Targeting γ-secretase triggers the selective enrichment of oligomeric APP-CTFs in brain extracellular vesicles from Alzheimer cell and mouse models
- 52 Downloads
We recently demonstrated an endolysosomal accumulation of the β-secretase-derived APP C-terminal fragment (CTF) C99 in brains of Alzheimer disease (AD) mouse models. Moreover, we showed that the treatment with the γ-secretase inhibitor (D6) led to further increased endolysosomal APP-CTF levels, but also revealed extracellular APP-CTF-associated immunostaining. We here hypothesized that this latter staining could reflect extracellular vesicle (EV)-associated APP-CTFs and aimed to characterize these γ-secretase inhibitor-induced APP-CTFs.
EVs were purified from cell media or mouse brains from vehicle- or D6-treated C99 or APPswedish expressing cells/mice and analyzed for APP-CTFs by immunoblot. Combined pharmacological, immunological and genetic approaches (presenilin invalidation and C99 dimerization mutants (GXXXG)) were used to characterize vesicle-containing APP-CTFs. Subcellular APP-CTF localization was determined by immunocytochemistry.
Purified EVs from both AD cell or mouse models were enriched in APP-CTFs as compared to EVs from control cells/brains. Surprisingly, EVs from D6-treated cells not only displayed increased C99 and C99-derived C83 levels but also higher molecular weight (HMW) APP-CTF-immunoreactivities that were hardly detectable in whole cell extracts. Accordingly, the intracellular levels of HMW APP-CTFs were amplified by the exosomal inhibitor GW4869. By combined pharmacological, immunological and genetic approaches, we established that these HMW APP-CTFs correspond to oligomeric APP-CTFs composed of C99 and/or C83. Immunocytochemical analysis showed that monomers were localized mainly to the trans-Golgi network, whereas oligomers were confined to endosomes and lysosomes, thus providing an anatomical support for the selective recovery of HMW APP-CTFs in EVs. The D6-induced APP-CTF oligomerization and subcellular mislocalization was indeed due to γ-secretase blockade, since it similarly occurred in presenilin-deficient fibroblasts. Further, our data proposed that besides favoring APP-CTF oligomerization by preventing C99 proteolysis, γ-secretase inhibiton also led to a defective SorLA-mediated retrograde transport of HMW APP-CTFs from endosomal compartments to the TGN.
This is the first study to demonstrate the presence of oligomeric APP-CTFs in AD mouse models, the levels of which are selectively enriched in endolysosomal compartments including exosomes and amplified by γ-secretase inhibition. Future studies should evaluate the putative contribution of these exosome-associated APP-CTFs in AD onset, progression and spreading.
KeywordsExtracellular vesicles C99 APP-CTFs Homo- and hetero-oligomerization Endosomes Lysosomes trans-Golgi network SorLA γ-Secretase inhibition Presenilin knockout Alzheimer’s disease
amyloid precursor protein
a-secretase inhibitor GI254023X
exosomal inhibitor GW4869
human embryonay kidney
human epitheloid cervix carcinoma
high molecular weight
mouse embryonary fibroblasts
nanoparticle tracking analysis
phosphate buffered saline
C99 expressing SH-SY5Y neuroblastoma
sortilin receptor with A-type repeats
A growing body of evidence indicate that the β-secretase-derived fragment of βAPP C99 (βCTF) accumulates in brains from various AD mouse models  as well as in brains from AD-affected patients [2, 3, 4, 5] or in induced pluripotent stem cells (iPSCs) derived from monogenic AD [6, 49]. C99 accumulation was found to mainly occur within endolysosomal compartments and be both a cause and consequence of lysosomal dysfunction [7, 8, 9]. Moreover, its accumulation was proposed to be linked to neuronal hyperactivity  and LTP alterations [7, 11]. Unexpectedly, our previous work on the 3xTgAD mouse model demonstrated that endolysosomal-associated C99 was only detected with N-terminal-directed and aggregate-specific antibodies but not with C-terminal-directed APP antibodies. This staining was strongly increased in γ-secretase inhibitor treated-mice, in which not only C99 but also C99-derived C83 (hereafter refered to as APP-CTFs) levels were enhanced and those of Aβ reduced, ruling out the possibility that it could be ascribed to Aβ. However, why this endolysosomal APP-CTF immunoreactivity presents an “aggregate-like” conformation remained to be established. Using the AAV-C99 mouse model, expressing C99 in absence of APP overexpression, we delineated the presence of two distinct APP-CTF immunolabelings. Athough we confirmed an endolysosomal-associated staining revealed by N-terminal directed and aggregate specific antibodies in these mice, an additional Golgi-associated staining was recovered by means of antibodies targeting the C-terminal moiety of APP-CTFs, thus proposing the presence of two distinct immunoreactive APP-CTF species. Strikingly, the N-terminal and aggregate-specific antibodies also revealed a diffuse and extracellular immunostaining in both AAV-C99 and 3xTgAD mice which was also strongly increased in γ-secretase inhibitor-treated animals. Since APP-CTFs are membrane-embedded, and thus, should not be recovered in the extracellular space, we postulated that this staining could correspond to exosome-associated APP-CTFs, which would then be particularly important in γ-secretase-treated animals. Amyloidogenic APP processing is thought to mainly occur in the endosomal system [12, 13], in which exosomes are formed , but previous studies have shown that in AD, the exosomal content of Aβ is actually quite low and accounts for less than 1% of the total Aβ content . Conversely, a growing body of evidence indicates that the membrane-bound APP-CTFs, including the C99 fragment, the direct precursor of Aβ , are abundant in extracellular vesicles [17, 18], but little is known about the mechanisms underlying their exosomal accumulation, role, fate and subsequent functional consequences. In the present work, we confirmed the strong abundance of APP-CTFs in purified extracellular vesicles (EVs) from various cellular and mouse AD models and found that the treatment with the γ-secretase inhibitor D6 led to further increased levels of C99 and C99-derived C83. Surprisingly, D6 treatment also unraveled the occurence of several high molecular weight (HMW) APP-CTFs, the levels of which were hardly detectable in cell or brain homogenates. Previous in vitro data suggest that C99 can exist not only as monomers but also as dimers and that dimerization is favored by three repeats of the glycine-xxx-glycine (GXXXG) motif present in the juxtamembrane and transmembrane regions . Thus, we questioned whether these HMW APP-CTFs could be dimeric/oligomeric APP-CTFs and used pharmacological and immunological approaches, as well as GXXXG dimerization mutants  to determine the exact nature of those APP-CTFs. Taken together, we establish that these HMW APP-CTFs indeed correspond to dimeric/oligomeric APP-CTFs, including not only C99 homodimers but also C83 homodimers and C99/C83 oligomers. Furthermore, our data showed that the selective recovery of dimeric/oligomeric APP-CTFs in EVs is linked to their intracellular mislocalization to the endolysosomal network due partly to a defective SorLA-mediated retrograde sorting of dimeric APP-CTFs to the TGN.
Materials and methods
Animals, viral infection and ELND006 treatment
3xTgAD (harboring βAPPswe, and TauP301L transgenes on a PS1M146V background) and non-transgenic (wild-type) mice  were generated from breeding pairs provided by Dr. LaFerla (Irvine, USA). AAV-10 production and AAV-mediated in vivo delivery was described previously . Briefly, 1 day old C57BL6 mice (Janvier Labs., France) were injected unilaterally with 4 μl of AAV virus, AAV-C99 or AAV-free (under control of the synapsin I promoter) (5.5 × 1012 vg/ml (viral genomes per ml)) into the left lateral ventricle and mice were analyzed at 2 months post-AAV delivery. 3xTgAD and wild-type mice, as well as AAV-infected mice were treated daily for 15 days with the γ-secretase inhibitor ELND006, referred to as D6 hereafter (30 mg/kg, Elan Pharmaceuticals, San Francisco) or with vehicle alone (methylcellulose/polysorbate 80, Sigma) via oral gavage, as described . For the purification of brain EVs (see below), mice were anesthetized by intraperitoneal injection of Ketamine (100 mg/kg) and Xylazine (24 mg/kg) and intracardiacally perfused with PBS before sacrifice. For immunohistochemistry, mice were perfused intracardically with PBS followed by paraformaldehyde 4% before collecting the brains. All animals were housed with a 12:12 h light/dark cycle and were given free access to food and water and experimental procedures were in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC) and local French legislation.
The pcDNA3 SPC99G33L construct was generated using the QuickChange II Site-Directed Mutagenesis Kit (Agilent Technologies) with pcDNA3 SPC99 previously described (flammang 2012) and appropriated primers: 5′- AAG GCG CAA TCA TTC TAC TCA TGG TGG GCG GTG - 3′ and 5′- CAC CGC CCA CCA TGA GTA GAA TGA TTG CGC CTT - 3′. The pcDNA3 SPC99G29L/G33L plasmid was obtained using the same protocol with the pcDNA3 SPC99G33L previously generated and the following primers: 5′- GGG TTC AAA CAA ACT CGC AAT CAT TCT ACT C - 3′ and 5′ - GAG TAG AAT GAT TGC GAG TTT GTT TGA ACC C - 3′). The doxycyclin-inductible pSBtet SPC99 construct used for stable cell line generation was obtained as following. First, the SPC99 fragment was amplified by PCR from the pcDNA3 SPC99 using the following primers (5′– ATA TTA GGC CTC TGA GGC CCC ACC ATG CTG CCC GGT TTG GCA C – 3′ and 5′– GAT GGC CTG ACA GGC CCT AGT TCT GCA TCT GCT CAA AGA ACT TG TAG GTT – 3′) to introduce the SfiI restriction site at both 5′ and 3′ end of fragment. The resulting product was then digested by SfiI and subcloned into the pSBtet vector. All constructs were verified by sequencing. Rab5-GFP, Rab7-GFP and Lamp1-GFP were from Addgene and the SorLAmyc construct was a kind gift from Peter St-George-Hyslop.
Cell culture and treatment
Human neuroblastoma (SH-SY5Y, ATTC or SH-SY5Y-APPswe ), human embryonic kidney cells (HEK293, ATTC), human epitheloid cervix carcinoma (HeLa, ATCC) and mouse embryonic fibroblasts (MEFs, wildtype or devoid of PS1 and PS2, PS1/2−/−)  were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum, penicillin (100 U/ml) and streptomycin (50 μg/ml) purchased from Life Technologies (CA, USA) at 37 °C/5% CO2. Transient transfections of cells were carried out using Lipofectamine 2000 (Life Technologies) for SH-5YSY and MEFs and JetPrime (Polyplus transfection) for HEK293 and HeLa cells, according to the manufacturer’s instructions, and cells were recovered 24–36 h post-transfection. For immunofluorescence analysis, some cells were co-transfected with C99 or C99G29L/G33L and plasmids expressing intracellular organelle-specific proteins (Rab5-GFP, Rab7-GFP and Lamp1-GFP, Addgene). Stable inducible HEK293 and SH-SY5Y cell lines were obtained by co-transfection of the Sleeping Beauty inductible vector (pSBtet SPC99) and the transposase SB100 using JetPrime and Lipofectamine 2000, as described above, and using puromycin selection. For stable cell lines, protein expression was induced by the addition of doxycyclin (10 μg/ml final concentration, Sigma) to the cells 24 h before cell treatments: D6, 1 μM (Elan Pharm.), GI254023X 5 μM (Sigma), or GW4869 10 μM (Sigma), which were added to OptiMEM. Cells for western blot analysis were lysed in RIPA buffer and sonicated. For immunocytochemistry, SH-SY5Y cells were grown on poly-D-lysine coated coverslips and treated as indicated above.
Immunostaining of tissue and cells
Paraformaldehyde-fixed brains (see above) were embedded in paraffin and cut on a microtome in 8 μm thick sections (Thermoscientific, France) or cut directly on a vibratome in 50 μm thick sections (Thermoscientific, France). Brain sections were treated with formic acid 50%/5 min and saturated with 5% BSA/0.1% Triton and then incubated at 4 °C overnight with primary antibodies (NU1 , 1:1000) or 4G8 (Covance, 1:2000), followed by Alexa Fluor-488 conjugated anti-mouse (Molecular Probes, 1:1000) and DAPI (Roche, 1:20000) or HRP-conjugated antibodies (Jackson ImmunoResearch, 1:1000) and DAB substrate (Vector). For co-labeling with Iba1 or GFAP vibratome sections were processed as described and incubated at 4 °C overnight with 4G8 (Covance, 1:2000) and Iba1 (Wako, 1:2000) or anti-GFAP (Abcam, 1:2000) followed by Alexa Fluor-488 and Alexa Fluor-594 conjugated antibodies (Molecular Probes, 1:1000). Cells in culture were paraformaldehyde fixed, permeabilized with 0.1% Triton-X 100 for 10 min, saturated in 5% BSA/0.1% Tween20, probed 1 h with primary antibodies (α-APPct, rabbit polyclonal, gift from Paul Fraser, 1:5000 and in some cases with α-TGN46 (Serotec, AHP500G, 1:1000) and detected with Alexa Fluor-488 or Alexa Fluor-590 conjugated antibodies (Molecular Probes, 1:1000). Confocal images were acquired using a Zeiss SP5 confocal microscope for cells and an Olympus Fluoview10 confocal microscope for tissue sections. Quantification of APP-CTF localization was performed in a double-blind manner. A total number of 400–500 cells were counted for each condition in 3 independent experiments. For each cell, APP-CTF localization was determined as “Golgi/ER” pattern or “endosome/lysosome” pattern.
Purification of EVs
Cells at about 80% of confluence (minimum 3 × 150 mm dishes per condition) were rinsed twice with PBS and cultivated for another 20-22 h in OptiMEM for exosomal secretion. When using inducible cell lines, protein expression was induced by the addition of doxycycline (10 μg/ml) 1 day before the change to OptiMEM, which also contained doxycycline. Media were harvested and spun at 2000 g for 20 min and filtered through a 0.22 μm filter. The supernatant was then sequentially centrifuged at 10000 g for 30 min and 100,000 g for 125 min. The pellet was washed in ice-cold PBS and centrifuged for another 120 min. The final pellet (P100) was resuspended in PBS. For western blot, RIPA buffer 5X× was added and EVs were sonicated in an ultrason bath for 15 min. EVs isolated from the brain extracelullar space were purified according to the protocol described by Perez-Gonzalez et al.  with minor modifications. As described above, mice were perfused intracardically with ice-cold PBS to remove blood extracellular vesicles. Then, hemibrains were cut in 4–5 pieces and enzymatically and mechanically dissociated using “the adult brain dissociation kit” (Miltenyi Biotec) and a GentleMACS Dissociator (Miltenyi Biotec). 10 ml of PBS was added to the cell suspension, which was then filtered through a 70 μm Smartstrainer. The suspension was then spun at 300 g for 10 min and the supernatant was used for exosome purification and the cell pellet for analysis by western blot. The supernatant was sequentially centrifuged at 2000 g/20 min, 10,000 g/30 min and 100,000 g for 90 min. The pellet was then washed in PBS, centrifuged another 90 min and the pellet was resuspended in 2 ml sucrose 0.95 M and introduced into a sucrose gradient (2, 1.65, 1.3, 0.95, 0.6 and 0.25 M, 2 ml each). The sucrose gradient was centrifuged at 200000 g/16 h. 1 ml fractions were collected from the top of the gradient and fractions flanking the interphase separating 2 neighboring sucrose layers were pooled together (a-top 1 ml fraction, b-2 ml, c-2 ml, d = 2 ml, e = 2 ml, f = 2 ml, and g = bottom 1 ml fractions) diluted with 10 ml PBS and centrifuged at 100000 g for 90 min. Accordingly to the original protocol, our preliminary experiments using wild-type mouse brains showed that 3 fractions (b, c and d) contain extracellular vesicles, thus in all further experiments, these fractions were pooled, centrifuged at 100000 g and resuspended in 200 μl of PBS. Then 20 μl was recovered for Nanoparticle tracking analysis (NTA) or electron microscopy (EM), and the remaining extracellular vesicles were lysed by the addition of RIPA5X, proteinase inhibitors and ultrasonicated. 30 μl of the lysate (corresponding to about 10% of the total volume) was used for each lane on the immuno blots.
Electron microscopy and nanoparticle tracking analysis (NTA)
For EM, a drop of EVs was added to a freshly ionized 300 mesh formvar/carbon coated grid and incubated for 5 min to allow adherence of the EVs to the grid. The grid was then washed through 5–7 puddles of ddH2O; and negatively stained in 2% aqueous uranyl acetate for 30 s, then visualized using a JEM 1400 electron microscope operating at 100 kV equipped with a Morada SIS camera. For NTA, vesicles diluted to the appropriate concentrations to permit counting in the range of 1 × 108-1 × 1010 particles/ml were injected into the Nanosight apparatus equipped with a syringe pump (Nanosight NS300, Malvern, France). The EVs were visualized by their scattering with a CCD video camera. Pump speed was set to 50, camera level at 16 and the detection threshold at 3. Particle size and number was determined from 3 (60 s) videos, and averaged using the NTA software.
Western blot analysis
Cell or brain homogenates (10 μg) and EV proteins (30 μl) were separated by Bio-Rad 12% stain-free™ TGX FastCast™ acrylamide gels or 16% tris-tricine gels (for the detection of APP-CTFs). Bio-Rad gels were photoactivated for the visualization of proteins and used for quantification of exosomal proteins (indicated in the Figures as protein stain) before being electrophoretically transferred to nitrocellulose membranes using the Bio-Rad Trans-Blot® Turbo™ Transfer System. Tris-tricine gels were directly transferred to nitrocellulose membranes and boiled in PBS before saturation with milk. Membranes were blotted with the following antibodies: α-APPcter (Rabbit polyclonal, gift from Paul Fraser, 1:1000), WO-2 (Sigma, 1:1000), α-Alix (Santa Cruz, 1A2A, sc-53,540, 1:1000), α-Hsc70 (Santa Cruz, B-6, sc-7298, 1:1000), α-Tsg101 (Santa Cruz, C-2, sc-7964, 1:1000), α-Flotilin-2 (Santa Cruz, H-90, sc-25,507, 1:500), α-Calnexin (Santa Cruz, AF18, sc-23,954, 1:500) or α-Actin (Sigma, 1:5000). After probing with primary antibodies, immunological complexes were revealed with HRP-conjugated antibodies (Jackson ImmunoResearch, 1:10000) followed by electrochemiluminescence (Westernbright™ Sirius™ and Quantum™ chemiluminescent HRP substrate, Advansta, France). Immunoblots for APP-CTFs were exposed for various times to ensure the detection of all APP-CTF immunoreactivities and the non-saturation of highly expressed APP-CTFs. Peak height of signal intensities from protein bands were quantified with ImageJ software.
Stable inducible C99-expressing HEK293 cells were transiently transfected with a myc-tagged SorLA construct using Lipofectamine 2000 and treated or not with D6, as described above. Cells were lysed in RIPA buffer and 500 μg lysates were precleared with 20 μl Protein A Sepharose (GE Healthcare) for 2 h at 4 °C. Then the supernatant was incubated for 1 h with 1 μl α-APPct antibody before adding 30 μl Protein A Sepharose beads. After several washes in RIPA buffer, the beads were resupended in SDS sample buffer (2×) and incubated for 5 min at 95 °C. 20 μl of the sample was loaded on either 8% or 12% Bio-Rad stain-free™ TGX FastCast™ acrylamide gels and subjected to western blot analysis using either an anti-myc antibody (clone 9E10, Sigma, 1:5000) or α-APPct (1:1000).
All quantitative data were subjected to non-parametric tests, the Mann-Whithney U test for single comparison and one-way ANOVA analysis for multiple comparisons followed by Dunnets post-hoc analysis using GraphPad Prism 7. Data are represented as means ± SEM. Statistical significance is represented by asterix * p < 0.05, **p < 0.01 and ***p < 0.001.
Purified EVs from C99- or APP- expressing cells are enriched in APP-CTFs and contain high molecular weight APP-CTFs, the levels of which are increased by γ-secretase inhibition
The treatment of SH-C99 cells with the inhibitor of exosomal production and secretion GW4869 leads to increased levels of intracellular high molecular weight APP-CTFs
The above described data indicated a high content of APP-CTFs in EVs and suggested that many of these EVs could correspond to exosomes. Exosome production is dependent on the conversion of sphingomyelin into ceramide by neutral sphingomyelinase (nSMase) . Thus, to strengthen our hypothesis, we assumed that GW4869 an inhibitor of nSMases should affect cellular HMW APP-CTF levels and immunological profile and mimic those observed in EVs. Indeed, the treatment of SH-C99 cells with GW4869 (GW) led to increased intracellular levels of D6-induced HMW APP-CTFs (Fig. 1i, k), the electrophoretic profile of which resembled those detected in EVs, while monomeric APP-CTF levels remained almost unaffected (Fig. 1i, j). This observation supported our previous statement that HMW APP-CTFs at least partly, could originate from intracellular compartments.
Purified EVs from C99-expressing animals are enriched in APP-CTFs and contain high molecular weight APP-CTFs, the levels of which are increased by γ-secretase inhibition
The dimerization C99 mutant favors the formation of high molecular weight APP-CTFs resembling those observed in D6-treated cells
Pharmacological blockade or genetic invalidation of γ-secretase unravels a distinct subcellular localization of monomeric and oligomeric APP-CTFs
D6-treatment reduces APP-CTF/SorLA physical interaction
Our previous work on the 3xTgAD and AAV-C99 mouse models showed that C99 accumulation was both a cause and consequence of early lysosomal dysfunction . We found that the treatment of these animals with the γ-secretase inhibitor D6 led to further increased endolysosomal-associated APP-CTFs and to exacerbated lysosomal-autophagic dysfunction . Here we demonstrate that γ-secretase inhibition also drastically increases exosomal APP-CTF levels reflected by a diffuse extracellular APP-CTF-associated immunostaining in C99-accumulating brain areas, an anatomical observation that was supported by further biochemical analysis. First, we confirmed the presence of both C99 and C83 in purified EVs from both C99- and APPswe-expressing cell media and mouse brains, in agreement with previous studies using purified EVs from cell media [17, 18, 36, 37] or brain tissue from Tg2575 mice . In our models, most C83 was found to derive from α-secretase-mediated cleavage of C99, since the treatment with the α-secretase inhibitor GI254023X led to strongly decreased C83 levels and concomitantly increased C99 levels. In agreement with previous works showing the presence of the α-secretase ADAM10 in EVs [36, 38], this cleavage was found to take place at least partly within exosomes, in which the C83/C99 ratio was even higher than the same intracellular ratio. More surprisingly, purified EVs from γ-secretase inhibitor-treated mice or cell media not only displayed further increased C99 and C83 levels but also HMW APP-CTF immunoreactivities, the levels of which were very low in whole cell or brain homogenates. These HMW APP-CTFs were found in cells expressing C99, but not in control cells, indicating that they were directly linked to C99 and not to other APP-cleavage products, such as the recently discovered CTFƞ [39, 40] proposed to be present in exosomes . Indeed, several lines of independent data from pharmacological (α-secretase inhibition), immunological (C- and N-terminal directed and aggregate specific- antibodies) or genetic (GXXXG mutants and presenilin invalidation) approaches indicated that these HMW APP-CTFs correspond to a heterogenous mix of oligomeric APP-CTFs composed of C99 and C83 homo- and hetero-oligomers. Indeed, in agreement with a selective enrichment of HMW APP-CTFs in exosomes, the treatment with the exosomal inhibitor GW4869 only slightly affected the intracellular levels of monomers, but triggered a significant retention of HMW APP-CTFs within cells. Furthermore, our immunocytochemical analysis indicated that the selective recovery of oligomers in exosomes was linked to their mislocalization within the endolysosomal network. Indeed, we found that monomers were localized mainly in the TGN and ER, whereas oligomers were present in endosomal and lysosomal compartments. It is well accepted that γ-secretase cleavage mainly occurs within acidic compartments of the endocytic pathway [12, 13]. Thus, γ-secretase inhibition-induced oligomer formation most likely takes place within these compartments explaining their specific accumulation in endosomes, lysosomes as well as exosomes. However, in addition our data suggested that this endolysosomal build-up was amplified due to defects in APP-CTF trafficking in γ-secretase inhibitor conditions. Both APP  and APP-CTFs  are trafficked between the TGN and endosomes by means of sorting molecules including SorLA (sortilin receptor with A-type repeats). This trafficking is mediated via the C-terminal tail of SorLA and its binding to cytosolic adaptor complexes, the retromer (for retrograde transport) and Golgi-localized γ-ear-containing ARF-binding proteins (GGAs) and AP1 and PACS1 (for anterograde transport) . On one hand, the disruption of retromer binding results in a retrograde-sorting defect with accumulation of SorLA and thus APP in endosomes and their depletion from the TGN. On the other hand, the disruption of the GGA interaction increases the levels of SorLA, and thus APP, in the TGN . Here, we showed that D6-treatment led to a strongly reduced SorLA/APP-CTF interaction suggesting that a defective retrograde transport of APP-CTF oligomers could contribute to the endosomal HMW APP-CTF accumulation. Accordingly, a recent paper showed that hippocampal neurons from mice expressing the APP mutant Y682G APPYG/YG, that poorly interacts with SorLA, display strong APP accumulation within late endosomes (LEs) and lysosomes, ensuing functional alterations of the lysosomal system . Here, we also found a strong lysosomal accumulation of APP-CTFs in D6-treated conditions, and previously showed that this APP-CTF accumulation could be a trigger of lysosomal and autophagic impairments . Our data suggest that defects in APP-CTF trafficking is due to the fact that oligomeric APP-CTFs poorly bind SorLA. However, it should be noted that SorLA itself undergoes regulated intramembrane proteolysis (RIP) with an initial ectodomain shedding by α-secretase yielding SorLA-CTFα that undergoes subsequent γ-secretase-mediated hydrolysis generating SorLA-CTFγ (SorLA-ICD) [34, 35]. Indeed, D6 treatment led to the accumulation of SorLA-CTFα within the same intracellular compartments as APP-CTF oligomers suggesting that endosomal accumulation of SorLA-CTFα perse could affect SorLA function and APP-CTF trafficking. Whatever the exact mechanisms, our findings indicate that γ-secretase activity is critical for a correct APP-CTF trafficking, since the phenotype elicited by the pharmacological blockade of γ-secretase was strictly mimicked by the genetic depletion of presenilins that correspond to the catalytic core of the enzymatic complex [44, 45].
Previous in vitro data have indicated that C99 can dimerize , but our work is the first to demonstrate that APP-CTFs can also heterodimerize and form oligomers and to show the presence of oligomeric APP-CTFs in vivo in AD mouse models. We found that the GXXXG mutant C99G29L/G33L mimics the γ-secretase inhibitor and blocks oligomeric APP-CTFs within compartments of the endosomal network. Probably a modified conformation of this mutant makes it more prone for dimerization, but the fact that dimers selectively accumulate in endosomes suggest that dimerization takes place within these compartments and indicates that oligomerization requires a particular biological environment (membrane lipid composition, pH etc). Indeed, previous works have shown that C99 dimerization is strongly affected by changes in both membrane thickness and lipid composition . Thus taken together, our data indicate that APP-CTF oligomerization depends on the structural conformation of C99, the levels of γ-secretase activity and the intracellular localization, and propose that the targeting of APP-CTF oligomers to late endosomes, lysosomes and exosomes is also linked to a defective APP-CTF retrograde transport to the TGN. Oligomer misdistribution may well contribute to the previously described endolysosomal defects in mouse AD models [7, 8], as well as in human AD affected brains  or primary fibroblasts from individuals with Down Syndrome [9, 48]. Although our studies are based on overexpression models, the conclusions concernant endolysosomal accumulation of aggregated C99 are strengthened by recent works based on either isogenic knockin human iPSCs modified by CRISPR/Cas9 or monogenic iPSCs carrying either APP or PS1 mutations, and which both demonstrate C99-mediated and Aβ-independent alterations of the endolysosomal network . Whereas the paper of Kwart observed early endosomal dysfunction, the work from Hung and Livesey demonstrated lysosomal-autophagic pathology . Similar findings were described by the group of Dr. Nixon using fibroblasts from Down’s patients known to display an elevation of C99. These cells were also found to display defects in early endososomes [4, 50] as well as in lysosomal degradation . Again, these alterations were found to be mediated by C99 and to be Aβ independent.
Interestingly, the data obtained in our study reconcile our immunohistochemical studies describing two distinct conformation- and localization-dependent APP-CTFs in our mouse models. Thus, in situ, only monomeric APP-CTFs localizing to the TGN and the ER, are recognized by C-terminal directed antibodies, whereas oligomerized and monomeric but oligomerization-prone APP-CTFs present in endosomes, lysosomes and exosomes, are detected with aggregate-specific antibodies but not with C-terminal antibodies.
We wish to thank Dr. Laferla for providing the 3×TgAD mouse, Elan Pharmaceuticals for ELND006, Dr. Fraser for the APPct antibody, Dr. St-Georges-Hyslop for the SorLAmyc construct and Dr. De Strooper for providing MEF PSwt and PS1/2−/− cells. We also thank Mathilde Cohen-Tannoudji at the institute of myology in Paris for help in producing AAVs and Sophie Pagnotta at the center of microscopy in Nice for help with electron microscopy analysis.
IL conceived and supervised the study. IL and AB performed most parts of experiments and data analysis. ABo generated stable cells lines and performed experiments with C99 mutants, RP performed the viral injection of neonatal mice, IL and FC wrote the paper. All authors read and approved the final manuscript.
This work has been developed and supported through the LABEX (excellence laboratory, program investment for the future) DISTALZ (Development of Innovative Strategies for a Transdisciplinary approach to ALZheimer’s disease) and by Fondation Alzheimer. A. Be and A. Bo were granted from DISTALZ.
Ethics approval and consent participate
All animal experimental procedures were in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC) and local French legislation.
Consent for publication
The authors declare that they have no competing interests.
- 1.Lauritzen I, Pardossi-Piquard R, Bauer C, Brigham E, Abraham JD, Ranaldi S, Fraser P, St-George-Hyslop P, Le Thuc O, Espin V, et al. The beta-secretase-derived C-terminal fragment of betaAPP, C99, but not Abeta, is a key contributor to early intraneuronal lesions in triple-transgenic mouse hippocampus. J Neurosci. 2012;32:16243–16255a.PubMedPubMedCentralCrossRefGoogle Scholar
- 2.Hebert SS, Horre K, Nicolai L, Papadopoulou AS, Mandemakers W, Silahtaroglu AN, Kauppinen S, Delacourte A, De Strooper B. Loss of microRNA cluster miR-29a/b-1 in sporadic Alzheimer's disease correlates with increased BACE1/beta-secretase expression. Proc Natl Acad Sci U S A. 2008;105:6415–20.PubMedPubMedCentralCrossRefGoogle Scholar
- 3.Pera M, Alcolea D, Sanchez-Valle R, Guardia-Laguarta C, Colom-Cadena M, Badiola N, Suarez-Calvet M, Llado A, Barrera-Ocampo AA, Sepulveda-Falla D, et al. Distinct patterns of APP processing in the CNS in autosomal-dominant and sporadic Alzheimer disease. Acta Neuropathol. 2013;125:201–13.CrossRefGoogle Scholar
- 5.Pulina M, Hiopkins M, Haroutunian V, Greengard P, Bustos V: C99, not beta-amyloid, is associated with selective death of vulnerable neurons in Alzheimer’s disease. bioRxiv 2019.Google Scholar
- 7.Lauritzen I, Pardossi-Piquard R, Bourgeois A, Pagnotta S, Biferi MG, Barkats M, Lacor P, Klein W, Bauer C, Checler F. Intraneuronal aggregation of the beta-CTF fragment of APP (C99) induces Abeta-independent lysosomal-autophagic pathology. Acta Neuropathol. 2016;132:257–76.PubMedPubMedCentralCrossRefGoogle Scholar
- 9.Ying J, Sato Y, Im E, Berg M, Bordi M, Darji S, Kumar A, Mohan PS, Bandyopadhyay U, Diaz A, et al. Lysosomal dysfunction in Down syndrome is APP-dependent and mediated by APP-betaCTF (C99). J Neurosci. 2019.Google Scholar
- 17.Vingtdeux V, Hamdane M, Loyens A, Gele P, Drobeck H, Begard S, Galas MC, Delacourte A, Beauvillain JC, Buee L, Sergeant N. Alkalizing drugs induce accumulation of amyloid precursor protein by-products in luminal vesicles of multivesicular bodies. J Biol Chem. 2007;282:18197–205.PubMedCrossRefGoogle Scholar
- 19.Kienlen-Campard P, Tasiaux B, Van Hees J, Li M, Huysseune S, Sato T, Fei JZ, Aimoto S, Courtoy PJ, Smith SO, et al. Amyloidogenic processing but not amyloid precursor protein (APP) intracellular C-terminal domain production requires a precisely oriented APP dimer assembled by transmembrane GXXXG motifs. J Biol Chem. 2008;283:7733–44.PubMedPubMedCentralCrossRefGoogle Scholar
- 21.Brigham E, Quinn K, Kwong G, Willits C, Goldbach E, Motter R, Lee M, Hu K, Wallace, Kholodenko D, et al: pharmacokinetic and pharmacodynamic investigation of ELND006, a novel APP-selective gamma-secretase inhibitor, on amyloid concentrations in the brain, CSF and plasma of multiple nonclinical species following oral administration. International Conference on Alzheimer’s Disease 2010.Google Scholar
- 22.Oules B, Del Prete D, Greco B, Zhang X, Lauritzen I, Sevalle J, Moreno S, Paterlini-Brechot P, Trebak M, Checler F, et al. Ryanodine receptor blockade reduces amyloid-beta load and memory impairments in Tg2576 mouse model of Alzheimer disease. J Neurosci. 2012;32:11820–34.PubMedPubMedCentralCrossRefGoogle Scholar
- 27.Thery C, Witwer KW, Aikawa E, Alcaraz MJ, Anderson JD, Andriantsitohaina R, Antoniou A, Arab T, Archer F, Atkin-Smith GK, et al. Minimal information for studies of extracellular vesicles 2018 (MISEV2018): a position statement of the International Society for Extracellular Vesicles and update of the MISEV2014 guidelines. J Extracell Vesicles. 2018;7:1535750.PubMedPubMedCentralCrossRefGoogle Scholar
- 28.Jager S, Leuchtenberger S, Martin A, Czirr E, Wesselowski J, Dieckmann M, Waldron E, Korth C, Koo EH, Heneka M, et al. alpha-secretase mediated conversion of the amyloid precursor protein derived membrane stub C99 to C83 limits Abeta generation. J Neurochem. 2009;111:1369–82.PubMedCrossRefGoogle Scholar
- 29.Flammang B, Pardossi-Piquard R, Sevalle J, Debayle D, Dabert-Gay AS, Thevenet A, Lauritzen I, Checler F. Evidence that the amyloid-beta protein precursor intracellular domain, AICD, derives from beta-Secretase-generated C-terminal fragment. J Alzheimers Dis. 2012;30:145–53.PubMedCrossRefGoogle Scholar
- 31.Spoelgen R, von Arnim CA, Thomas AV, Peltan ID, Koker M, Deng A, Irizarry MC, Andersen OM, Willnow TE, Hyman BT. Interaction of the cytosolic domains of sorLA/LR11 with the amyloid precursor protein (APP) and beta-secretase beta-site APP-cleaving enzyme. J Neurosci. 2006;26:418–28.PubMedPubMedCentralCrossRefGoogle Scholar
- 38.Kowal J, Arras G, Colombo M, Jouve M, Morath JP, Primdal-Bengtson B, Dingli F, Loew D, Tkach M, Thery C. Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes. Proc Natl Acad Sci U S A. 2016;113:E968–77.PubMedPubMedCentralCrossRefGoogle Scholar
- 39.Baranger K, Marchalant Y, Bonnet AE, Crouzin N, Carrete A, Paumier JM, Py NA, Bernard A, Bauer C, Charrat E, et al. MT5-MMP is a new pro-amyloidogenic proteinase that promotes amyloid pathology and cognitive decline in a transgenic mouse model of Alzheimer's disease. Cell Mol Life Sci. 2015;73:217–36.PubMedPubMedCentralCrossRefGoogle Scholar
- 48.Cataldo AM, Petanceska S, Peterhoff CM, Terio NB, Epstein CJ, Villar A, Carlson EJ, Staufenbiel M, Nixon RA. App gene dosage modulates endosomal abnormalities of Alzheimer’s disease in a segmental trisomy 16 mouse model of Down syndrome. J Neurosci. 2003;23:6788–92.PubMedPubMedCentralCrossRefGoogle Scholar
- 50.Jiang Y, Mullaney KA, Peterhoff CM, Che S, Schmidt SD, Boyer-Boiteau A, Ginsberg SD, Cataldo AM, Mathews PM, Nixon RA. Alzheimer’s-related endosome dysfunction in Down syndrome is Abeta-independent but requires APP and is reversed by BACE-1 inhibition. Proc Natl Acad Sci U S A. 2010;107:1630–5.PubMedCrossRefGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.