co-transfected HEK293 cells with wtPS1, D257A, or endoproteolysis mutants and APPsw, a FL-
APP construct that contains the Swedish mutation (KM670/ 671NL), which results in an increase in h-secretase cleavage (Scheuner et al., 1996) (Fig. 1D,E). APP expression was assessed by western blot analysis and Ah production was measured by ELISA on conditioned media collected from the same cells. Cells expressing only endogenous presenilin or transfected with wtPS1 were able to completely process the C99/C83 fragments, as evidenced by their absence in lanes 1 and 2 (Fig. 1D, lower blot). However, cells expressing D257A showed an accumulation of the C99/C83 fragments and a significant reduction in total Ah production (Fig. 1E) despite the presence of endogenous wtPS1 and wtPS2 in these cells. The M292D, V293G, and M292D/ V293K mutants did not cause a significant accumulation of the g-
68
A.L. Brunkan et al. / Mol. Cell. Neurosci. 29 (2005) 65–73
secretase substrates or affect total Ah levels and thus do not interfere with endogenous PS activity. Curiously, the V293K mutation had an FAD-like effect on the Ah42:Ah42+40 ratio (Fig. 1E, lower graph). The V293K mutation thus reduces endoproteolysis and NICD production in PS1/2KO cells and produces an FADlike phenotype with respect to Ah42 levels in HEK cells expressing mutant and wild-type PS. Presenilin endoproteolysis is an intramolecular autocatalytic event We have previously hypothesized that active g-secretase may contain two PS molecules (Schroeter et al., 2003), supporting the proposal that PS endoproteolysis may be an intermolecular cleavage. To distinguish between intermolecular and intramolecular endoproteolytic mechanisms, we co-expressed a tagged wtPS1 molecule (H6X-PS1wt) that has a 6-His and an Express tag at its Nterminus (Steiner et al., 2002) with either the D257A construct or the M292D/V293K construct and assayed the ability of the H6XPS1wt to cleave the mutant PS1. D257A has a mutation in its active site, but the endoproteolytic site sequence is intact. Although it is possible that D257A has an abnormal conformation that makes it uncleavable by a wtPS1 molecule, currently there is no evidence to support a change in conformation but rather data collected on this mutation defines a change in a critical active site residue. Conversely, the M292D/V293K mutation is at the endoproteolytic site while its active site is unaltered. We reasoned that if PS1 exists as a dimer and PS1 autocatalysis occurs through an intermolecular cleavage, the active site in the H6X-PS1wt molecule should be able to cleave the D257A molecule that has a mutation in its own active site. However, since the M292D/V293K mutation is at the cleavage site, intermolecular cleavage by a wtPS1 molecule (H6X-PS1wt) would still be inhibited by this mutation. Alternatively, if endoproteolysis is an intramolecular cleavage, coexpression of wild-type and D257A PS will have no impact on endoproteolysis of D257A because of the active site mutation. Endoproteolysis of the H6X-PS1wt molecule can be distinguished from the mutants because the N-terminal tag causes it to migrate to a higher position in SDS-PAGE. We co-transfected PS1/2KO MEFs with these constructs either individually or in pairs as indicated (Fig. 2). D257A is unable to cleave itself in these cells, producing no NTF fragments, and M292D/V293K has little presenilinase activity and produces very few NTF fragments, while H6X-PS1wt produces normal levels of NTF and CTF fragments. However, when coexpressed with D257A or M292D/V293K, H6X-PS1wt was unable to cleave either mutant molecule. The presence of the uncleaved molecule did not affect the ability of H6X-PS1wt to cleave or activate itself, as evidenced by the production of the H6X-NTF as well as the cleavage of DE to form NICD and of C99 to form CTFg. One possible explanation for the apparent lack of transcleavage activity and the retention of H6X-PS1wt activity in the presence of uncleaved PS1 is that the tagged molecule forms dimers predominately with other H6X-PS1wt molecules, while the endoproteolysis mutants form dimers predominately with other endoproteolysis mutants. The D257A mutation could potentially interfere with a putative PS-PS interaction. To address this possibility, we co-transfected PS1/2KO MEFs with H6X-PS1wt and V5-tagged-D257A and tested whether the wtPS and the D257A molecule are able to co-immunoprecipitate (Fig. 3). Immunoprecipitation of H6X-PS1wt with a-HisG antibody brought down V5-D257A-FL only in cells that were co-transfected with
Fig. 2. PS1 proteolysis does not occur via a trans-cleavage mechanism. PS1/2KO MEFs were transiently co-transfected with C99-6MT, a Nterminal-tagged wtPS1 construct (H6X-PS1), and either M292D/V293K or D257A as indicated. (A) Cells were lysed in co-IP buffer, immunoprecipitated with 7N14, and western blotted with NT1 antibody. (B) Cell lysates were western blotted with 9E10 to detect Notch DE and NICD. (C) Cell lysates were western blotted with 9E10 to detect C99-6MT and CTFg.
both constructs (Fig. 3A, lane 5) but not in cells transfected with only H6X-PS1wt (Fig. 3A, lane 4) or only V5-D257A (Fig. 3A, lane 2). When lysates from cells expressing only H6X-PS1wt or only V5-D257A were mixed, no co-immunoprecipitation was observed (Fig. 3A, lane 3), confirming that the interaction occurs within the cells. Similarly, immunoprecipitation of V5-D257A with a-V5 antibody brought down H6X-PS1wt-FL only in cells that were co-transfected with both constructs (Fig. 3B, lane 5) but not in singly transfected cells (Fig. 3B, lanes 2, 4) or mixed lysates (Fig. 3B, lane 3). Thus, even though the inactive D257A molecule and wtPS1 molecule are able to interact, the wtPS1 molecule is still unable to cleave the D257A molecule. The lack of a dimerization requirement for presenilinase activity is further supported by an experiment in which D257A molecules far outnumbered wtPS1 molecules, such that wtPS1:wtPS1 dimers would not be expected to form; in this conditions, wtPS1 still underwent endoproteolysis (data not shown). The lack of transcleavage by the wtPS1 molecule suggests that endoproteolysis of PS1 is an intramolecular rather than an intermolecular event.
Discussion PS1 undergoes endoproteolysis between residues M292 and V293 to form a stable NTF/CTF complex that is putatively the form of PS1 present in the active g-secretase complex (Thinakaran et al., 1996; Yu et al., 1998). Endoproteolysis occurs towards the end of g-secretase complex formation and maturation and may be the final step required to activate the precursor FL-PS1 molecule. Several studies have demonstrated that NCT and APH1 form complexes initially and that full-length PS subsequently binds to this complex. PS endoproteolysis and g-secretase activity appear to require association of PEN-2 with this complex (Fraering et al.,
A.L. Brunkan et al. / Mol. Cell. Neurosci. 29 (2005) 65–73
69
Fig. 3. H6X-PS1 and V5-D257A are able to interact. PS1/2KO MEFs were transiently transfected with vector alone (-PS1), vector + V5-D257A or H6X-PS1, or V5-D257A + H6X-PS1. Cells were lysed in co-IP buffer and lysates from V5-D257A and H6X-PS1 cells were mixed (lysate mix). (A) Lysates were subjected to immunoprecipitation with monoclonal antibody anti-HisG and co-immunoprecipitated fragments were detected with anti-V5 monoclonal antibody. (B) Lysates were subjected to immunoprecipitation with monoclonal antibody anti-V5 and co-immunoprecipitated fragments were detected with anti-HisG monoclonal antibody. (C, D) Lysates were western blotted with either anti-V5 (C) or anti-HisG (D) to visualize protein expression.
2004; Gu et al., 2003; Hu and Fortini, 2003; LaVoie et al., 2003; Luo et al., 2003; Takasugi et al., 2003). The stoichiometry of these molecules in the high molecular weight complex is currently unknown. The association of multiple PS molecules has been demonstrated by co-immunoprecipitation and crosslinking experiments (Cervantes et al., 2001; Schroeter et al., 2003), leading to the hypothesis that PS1 may exist as a multimer in the active gsecretase complex. The simplest such complex would contain a PS dimer. g-secretase cleavage at the S3 site of Notch and the S3-like site of APP, and PS endoproteolysis by the bpresenilinaseQ share many similarities. First, both g-secretase and presenilinase appear to be aspartyl proteases and the aspartate residues at positions 257 and 385 are required for both cleavage events (Kimberly et al., 2000; Steiner et al., 1999b; Wolfe et al., 1999a,b; Yu et al., 2000a). Second, both activities can be inhibited by g-secretase inhibitors that are transition state analogues (Beher et al., 2001; Campbell et al., 2002, 2003). Finally, the substrates for each activity contain a valine residue at the P1V site of cleavage. These characteristics suggest that the presenilinase responsible for PS endoproteolysis may be g-secretase. However, non-transition state analogue gsecretase inhibitors do not affect presenilinase activity, suggesting that there are biochemical distinctions between the two activities (Campbell et al., 2003; Petit et al., 2001). Subtle changes to PS structure seem to be sufficient to create an FAD phenotype of increased Ah42 production, suggesting that PS conformation is essential to g-secretase activity and specificity. Thus, g-secretase may carry out both the g-secretase and presenilinase activities, but the complexes may not have identical components or PS conformation. This study sought to further explore the structural requirements for presenilinase and the relationship between PS endoproteolysis and g-secretase activity.
We introduced point mutations at the P1V site, similar to mutations that have been studied in APP and Notch (Hecimovic et al., 2004; Huppert et al., 2000; Schroeter et al., 1998). Comparison of PS1 cleavage site mutations (Fig. 1) shows some similarities between presenilinase and g-secretase: V-NK substitution at the P1V residue of Notch, APP, and PS1 all produce molecules with reduced cleavage. However, other mutations that reduce gsecretase cleavage when introduced into Notch, have no effect on PS endoproteolysis. These results suggest subtle differences between g-secretase and presenilinase activity. Steiner et al. (1999a) argued that an M292D mutation at the site of endoproteolysis also blocked PS1 cleavage but retained g-secretase activity. However, this mutant was characterized in HEK cells expressing endogenous PS1 that may account for the g-secretase activity seen in the presence of M292D mutant PS1. Similarly, Campbell et al. (2003) described inhibitors that block endoproteolysis but allow g-secretase activity, but assayed cleavage and gsecretase activity in a CHO system that also retains endogenous PS1. Thus, it was still unclear whether endoproteolysis is required to activate a FL-PS1 zymogen. In this study, we tested the activity of the endoproteolysis point mutants in PS1/2KO cells that have no endogenous PS1, allowing us to sensitively examine PS1 endoproteolysis and activity. We looked at the Steiner et al. (1999a) M292D PS1 as well as a second non-conservative mutation at the P1 site, M292E. Although the presenilinase activity of these molecules is greatly inhibited (Fig. 1A), residual fragments are produced that may be sufficient to support wild-type levels of g-secretase activity (Figs. 1B–E). Still seeking an uncleaved PS1 molecule, we made the double mutant M292D/V293K. This mutation allowed residual fragments similar to those produced from the M292D mutant (Fig. 1A) but retained g-secretase activity similar to that of the V293K mutant. The
70
A.L. Brunkan et al. / Mol. Cell. Neurosci. 29 (2005) 65–73
fragments created from both the M292D and M292D/V293K mutants included an NTF band at the expected molecular weight as well as a novel NTF band that runs slightly more slowly in SDSPAGE. This may be a minor NTF species that is normally produced but in such small quantities that it is usually overwhelmed by the major NTF band, or it could be a fragment produced from an alternate cleavage site. Expression of wtPS1 at levels that are undetectable by western blot in this system or in PS1 / cells are able to produce CTFg from C99 and NICD from DE (data not shown, (Ray et al., 1999b)). Lai et al. (2003) have reported that only 6% of cellular PS is associated with active g-secretase, thus it is not surprising that western blotting would not have the sensitivity to detect the small number of active molecules necessary to sustain g-secretase activity (Beher et al., 2003). We were unable to generate a PS1 mutant that eliminated cleavage by introducing point mutations at the site of endoproteolysis (Fig. 1). As a result, we were unable to determine whether a PS1 molecule with such a mutation would have g-secretase activity. At this point, our results cannot refute the hypothesis that PS endoproteolysis is essential for g-secretase activity, except in the case of the naturally occurring FAD mutation DE9 that removes 30 amino acids, including the site of endoproteolysis. We and others have recently demonstrated that PS and its homologues can form multimers (Cervantes et al., 2001; Nyborg et al., 2004; Schroeter et al., 2003). Indeed, it is the homodimeric form of the presenilin homologue SPP that is bound by g-secretase inhibitors (Nyborg et al., 2004), suggesting that PS dimerization may be important for g-secretase activity. We sought to determine whether dimerization is required for PS1 endoproteolysis by testing the ability of one PS1 molecule to perform the presenilinase cleavage of a second PS1 molecule. In the PS1/2KO cells that have no endogenous wtPS1, a transfected PS1 molecule must have an intact active site in order to perform the presenilinase function. If the g-secretase complexes for presenilinase and g-secretase activity differ, then lack of g-secretase activity by mutants assayed in these cells could be due to a direct block of the g-secretase active site or inhibition of endoproteolysis that results in no production of active PS1 fragments and thus no g-secretase activity. Introduction of mutations into the DE9 molecule bypasses the endoproteolysis requirement and allows the direct assessment of mutations on gsecretase activity. The active site mutations D257A, D385A, and P433L remain inactive when present in the DE9 molecule (Wang et al., 2004; Wolfe et al., 1999a), arguing that these mutations truly affect the g-secretase active site rather than simply abolishing a separate presenilinase activity. Thus, in the PS1/2KO cells, we can test whether a wtPS1 molecule with an intact active site is able to process the active site mutant D257A, which has an intact endoproteolytic cleavage site. Although we were unable to use subtle point mutations to completely abolish PS1 endoproteolysis, the decrease in PS1 processing of the M292D/V293K mutant was sufficient to distinguish the fragments produced from this mutant from the normal levels of fragments produced from wtPS1. In the PS1/2KO cells, we would not expect a wtPS1 molecule to restore fragment levels from the M292D/V293K mutant because this mutant undergoes reduced cleavage due to mutation at the site of endoproteolysis while it retains the active site residues. As expected, co-transfection of wtPS1 with the M292D/V293K mutant did not restore wild-type levels of fragment production from the mutant molecule (Fig. 2). Cotransfection of wtPS1 with the active site mutant D257A also did not restore fragment production from this mutant (Fig. 2). The lack of
transcleavage of D257A by the wtPS1 molecule was not due to the absence of formation of wt-mutant dimers, since V5 tagged D257A could be co-immunoprecipitated by H6X-PS1wt and vice versa (Fig. 3). Furthermore, if presenilinase contains a PS dimer at its catalytic center then coexpression of wild type and M292D/V293K or D257A PS might be expected to lead to reduced endoproteolysis of the wild type molecule due to the formation of wild-type/mutant dimers with altered activity. No change in wild type NTF levels was observed despite reduced g-secretase activity (Fig. 2). These data show that the presenilinase activity of PS1 is an intramolecular autocatalytic event. Endoproteolysis does not occur in trans, despite the fact that tagged PS can co-immunoprecipitate (Li et al., 2000; Schroeter et al., 2003). Indeed our data suggest that an individual PS molecule might utilize its own active site to perform the presenilinase cleavage on itself. The structure of PS is unknown but three models have been proposed in which PS spans the membrane 6, 7, or 8 times (Doan et al., 1996; Dewji and Singer, 1997; Li and Greenwald, 1996, 1998; Lehmann et al., 1997). In all three models, the site of endoproteolysis is present in a cytoplasmic but highly hydrophobic domain following the 6th TM. However, it may be associated with or embedded in the membrane and present a cleavage site that is located in a similar position within the membrane as the Notch S3 and APP S3-like cleavage sites. At present, there is no structural data available to determine if this domain is located in the cytoplasm or whether it is actually embedded in the membrane. We note however that the amino acid substitutions that produced the most significant effects on endoproteolysis were those that introduced a charged residue and thus may reduce the likelihood that this domain is present in the membrane. The reduced cleavage may then result not from sequence specificity of the enzyme but altered structure of the zymogen. The data presented here suggest a model (Fig. 4) for PS’s joint presenilinase and g-secretase activities. The membrane-embedded domain that contains the site of endoproteolysis occludes the active site in the conformation adapted by a FL-PS1 molecule (I). Upon association of the g-secretase complex and appropriate cofactors, the NTF and CTF associate and PS1 takes on the presenilinase conformation (II), allowing the PS1 active site to cleave PS1 between residues M292 and V293. This cleavage disrupts the hydrophobic domain around the site of cleavage, resulting in withdrawal of this domain from the membrane (III). Active site residues are then accessible for cleavage of g-secretase substrates as PS1 assumes its g-secretase conformation (IV). In the DE9 PS1 molecule, which is active as a FL-PS1, the deletion of residues 290–319 removes the latter half of the hydrophobic region surrounding the site of endoproteolysis. In the model proposed here, the loss of this hydrophobic domain and endoproteolytic site would prevent insertion in the membrane. This would produce a molecule similar in structure to that depicted in Fig. 4 part III, resulting in an active site that is accessible for interaction with gsecretase substrates despite the lack of presenilinase cleavage. This model is supported by data demonstrating that a peptide comprised of the exon 9 sequence is able to block g-secretase activity (Knappenberger et al., 2004). In conclusion, our results demonstrate that bpresenilinaseQ and g-secretase have similar but not identical sequence specificities. Furthermore, while published data suggests that a PS dimer is at the catalytic center of g-secretase, a hypothesis supported by our observation that D257A inhibits wtPS1 activity, our data show that bpresenilinaseQ is an intramolecular cleavage. These observations suggest that presenilinase and g-secretase may be different
A.L. Brunkan et al. / Mol. Cell. Neurosci. 29 (2005) 65–73
71
with NT1 (1:1500) (monoclonal antibody provided by P. Matthews, Nathan Kline Institute, Orangeburg, NY) (Mathews et al., 2000). DE and C99-6MT were detected with a-myc antibody 9E10 (1:1000) (monoclonal antibody, Sigma, St. Louis, MO). APPsw was detected with 13G8 (1:1000) (monoclonal antibody provided by P. Seubert, Elan Pharmaceuticals, Inc., South San Francisco, CA). Cells lines and transient transfection
Fig. 4. Model for role of the endoproteolytic site in regulation of presenilin’s presenilinase and g-secretase activities. (I) The hydrophobic domain containing the endoproteolytic site sits in the membrane and occludes the active site in a newly synthesized FL-PS1 molecule. (II) Upon interaction with other members of the complex, PS adopts the presenilinase conformation that positions the endoproteolytic site within the active site, allowing cleavage between residues M292 and V293. (III) Endoproteolysis disrupts the hydrophobic domain and it is released from the membrane. (IV) The PS active site is now accessible for g-secretase cleavage of other substrates. For simplicity, only one PS1 molecule is depicted in (IV) although dimerization may be required for g-secretase activity.
molecular configurations of the same protease activity, a hypothesis supported by differences in the inhibitor profiles for the two activities. Elucidation of the stoichiometry of the g-secretase complex members and their interactions within the complex will further our understanding of the differences between presenilinase and g-secretase.
Experimental methods Plasmids APP695 with the Swedish mutation (APPsw) and PS1 cDNAs were expressed from the pcDNA3 vector (Invitrogen Corporation, Carlsbad, CA). APPsw, wtPS1, D257A, M292D, and H6X-PS1 have been previously described (Ray et al., 1999a; Steiner et al., 1999a, 2002). The M292E, V293G, V293L, V293A, V293K, and M292D/ V293K mutations were made in wtPS1 using a site-directed mutagenesis protocol based on the QuikChange kit from Stratagene (La Jolla, CA). Site-directed mutagenesis was also used to add a V5 tag at the N-terminus of wtPS1 (V5-PS1), and then to make V5D257A in V5-PS1. Notch DE (M1726V) in pCS2+ (NDE) has a 6-myc tag at its C-terminus (Kopan et al., 1996). C99-6MT is the Cterminal 100 residues of APP and like NDE it is expressed from pCS2 and has a 6-myc tag at its C-terminus (Hecimovic et al., 2004). NDE and C99-6MT both serve as substrates for ligand-independent, g-secretase-dependent proteolysis. Antibodies PS1 was immunoprecipitated with 7N14 (1:80) (polyclonal rabbit antibody against residues 1–14 of human PS1) and detected
Human embryonic kidney (HEK) 293 cells (ATCC, Rockville, MC) and PS1/PS2 knockout (PS1/2KO) mouse embryonic fibroblasts (MEFs) (provided by B. DeStrooper) (Herreman et al., 2000) were grown at 378C under 5% CO2 in DMEM containing 10% FBS (Gibcok, Invitrogen) and 100 Ag/Al penicillin and streptomycin. HEKs and MEFs were plated in 35 mm or 100 mm dishes, respectively, and transiently transfected the following day using FuGenek6 transfection reagent (Roche Diagnostics Corporation, Indianapolis, IN). Cells were harvested 48–76 h post-transfection. Immunoprecipitations, co-immunoprecipitations, and immunoblots Cells were lysed in either HA buffer (0.1% SDS, 1.0% TX-100, 0.5% BSA in PBS) for HEK cells or co-IP lysis buffer (Ray et al., 1999a) for MEFs, and immunoprecipitations were performed essentially as described (Ray et al., 1999a). Proteins were eluted into SDS sample buffer (2% SDS, 62.5 mM Tris pH 6.8, 0.5% dithiothreitol, 10% glycerol) and separated by SDS/PAGE: PS proteins were separated on 10–20% or 4–20% Ready Gels, APP proteins on 10–20% or 12% Ready Gels, and Notch proteins on 7.5% Criterion Gels (Bio-Rad Laboratories, Hercules, CA). Proteins were transferred to PVDF membranes (Immobilonk-P, Millipore Corporation, Billerica, MA), incubated in primary antibody overnight at 48C, and visualized by chemiluminescence (Pierce Biotechnology, Rockford, IL). For co-immunoprecipitations, lysates from singly transfected cells were mixed for the bmixed lysatesQ control (see text), lysates and the mixed lysates sample were each split into two portions, and each portion was immunoprecipitated with either a-V5 or a-HisG antibody. Proteins were eluted into SDS sample buffer lacking dithiothreitol and analyzed by SDS/PAGE and western blotting similar to lysate and immunoprecipitation samples. Ab ELISA HEK cells were transfected with APPsw and PS1 constructs. 48 h post-transfection, the medium was replaced and conditioned for 20 h. The conditioned medium was centrifuged at 48C for 10 min at 14,000 g and assayed for Ah40 and Ah42 levels using colorimetric ELISA kits (BioSource International Inc., Camarillo, CA) according to the manufacturer’s instructions.
Acknowledgments We thank Dr. R. Kopan for the Notch DE construct, Dr. C. Haass for the M292D and H6X-PS1 constructs, Dr. P. Matthews for the NT1 antibody, Dr. P. Seubert for the 13G8 antibody, and Dr. J. Cirrito for technical advice. This work was supported by NIH grant AG017050.
72
A.L. Brunkan et al. / Mol. Cell. Neurosci. 29 (2005) 65–73
References Beher, D., Wrigley, J.D., Nadin, A., Evin, G., Masters, C.L., Harrison, T., Castro, J.L., Shearman, M.S., 2001. Pharmacological knock-down of the presenilin 1 heterodimer by a novel gamma–secretase inhibitor: implications for presenilin biology. J. Biol. Chem. 276, 45394 – 45402. Beher, D., Fricker, M., Nadin, A., et al., 2003. In vitro characterization of the presenilin-dependent gamma–secretase complex using a novel affinity ligand. Biochemistry 42, 8133 – 8142. Blaumueller, C.M., Qi, H., Zagouras, P., Artavanis-Tsakonas, S., 1997. Intracellular cleavage of Notch leads to a heterodimeric receptor on the plasma membrane. Cell 90, 281 – 291. Borchelt, D.R., Thinakaran, G., Eckman, C.B., et al., 1996. Familial Alzheimer’s disease-linked presenilin 1 variants elevate Abeta1-42/1– 40 ratio in vitro and in vivo. Neuron 17, 1005 – 1013. Brou, C., Logeat, F., Gupta, N., et al., 2000. A novel proteolytic cleavage involved in Notch signaling: the role of the disintegrin-metalloprotease TACE. Mol. Cell 5, 207 – 216. Busciglio, J., Gabuzda, D.H., Matsudaira, P., Yankner, B.A., 1993. Generation of beta-amyloid in the secretory pathway in neuronal and nonneuronal cells. Proc. Natl. Acad. Sci. U. S. A. 90, 2092 – 2096. Campbell, W.A., Iskandar, M.K., Reed, M.L., Xia, W., 2002. Endoproteolysis of presenilin in vitro: inhibition by gamma–secretase inhibitors. Biochemistry 41, 3372 – 3379. Campbell, W.A., Reed, M.L., Strahle, J., Wolfe, M.S., Xia, W., 2003. Presenilin endoproteolysis mediated by an aspartyl protease activity pharmacologically distinct from gamma–secretase. J. Neurochem. 85, 1563 – 1574. Cao, X., Sudhof, T.C., 2001. A transcriptionally [correction of transcriptively] active complex of APP with Fe65 and histone acetyltransferase Tip60. Science 293, 115 – 120. Capell, A., Grunberg, J., Pesold, B., Diehlmann, A., Citron, M., Nixon, R., Beyreuther, K., Selkoe, D.J., Haass, C., 1998. The proteolytic fragments of the Alzheimer’s disease-associated presenilin-1 form heterodimers and occur as a 100–150-kDa molecular mass complex. J. Biol. Chem. 273, 3205 – 3211. Cervantes, S., Gonzalez-Duarte, R., Marfany, G., 2001. Homodimerization of presenilin N-terminal fragments is affected by mutations linked to Alzheimer’s disease. FEBS Lett. 505, 81 – 86. Cupers, P., Orlans, I., Craessaerts, K., Annaert, W., De Strooper, B., 2001. The amyloid precursor protein (APP)-cytoplasmic fragment generated by gamma–secretase is rapidly degraded but distributes partially in a nuclear fraction of neurones in culture. J. Neurochem. 78, 1168 – 1178. Dewji, N.N., Singer, S.J., 1997. The seven-transmembrane spanning topography of the Alzheimer disease-related presenilin proteins in the plasma membranes of cultured cells. Proc. Natl. Acad. Sci. U. S. A. 94, 14025 – 14030. Doan, A., Thinakaran, G., Borchelt, D.R., et al., 1996. Protein topology of presenilin 1. Neuron 17, 1023 – 1030. Edbauer, D., Winkler, E., Regula, J.T., Pesold, B., Steiner, H., Haass, C., 2003. Reconstitution of gamma–secretase activity. Nat. Cell Biol. 5, 486 – 488. Esch, F.S., Keim, P.S., Beattie, E.C., Blacher, R.W., Culwell, A.R., Oltersdorf, T., McClure, D., Ward, P.J., 1990. Cleavage of amyloid beta peptide during constitutive processing of its precursor. Science 248, 1122 – 1124. Fraering, P.C., LaVoie, M.J., Ye, W., Ostaszewski, B.L., Kimberly, W.T., Selkoe, D.J., Wolfe, M.S., 2004. Detergent-dependent dissociation of active gamma–secretase reveals an interaction between Pen-2 and PS1NTF and offers a model for subunit organization within the complex. Biochemistry 43, 323 – 333. Francis, R., McGrath, G., Zhang, J., et al., 2002. aph-1 and pen-2 are required for Notch pathway signaling, gamma–secretase cleavage of betaAPP, and presenilin protein accumulation. Dev. Cell 3, 85 – 97. Gao, Y., Pimplikar, S.W., 2001. The gamma -secretase-cleaved C-terminal
fragment of amyloid precursor protein mediates signaling to the nucleus. Proc. Natl. Acad. Sci. U. S. A. 98, 14979 – 14984. Goutte, C., Tsunozaki, M., Hale, V.A., Priess, J.R., 2002. APH-1 is a multipass membrane protein essential for the Notch signaling pathway in Caenorhabditis elegans embryos. Proc. Natl. Acad. Sci. U. S. A. 99, 775 – 779. Gu, Y., Chen, F., Sanjo, N., et al., 2003. APH-1 interacts with mature and immature forms of presenilins and nicastrin and may play a role in maturation of presenilin–nicastrin complexes. J. Biol. Chem. 278, 7374 – 7380. Haass, C., Selkoe, D.J., 1993. Cellular processing of beta-amyloid precursor protein and the genesis of amyloid beta-peptide. Cell 75, 1039 – 1042. Haass, C., Schlossmacher, M.G., Hung, A.Y., et al., 1992. Amyloid betapeptide is produced by cultured cells during normal metabolism. Nature 359, 322 – 325. Hecimovic, S., Wang, J., Dolios, G., Martinez, M., Wang, R., Goate, A.M., 2004. Mutations in APP have independent effects on Abeta and CTFgamma generation. Neurobiol. Dis. 17, 205 – 218 (E-published 11 Septemer 2004). Herreman, A., Serneels, L., Annaert, W., Collen, D., Schoonjans, L., De Strooper, B., 2000. Total inactivation of gamma–secretase activity in presenilin-deficient embryonic stem cells. Nat. Cell Biol. 2, 461 – 462. Hu, Y., Fortini, M.E., 2003. Different cofactor activities in gamma– secretase assembly: evidence for a nicastrin-Aph-1 subcomplex. J. Cell Biol. 161, 685 – 690. Huppert, S.S., Le, A., Schroeter, E.H., Mumm, J.S., Saxena, M.T., Milner, L.A., Kopan, R., 2000. Embryonic lethality in mice homozygous for a processing-deficient allele of Notch1. Nature 405, 966 – 970. Jarrett, J.T., Berger, E.P., Lansbury Jr., P.T., 1993. The C-terminus of the beta protein is critical in amyloidogenesis. Ann. N. Y. Acad. Sci. 695, 144 – 148. Kimberly, W.T., Xia, W., Rahmati, T., Wolfe, M.S., Selkoe, D.J., 2000. The transmembrane aspartates in presenilin 1 and 2 are obligatory for gamma–secretase activity and amyloid beta-protein generation. J. Biol. Chem. 275, 3173 – 3178. Kimberly, W.T., Zheng, J.B., Guenette, S.Y., Selkoe, D.J., 2001. The intracellular domain of the beta-amyloid precursor protein is stabilized by Fe65 and translocates to the nucleus in a notch-like manner. J. Biol. Chem. 276, 40288 – 40292. Kimberly, W.T., LaVoie, M.J., Ostaszewski, B.L., Ye, W., Wolfe, M.S., Selkoe, D.J., 2003. Gamma–secretase is a membrane protein complex comprised of presenilin, nicastrin, Aph-1, and Pen-2. Proc. Natl. Acad. Sci. U. S. A. 100, 6382 – 6387. Knappenberger, K.S., Tian, G., Ye, X., Sobotka-Briner, C., Ghanekar, S.V., Greenberg, B.D., Scott, C.W., 2004. Mechanism of gamma– secretase cleavage activation: is gamma–secretase regulated through autoinhibition involving the presenilin-1 exon 9 loop? Biochemistry 43, 6208 – 6218. Kopan, R., Schroeter, E.H., Weintraub, H., Nye, J.S., 1996. Signal transduction by activated mNotch: importance of proteolytic processing and its regulation by the extracellular domain. Proc. Natl. Acad. Sci. U. S. A. 93, 1683 – 1688. Lai, M.T., Chen, E., Crouthamel, M.C., et al., 2003. Presenilin-1 and presenilin-2 exhibit distinct yet overlapping gamma–secretase activities. J. Biol. Chem. 278, 22475 – 22481. La Voie, M.J., Fraering, P.C., Ostaszewski, B.L., Ye, W., Kimberly, W.T., Wolfe, M.S., Selkoe, D.J., 2003. Assembly of the gamma–secretase complex involves early formation of an intermediate subcomplex of Aph-1 and nicastrin. J. Biol. Chem. 278, 37213 – 37222. Lehmann, S., Chiesa, R., Harris, D.A., 1997. Evidence for a sixtransmembrane domain structure of presenilin 1. J. Biol. Chem. 272, 12047 – 12051. Li, X., Greenwald, I., 1996. Membrane topology of the C. elegans SEL-12 presenilin. Neuron 17, 1015 – 1021. Li, X., Greenwald, I., 1998. Additional evidence for an eight-trans-
A.L. Brunkan et al. / Mol. Cell. Neurosci. 29 (2005) 65–73 membrane-domain topology for Caenorhabditis elegans and human presenilins. Proc. Natl. Acad. Sci. U. S. A. 95, 7109 – 7114. Li, Y.M., Xu, M., Lai, M.T., et al., 2000. Photoactivated gamma–secretase inhibitors directed to the active site covalently label presenilin 1. Nature 405, 689 – 694. Logeat, F., Bessia, C., Brou, C., LeBail, O., Jarriault, S., Seidah, N.G., Israel, A., 1998. The Notch1 receptor is cleaved constitutively by a furin-like convertase. Proc. Natl. Acad. Sci. U. S. A. 95, 8108 – 8112. Luo, W.J., Wang, H., Li, H., et al., 2003. PEN-2 and APH-1 coordinately regulate proteolytic processing of presenilin 1. J. Biol. Chem. 278, 7850 – 7854. Mathews, P.M., Cataldo, A.M., Kao, B.H., et al., 2000. Brain expression of presenilins in sporadic and early-onset, familial Alzheimer’s disease. Mol. Med. 6, 878 – 891. Mumm, J.S., Schroeter, E.H., Saxena, M.T., Griesemer, A., Tian, X., Pan, D.J., Ray, W.J., Kopan, R., 2000. A ligand-induced extracellular cleavage regulates gamma–secretase-like proteolytic activation of Notch1. Mol. Cell 5, 197 – 206. Murphy, M.P., Hickman, L.J., Eckman, C.B., Uljon, S.N., Wang, R., Golde, T.E., 1999. gamma–secretase, evidence for multiple proteolytic activities and influence of membrane positioning of substrate on generation of amyloid beta peptides of varying length. J. Biol. Chem. 274, 11914 – 11923. Nyborg, A.C., Kornilova, A.Y., Jansen, K., Ladd, T.B., Wolfe, M.S., Golde, T.E., 2004. Signal peptide forms a homodimer that is labeled by an active site directed g-secretase inhibitor. J. Biol. Chem. 279, 15153 – 15160. Petit, A., Bihel, F., Alves da Costa, C., Pourquie, O., Checler, F., Kraus, J.L., 2001. New protease inhibitors prevent gamma–secretase-mediated production of Abeta40/42 without affecting Notch cleavage. Nat. Cell Biol. 3, 507 – 511. Ratovitski, T., Slunt, H.H., Thinakaran, G., Price, D.L., Sisodia, S.S., Borchelt, D.R., 1997. Endoproteolytic processing and stabilization of wild-type and mutant presenilin. J. Biol. Chem. 272, 24536 – 24541. Ray, W.J., Yao, M., Nowotny, P., Mumm, J., Zhang, W., Wu, J.Y., Kopan, R., Goate, A.M., 1999a. Evidence for a physical interaction between presenilin and Notch. Proc. Natl. Acad. Sci. U. S. A. 96, 3263 – 3268. Ray, W.J., Yao, M., Mumm, J., Schroeter, E.H., Saftig, P., Wolfe, M., Selkoe, D.J., Kopan, R., Goate, A.M., 1999b. Cell surface presenilin-1 participates in the gamma–secretase-like proteolysis of Notch. J. Biol. Chem. 274, 36801 – 36807. Scheuner, D., Eckman, C., Jensen, M., et al., 1996. Secreted amyloid betaprotein similar to that in the senile plaques of Alzheimer’s disease is increased in vivo by the presenilin 1 and 2 and APP mutations linked to familial Alzheimer’s disease. Nat. Med. 2, 864 – 870. Schroeter, E.H., Kisslinger, J.A., Kopan, R., 1998. Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature 393, 382 – 386. Schroeter, E.H., Ilagan, M.X., Brunkan, A.L., et al., 2003. A presenilin dimer at the core of the gamma–secretase enzyme: insights from parallel analysis of Notch 1 and APP proteolysis. Proc. Natl. Acad. Sci. U. S. A. 100, 13075 – 13080. Seubert, P., Oltersdorf, T., Lee, M.G., et al., 1993. Secretion of betaamyloid precursor protein cleaved at the amino terminus of the betaamyloid peptide. Nature 361, 260 – 263.
73
Shoji, M., Golde, T.E., Ghiso, J., et al., 1992. Production of the Alzheimer amyloid h protein by normal proteolytic processing. Science 258, 126 – 129. Steiner, H., Romig, H., Pesold, B., Philipp, U., Baader, M., Citron, M., Loetscher, H., Jacobsen, H., Haass, C., 1999a. Amyloidogenic function of the Alzheimer’s disease-associated presenilin 1 in the absence of endoproteolysis. Biochemistry 38, 14600 – 14605. Steiner, H., Duff, K., Capell, A., et al., 1999b. A loss of function mutation of presenilin-2 interferes with amyloid beta-peptide production and notch signaling. J. Biol. Chem. 274, 28669 – 28673. Steiner, H., Winkler, E., Edbauer, D., Prokop, S., Basset, G., Yamasaki, A., Kostka, M., Haass, C., 2002. PEN-2 is an integral component of the gamma -secretase complex required for coordinated expression of presenilin and nicastrin. J. Biol. Chem. 277, 39062 – 39065. Takasugi, N., Tomita, T., Hayashi, I., Tsuruoka, M., Niimura, M., Takahashi, Y., Thinakaran, G., Iwatsubo, T., 2003. The role of presenilin cofactors in the gamma–secretase complex. Nature 422, 438 – 441. Thinakaran, G., Borchelt, D.R., Lee, M.K., et al., 1996. Endoproteolysis of presenilin 1 and accumulation of processed derivatives in vivo. Neuron 17, 181 – 190. Thinakaran, G., Harris, C.L., Ratovitski, T., Davenport, F., Slunt, H.H., Price, D.L., Borchelt, D.R., Sisodia, S.S., 1997. Evidence that levels of presenilins (PS1 and PS2) are coordinately regulated by competition for limiting cellular factors. J. Biol. Chem. 272, 28415 – 28422. Wang, R., Sweeney, D., Gandy, S.E., Sisodia, S.S., 1996. The profile of soluble amyloid beta protein in cultured cell media. Detection and quantification of amyloid beta protein and variants by immunoprecipitation-mass spectrometry. J. Biol. Chem. 271, 31894 – 31902. Wang, J., Brunkan, A.L., Hecimovic, S., Walker, E., Goate, A., 2004. Conserved bPALQ sequence in presenilins is essential for gamma– secretase activity, but not required for formation or stabilization of gamma–secretase complexes. Neuroibiol. Dis. 15, 654 – 666. Wolfe, M.S., Xia, W., Ostaszewski, B.L., Diehl, T.S., Kimberly, W.T., Selkoe, D.J., 1999a. Two transmembrane aspartates in presenilin-1 required for presenilin endoproteolysis and gamma–secretase activity. Nature 398, 513 – 517. Wolfe, M.S., Xia, W., Moore, C.L., Leatherwood, D.D., Ostaszewski, B., Rahmati, T., Donkor, I.O., Selkoe, D.J., 1999b. Peptidomimetic probes and molecular modeling suggest that Alzheimer’s gamma– secretase is an intramembrane-cleaving aspartyl protease. Biochemistry 38, 4720 – 4727. Yu, G., Chen, F., Levesque, G., et al., 1998. The presenilin 1 protein is a component of a high molecular weight intracellular complex that contains beta-catenin. J. Biol. Chem. 273, 16470 – 16475. Yu, G., Chen, F., Nishimura, M., et al., 2000a. Mutation of conserved aspartates affects maturation of both aspartate mutant and endogenous presenilin 1 and presenilin 2 complexes. J. Biol. Chem. 275, 27348 – 27353. Yu, G., Nishimura, M., Arawaka, S., et al., 2000b. Nicastrin modulates presenilin-mediated notch/glp-1 signal transduction and betaAPP processing. Nature 407, 48 – 54. Zhang, J., Kang, D.E., Xia, W., Okochi, M., Mori, H., Selkoe, D.J., Koo, E.H., 1998. Subcellular distribution and turnover of presenilins in transfected cells. J. Biol. Chem. 273, 12436 – 12442.
