Inhibition of γ-Secretase Leads to an Increase in Presenilin-1
- 880 Downloads
γ-Secretase inhibitors (GSIs) are potential therapeutic agents for Alzheimer’s disease (AD); however, trials have proven disappointing. We addressed the possibility that γ-secretase inhibition can provoke a rebound effect, elevating the levels of the catalytic γ-secretase subunit, presenilin-1 (PS1). Acute treatment of SH-SY5Y cells with the GSI LY-374973 (N-[N-(3,5-difluorophenacetyl)-l-alanyl]-S-phenylglycine t-butyl ester, DAPT) augments PS1, in parallel with increases in other γ-secretase subunits nicastrin, presenilin enhancer 2, and anterior pharynx-defective 1, yet with no increase in messenger RNA expression. Over-expression of the C-terminal fragment (CTF) of APP, C99, also triggered an increase in PS1. Similar increases in PS1 were evident in primary neurons treated repeatedly (4 days) with DAPT or with the GSI BMS-708163 (avagacestat). Likewise, rats examined after 21 days administered with avagacestat (40 mg/kg/day) had more brain PS1. Sustained γ-secretase inhibition did not exert a long-term effect on PS1 activity, evident through the decrease in CTFs of APP and ApoER2. Prolonged avagacestat treatment of rats produced a subtle impairment in anxiety-like behavior. The rebound increase in PS1 in response to GSIs must be taken into consideration for future drug development.
KeywordsAlzheimer’s disease Presenilin-1 γ-Secretase inhibitor Therapy
Alzheimer’s disease (AD) is the most common dementia in the elderly, and it is characterized by extracellular deposits of aggregated β-amyloid (Aβ) peptides and accumulation of intracellular tangles of the abnormally hyperphosphorylated microtubule-associated protein tau (P-tau) . According to the amyloid cascade hypothesis, which is the most prevalent view on AD pathogenesis, the disease pathophysiology is triggered by an excess of neurotoxic Aβ peptides, potentially in combination with other genetics and risk factors . Drug candidates targeting Aβ have dominated AD drug development programs for the past three decades , and accordingly, targets for each individual step in this cascade have been developed, with β/γ-secretase inhibitors representing one particular opportunity for front-line therapy.
The Aβ peptide is generated by successive proteolytic processing of the amyloid precursor protein (APP) by secretases. APP is a type I transmembrane spanning glycoprotein that is first processed by either α- or β-secretase, followed by γ-secretase cleavage. β- and γ-secretase cleavage generate Aβ peptides of variable amino acid length, being the most abundant the Aβ40 peptide  while Aβ42 appears to be the most amyloidogenic . The major neuronal β-secretase is the beta-site APP-cleaving enzyme 1 (BACE1) , while the γ-secretase enzyme complex contains four essential subunits: presenilin-1 (or presenilin-2), nicastrin, anterior pharynx-defective 1 (APH1), and presenilin enhancer 2 (PEN2) . γ-Secretase acts an aspartyl protease, which catalytic core is presenilin-1 (PS1), being its dysfunction associated with the pathological development of AD . Thus, compounds that inhibit γ-secretase, targeting PS1, are potential therapeutic agents for AD.
Preclinical studies clearly established that γ-secretase inhibitors (GSIs) reduce brain Aβ in rodent models and also reverse Aβ-induced cognitive deficits in the AD Tg2576 mice . However, the therapeutic effect of such drugs in humans has fallen below expectation, with no demonstrated efficacy in clinical trials and even impaired cognitive function in long-term treated subjects . Problems of tolerability and dose-limiting effects during clinical trials with GSIs may have compromised target engagement for arriving to the minimum extent of Aβ lowering for significant cognitive benefit in AD patients (discussed in Toyn and Ahlijanian ). On the other hand, a paradoxical increase of plasma Aβ levels has been observed upon chronic treatment with a classical GSI in transgenic animal . Treatment of transgenic mice and humans with other GSIs, including compounds involved in clinical trials, may cause late rebound effects on plasma Aβ levels [13, 14, 15]. These changes may be illustrative of a rebound effect in reaction to inhibition by a GSI-based therapy. To decipher why current GSIs fail to improve the disease state may help to optimize future drug development.
Upregulation of enzyme isoforms [16, 17], and also of the specific enzyme targeted by the drug [18, 19, 20], is not an uncommon phenomenon in reaction to inhibition, although to our knowledge, this possible effect remains unexplored in terms of GSI treatment. Interestingly, we recently reported that an increase in acetylcholinesterase could block γ-secretase activity and that this inhibition initiates a feedback process that leads to a rebound effect, elevating PS1 levels . Here, we tested how GSIs affect PS1 levels in cellular and animal models. As such, we provide evidence that γ-secretase inhibition could provoke a rebound increase in PS1, which may be of particular importance for the design of specific AD therapies based on GSIs and related drugs.
Materials and Methods
Cell Cultures and Pharmacological Treatment with GSIs
SH-SY5Y cells (700,000 cells/well) were grown in six-well plates for 24 h in Dulbecco’s modified Eagle’s medium (DMEM) + GlutaMAX™ (Gibco® Life Technologies, Paisley, UK) supplemented with 10% fetal bovine serum (FBS; Gibco) and 100 μg/mL penicillin/streptomycin (Gibco). The cells were treated with 5 μM of γ-secretase inhibitor LY-374973: N-[N-(3,5-difluorophenacetyl)-l-alanyl]-S-phenylglycine t-butyl ester (DAPT; Calbiochem®, Merck KGaA) or the dimethyl sulfoxide (DMSO) vehicle alone. Following an 18-h treatment, the cells were washed twice with cold phosphate-buffered saline (PBS) and resuspended in 100 μL of ice-cold extraction buffer supplemented with a cocktail of protease inhibitors: 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 5 mM EDTA, 1% (w/v) Nonidet P-40, and 0.5% (w/v) Triton X-100. Cell lysates were sonicated and centrifuged for 1 h at 70,000×g and 4 °C, and the extracts were frozen at −80 °C for future analysis.
For some experiments, SH-SY5Y cells were transfected with 4 μg of a construct that encodes the C-terminal 99 amino acids of APP (amino acids 597–695), extending from the β-secretase cleavage site to the C-terminus (a generous gift from David H. Small). A pCI empty vector (Promega) served as the negative control. These cells (7 × 105 cells/well) were then seeded on 35-mm tissue culture dishes and transfected using Lipofectamine® 2000 (Thermo Scientific™) according to the manufacturer’s instructions. After 2 days in culture, the cells and culture supernatants were harvested separately, and the cell culture supernatants were cleared by centrifugation at 1000×g for 10 min at 4 °C. The cells were then washed with PBS and solubilized as described above. C-terminal fragment of APP (APP-CTF) levels were assayed in Western blots to determine transfection efficiency.
To culture primary cortical neurons, cortical lobes from E16.5 mice embryos were trypsinized and dissociated in Hank’s balanced salt solution (Life Technologies). Neurons were plated onto 35-mm dishes (1.3 × 106 cells/dish) and maintained in Neurobasal medium (Invitrogen) containing B27 supplement (Gibco BRL), 100 IU/mL penicillin, 100 μg/mL streptomycin, and 2 mM glutamine. After 7 days in culture, the cortical neurons were treated with 2 μM of DAPT or the GSI avagacestat (BMS-708163; from Bristol-Myers Squibb) for four consecutive days and analyzed on day 5, 18 h after the last dose. The cells were washed with PBS and solubilized as described above.
Cell viability was measured using the tetrazolium assay (MTS; CellTiter 96® AQueous Assay, Promega) according to the manufacturer’s instructions. Cells were cultured in 96-well plates and treated with GSIs as previously stated. MTS was added after GSI treatment, cells were incubated for 4 h, and then viability was determined by measuring the absorbance at 490 nm in a microplate reader (Infinite M200, Tecan).
Animals and Tissue Preparation
All animal procedures were approved by the Animal Care and Use Committees at the Universidad Miguel Hernández and by Centro Principe Felipe (2016A/SC/PEA/00127). Wistar male rats that weighed 250–300 g at the beginning of GSI administration were used. The rats were orally administered the avagacestat (40 mg/kg) or vehicle alone (polyethylene glycol) using a single or once-a-day dose for 4 or 21 days (n = 10 for each group), and they were sacrificed ~ 4 h after the final administration of avagacestat. Cerebrospinal fluid (CSF) samples (50–60 μL) were collected by cisternal puncture with a needle inserted in the suboccipital region through the atlanto-occipital membrane, with a single incision into the subarachnoid space . CSF samples were centrifuged at 1000×g for 10 min at 4 °C, and the supernatants were stored at −80 °C. In addition, the rat’s brain was removed and their cerebral cortices were dissected out and stored at −80 °C. Hemi-cortices were thawed slowly at 4 °C and homogenized (10% w/v) in extraction buffer: 50 mM Tris-HCl (pH 7.4)/500 mM NaCl/5 mM EDTA/1% (w/v) Nonidet P-40/0.5% (w/v) Triton X-100, supplemented with a cocktail of protease inhibitors . The homogenates were sonicated and centrifuged, as indicated above, and the supernatants were collected and frozen at −80 °C. Protein concentrations were determined using the bicinchoninic acid method (Pierce, Rockford, IL, USA). The other hemi-cortices were reserved for messenger RNA (mRNA) analysis (see below).
Cell (20 μg) and brain extracts (40 μg) and CSF samples (30 μL) were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under fully reducing conditions. Samples were denatured at 50 °C for 15 min to analyze PS1 or, alternatively, at 98 °C for 5 min for other proteins. The proteins separated were transferred to nitrocellulose membranes (Schleicher and Schuell Bioscience GmbH) and probed with a PS1 antibody raised against amino acids 1–20 (antibody 98/1; see Evin et al. ). Protein extracts from cell cultures were also probed for other γ-secretase subunits using the following antibodies: mouse anti-nicastrin (Millipore), rabbit anti-PEN2 (Sigma), and rabbit anti-APH1 (which recognizes both the APH1A and APH1B homologs; Sigma).
Brain extracts were also assayed for the CTF of APP or ApoER2 using the monoclonal anti-APP C-terminal antibody C1-6.1 (Covance) or a polyclonal antiserum against the C-terminal of ApoER2 (Abcam). Alternatively, the anti-APP monoclonal antibody 6E10 (Covance) was used. A rabbit anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) antibody (Abcam) was used as a loading control. Western blots for different antibodies were performed separately to avoid re-probing the membranes. Antibody binding was detected with the corresponding conjugated secondary antibody (IRDye 680CW goat anti-mouse and IRDye 800RD goat anti-rabbit; LI-COR Biosciences) and visualized on an Odyssey CLx Infrared Imaging System (LI-COR Biosciences). Densitometric quantification of the signal from immunoreactive bands was obtained using LI-COR software (Image Studio Lite).
RNA Isolation and the Analysis of γ-Secretase Subunit Transcripts by qRT-PCR
The transcripts encoding PS1, nicastrin, PEN2, and two forms of APH1 (APH1A and APH1B) were assayed. The total RNA from rat brain hemi-cortices, SH-SY5Y cells, and mouse cortical neurons was isolated with the TRIzol® Reagent using the PureLink™ Micro-to-Midi Total RNA Purification System (Invitrogen™ Life Technologies), following the manufacturer’s instructions. First-strand complementary DNAs (cDNAs) were synthesized by reverse transcription of 1.5 of total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems; Life Technologies) according to the manufacturer’s instructions. Quantitative PCR amplification was performed using a StepOne™ Real-Time PCR System (Applied Biosystems) and TaqMan PCR Master Mix with specific TaqMan Gene Expression Assays: Hs00997789 for PS1, Hs00950933_m1 for nicastrin, Hs00708570_s1 for PEN2, HS00211268_m1 for APH1A, and Hs0029911_m1 for APH1B on SH-SY5Y cell RNA; Mm00501184_m1 for PS1 on mouse cortical neuron RNA; and Rn00569763_m1 for PS1 on rat brain hemi-cortex RNA. GAPDH was amplified as a housekeeping marker (Hs03929097 for SH-SY5Y cells, Mm99999915_g1 for mouse cortical neurons, and Rn014626662_g1 for rat brain hemi-cortices), and the transcript levels were calculated relative to GAPDH using the comparative 2−ΔCt method.
The Y-maze alternation, active avoidance, and beam walking tests were performed to analyze memory and learning functions, as well as motor coordination. The tests were performed 2–4 h after the final administration of avagacestat.
Y-Maze Novel Spatial Recognition Memory
This test is based on the rodents’ natural curiosity to explore novel areas, and the rats were tested as described elsewhere . Briefly, rats were placed into one of the arms of the Y-maze (start arm) and allowed to explore the maze with one of the arms closed for 3 min (training trial). After a 30-min inter-trial interval, the rats were returned to the Y-maze, placing them in the start arm, and then the rats were allowed to freely explore all three arms of the maze for 3 min (test trial). The number of entries into and the time spent in each arm were registered manually by an observer blinded to the rat’s treatment. The discrimination ratio is a measure of the preference for the novel arm over the familiar (old) arm, calculated as the Time spent in Novel / Time spent in the (Novel + Old).
The active avoidance task is designed to test the ability of the rats to avoid an aversive event by first learning to perform a specific behavior in response to a stimulus. The test was performed on a single day and involved 50 trials per animal, as described previously .
Beam Walking Test
The beam walking test assesses deficits in fine motor coordination , although it is also a useful assay to test for anxiety-like behavior  as it also causes some anxiety in the animal. Motor coordination was tested on a 1-m-long wooden stick (20 mm in diameter) situated approximately 1 m above the ground as described elsewhere . The number of slips (foot faults) and the latency to cross (the time spent on the apparatus as an estimate of anxiety) are scored.
All data were analyzed using SigmaStat (version 3.5; Systat Software, Inc.), determining exact p values by applying a Student’s t test (two-tailed) or the Mann-Whitney rank-sum test, when normality was rejected. The results are presented as the means ± SEM.
Inhibition of γ-Secretase by the GSI DAPT Increases the PS1 in SH-SY5Y and Primary Neuronal Cultures
Likewise, repeated DAPT treatment of mouse primary neuronal cultures grown for 2 weeks and then treated daily with DAPT (2 μM) over 4 days also augmented the amount of PS1 protein (64 ± 11%, p < 0.001; Fig. 1b), with unaltered mRNA levels (Fig. 1b). Again, no cytotoxicity was observed during the treatment (p = 0.4, as compared with cell viability in cells treated with vehicle). Hence, the change in PS1 content persisted when γ-secretase inhibition was maintained.
Effects of APP-CTF Over-expression on PS1 in SH-SY5Y Neuroblastoma Cells
The GSI Avagacestat Alters the PS1 in Cultured Cells and Its Content In Vivo
The possibility that levels and activities of secretases are affected in the brain of AD subjects has been studied intensively [5, 8]. However, whether their inhibition by GSIs can induce persistent compensatory changes in the brain has yet to be addressed. It is known that neurotransmitter transporters potentially undergo alterations to gene transcription, mRNA translation/stability, post-translational, protein trafficking, cytoskeletal interactions, and oligomerization in response to chronic drug administration . Indeed, an upregulation of proteins targeted by pharmacological inhibition has also been documented [18, 19, 20]. Here, we demonstrate that GSIs can induce a feedback mechanism that results in accumulation of PS1 in different cell models. A similar elevation of brain PS1 was identified in 21-day avagacestat-treated rats, which also displayed an increasing rate of processing of the γ-secretase substrates APP-CTF and ApoER2-CTF, indicative of a rebound effect. These effects could be related to the reported failure of GSIs to achieve long-term Aβ regulation and their contribution to rather than the palliation of the AD pathology.
We found an increase in PS1 after a single day of DAPT administration to SH-SY5Y cells. Similar results were obtained in primary neuronal cultures treated for 4 days with DAPT and in rats treated for 21 days with avagacestat. The increase in PS1 protein was not paralleled by changes in PS1 mRNA content, indicating that this increase is not mediated by transcriptional upregulation. At present, the mechanism by which PS1 levels are enhanced by GSI administration remains unknown. Interestingly, over-expression of the β-secretase-derived APP fragment C99 could also mediate an increase in PS1. There is evidence that the accumulation of APP-C99 may be directly implicated in neurodegeneration and cognitive alterations . Previous evidences indicate that excess in other γ-secretase substrates can compromise γ-secretase catalytic activity, being accompanied by an increase in PS1 levels . Remarkably, it has been demonstrated that accumulation of APP-C99 can cause an impaired lysosomal-autophagic function . Hence, it seems desirable to investigate whether an excess of γ-secretase substrates may result in transient stabilization of PS1/γ-secretase substrate complexes, interfering in the effective clearance of PS1. Similarly, the stabilization of PS1/GSI complexes during sustained γ-secretase inhibition could interfere in the clearance/turnover of PS1. Indeed, decreased intracellular clearance of PS1 may also reflect the reduction of the CSF-PS1 complex levels, although how PS1 reaches the CSF is unknown.
In this context, it has been demonstrated that changes in PS1 ubiquitination can alter cellular levels of PS1 and other γ-secretase subunits, leading to an alteration in the metabolism of APP [45, 46]. Therefore, a chronic treatment with GSIs may cause a sustained accumulation of PS1 leading a rebound effect with gain in γ-secretase activity. In this regard, although avagacestat has demonstrated effect in the accumulation of APP-CTF (acute treatment in rats), prolonged exposition to the GSI (21 days of treatment in rats) has lead to PS1 accumulation. An increase in PS1 levels, even maintained in GSI treatment, could result in an increasing rate of substrate processing during the oscillations in the effective inhibitory concentration of the drug, derived of the half-life and QD dosing.
Moreover, other alternatives are suitable. There are subtle differences in the subcellular accumulation of APP-CTF in PS1-deficient cells, with no obvious redistribution of the full-length protein . Distinct subcellular locations of PSs have been shown to contribute to substrate specificity , and changes in the subcellular distribution of BACE1 induced by Aβ oligomers have been related to the pathogenesis of AD . In brief, both post-translational and turnover/degradation mechanisms may participate in the pernicious response to GSI and deserve investigation. Moreover, we cannot discard that other enzymes distinct from PS1, or acting in parallel, could be involved in the rebound effect, with an increased rating of γ-secretase substrate processing during prolonged inhibition.
Chronic inhibition of PS1 with GSI has led to toxic side effects in clinical trials [37, 50, 51]. These adverse effects were thought to be related with the regulation of Notch activity by γ-secretase, a protein that is important for cell-to-cell communication and that has also been implicated in cancer . Toxic side effects have been noted in clinical trials conducted with Notch-sparing GSIs as well as non-selective GSIs, although the true selectivity of the former is not clear . Indeed, dozens of additional substrates for γ-secretase have been identified and, thus, non-selective GSIs would probably interfere with multiple cellular events [53, 54]. Currently, clinical trials with semagacestat (LY450139), an earlier-generation GSI that does not discriminate well between APP and Notch, have been discontinued, similar to clinical trials with avagacestat. Furthermore, the development of another Notch-sparing GSI, begacestat (GSI-953) , has also been discontinued for reasons that are not clear (discussed in De Strooper and Chávez-Gutiérrez ).
The therapeutic effect of GSIs appears to be transient, and the possibility of decelerating or halting cognitive deterioration also falls below expectations. At 2 years, no significant differences were observed in key clinical outcome measures in an avagacestat phase 2 trial, yet progression to dementia was more frequent in the prodromal AD cohort vs the observational cohort . Similarly, semagacestat made AD patients cognitively worse in a phase 3 trial . In Tg2576 mice, a 1-day treatment with two GSIs significantly ameliorated cognitive deficits (acute effects) but these effects disappeared when an 8-day treatment schedule was employed. Indeed, prolonged treatment with GSIs impairs spatial working memory and cognitive function . In our study, an augmented latency time in the beam walking test in wild-type rats treated for 21 days with avagacestat suggests that some behavioral issues are affected by GSIs. This phenomenon is consistent with the dampening of initiative and the anxiety that are common neuropsychiatric features of AD [59, 60]. Interestingly, the conditional double presenilin knockout mice has observably altered anxiety-like behavior , and less anxiety is also displayed by transgenic mice expressing mutants PS1-A246E  and PS2-N141I . An association of PS1 with altered anxiety-like behavior has been suggested  and is worthy of further investigation. The subtle alterations in behavioral tests in wild-type rats are inconclusive since we did not use an animal model with an impaired condition, and nor did we demonstrate a direct association between altered anxiety-like behavior and increased brain PS1 levels. However, we speculate that part of the impairment observed in clinical trials involving GSI use on humans and in chronically treated animals could be due to rebound increases in PS1.
Although simple in concept, the validation of amyloid drug targets, and specifically that of GSIs, has proved complex in practice. Earlier studies indicated that the acute oral administration of DAPT to APPV717F transgenic mice reduces the Aβ in the brain . The use of canine  and non-human primate  models also served to demonstrate that GSIs decrease the Aβ peptides in the CSF. However, it is well established that the levels of AD CSF diminish when there is an increase in brain deposition of Aβ. Thus, changes in CSF-Aβ are unlikely to provide significant information about therapies aimed at reducing Aβ production, and a lowering of CSF-Aβ levels is unlikely to be a suitable measure of target engagement .
In this regard, acute administration of avagacestat robustly reduces CSF Aβ40 and Aβ42 levels similarly in rats and dogs . Moreover, the administration of a single dose of avagacestat to healthy humans, as well over a 28-day schedule, also markedly decreases Aβ40 and Aβ42 concentrations in the CSF [68, 69]. However, exploratory CSF amyloid isoforms displayed a dose-dependent but not significant reduction in a small subset of patients in a phase 2 trial, and while well tolerated, lower doses did not affect the Aβ40 and Aβ42 levels in treated patients . Similarly, earlier studies with semagacestat in volunteers indicated unchanged levels of CSF-Aβ , although in another study, single oral doses of semagacestat appeared to decrease Aβ levels in the CSF of healthy volunteers . No significant reduction in CSF Aβ42 or Aβ40 level was detected in a phase 2 safety trial , a finding verified by mass spectrometry analysis of the same samples . Instead, an increase in shorter Aβ peptides (Aβ1–14, Aβ1–15, and Aβ1–16) was identified, probably due to increased substrate availability (APP-C99) for α-secretase . Interestingly, semagacestat produced a decrease in plasma Aβ concentrations in a 6-h interval following drug administration, returning to baseline and then transiently increasing the Aβ concentrations . It was suggested that semagacestat might lower Aβ at high concentrations but cause Aβ elevation at low concentrations . A structurally related γ-secretase inhibitor, LY-411575, also elevated plasma Aβ40 and Aβ42 in Tg2576 mice . A biphasic activation-inhibition dose-response curve for GSIs was proposed to explain these changes in Aβ secretion . However, these changes may also be indicative of a transient overshooting or rebound effect, since an increase in plasma Aβ40 and Aβ42 has been described in Tg2576 mice chronically treated with DAPT .
Here, we addressed the efficiency of GSIs to inhibit PS1 by assessing changes in the cellular γ-secretase substrate APP-CTF. As expected, the accumulation of APP-CTF served to assess the inhibitory effect of DAPT on PS1 in cellular models and also that of avagacestat. Accordingly, we were able to detect accumulation in the brain levels of APP-CTF in acutely treated rats (sacrificed 4 h after a single dose). However, sustained inhibition of γ-secretase activity over 21 days revealed decreased APP-CTF levels, suggesting that the consolidation of higher PS1 levels in reaction to chronic inhibition results in an increase in γ-secretase activity, at least in the intervals between GSI administration. The consolidation of higher levels of PS1 might indiscriminately affect all γ-secretase substrates, such as ApoER2 and others, further exacerbating the AD pathology. Interestingly, administration of GSIs increased APP-CTF in H4 cells over-expressing APP, although this increase was unexpectedly attenuated at high concentrations . Elsewhere, APP-C99 levels increase in CHO cells co-expressing APP and PS1 relative to cells expressing APP alone, and PS1 can stabilize APP-CTF independent of γ-secretase activity . Hence, the relationship between the substrate and the catalytic enzyme appears to be more complex than might at first appear.
PS1 also participates in other cell functions [76, 77], and therefore, the increase in PS1 after GSI administration may influence distinct cellular effects, even if this subunit is not-catalytically active. In this regard, PS1 has been implicated in the physiological maturation and glycosylation of several key proteins implicated in AD, such as nicastrin , BACE1 , acetylcholinesterase , and others, including APP . Hence, the over-expression of either the wild-type or mutant PS1 disturbs glycoprotein processing . Further research will be needed to clarify the influence of increased PS1, under prolonged GSI administration, in the role of PS1 in their non-proteolytic functions, and possible interference with the therapeutic response.
We show here that administration of GSIs result in a rebound increase in PS1 levels in cellular and animal models, which must be taken into consideration when using such compounds in AD therapy. Indeed, our results indicate that the effect of GSI inhibitors on APP processing failed to have a long-term effect in treated rats, possibly due to the persistent PS1 elevation in reaction to chronic inhibition.
The outcomes of the clinical trials with GSIs have been disappointing, although this may not represent the end of the development of these drugs to treat AD. The data presented here indicate that the therapeutic benefits of GSIs and related drugs should continue to be explored, or at least, we can extract information that will help understand the failure of GSIs in AD trials . Hence, synthesizing new GSIs that distinguish strongly between APP and Notch may serve to lower the required dose, yet it still might not solve the unexplained and unexpected problem of the facilitation of toxic side effects and the AD-derived pathogenesis. The failure of GSIs in clinical trials highlights the need for a systematic re-examination of γ-secretase biology, including further characterization of the mechanisms related to the response to chronic inhibition. Elucidating the mechanisms involved, the complex self-regulation of γ-secretase is also important to optimize therapies based on γ-secretase modulation. A potent inhibition/modulation of secretase activities will result in the unbalanced generation of proteolytic fragments of APP (and fragments from other substrates), which could determine a self-regulatory response that will require further analysis for new secretase inhibitors/modulators designed to specifically inhibit the Alzheimer process.
In this regard, γ-secretase modulators (GSMs), which only block the γ-secretase cleavage of APP to generate the Aβ42, with no changes in the production of total Aβ, were also noticed to have negative outcomes . The clinical development of BACE1 inhibitors is also being intensely pursued, and several promising BACE1 inhibitors have entered human clinical trials , but a sign of toxicity forced to stop the earliest trials (discussed in Lahiri et al. ). For a successful development of new secretase inhibitors/modulators, it is needed to better understand the cellular response to the sustained inhibition/modulation of the secretase activity.
Despite its enzymatic capacity, γ-secretase activity appears tightly regulated by many cellular components, including its own subunits, modulatory partners, and substrates, as well as by an array of cellular events . Furthermore, GSIs are presently explored in clinical trials as potential therapeutic agents in cancer, targeting Notch, although a number of mechanism-based adverse events again emerge . As the therapeutic benefits of GSIs and related drugs continue to be explored, a better understanding of the response of PS1 to chronic inhibition will become more necessary.
We thank Dr. J. Culvenor and Dr. G. Evin (Department of Pathology, The University of Melbourne, Australia) for the gift of the PS1 antibodies and Dr. D. H. Small (Menzies Institute for Medical Research, University of Tasmania, Australia) for the advice and for providing the C99 cDNA. We also thank Bristol-Myers Squibb (Wallingford, CT, USA) for the generous gift of avagacestat. This study was funded in part by the EU BIOMARKAPD-Joint Programming on Neurodegenerative Diseases (JPND) project by the Instituto de Salud Carlos III (ISCIII grants PI11/03026 and PI15/00665 for JSV and PI14/00566 for MSGA), co-financed by the Fondo Europeo de Desarrollo Regional (FEDER) “Investing in your future,” and through CIBERNED (ISCIII). We also acknowledge financial support from the Spanish Ministerio de Economía y Competitividad, through the “Severo Ochoa” Programme for Centres of Excellence in R&D (SEV-2013-0317). The funding bodies played no role in the study design, the data collection and analysis, the decision to publish, or the preparation of the manuscript.
Compliance with Ethical Standards
All animal procedures were approved by the Animal Care and Use Committees at the Universidad Miguel Hernández and by Centro Principe Felipe (2016A/SC/PEA/00127).
Conflict of Interest
The authors declare that they have no competing interests.
- 13.Siemers E, Skinner M, Dean RA, Gonzales C, Satterwhite J, Farlow M et al (2005) Safety, tolerability, and changes in amyloid beta concentrations after administration of a gamma-secretase inhibitor in volunteers. Clin Neuropharmacol 28:126–132Google Scholar
- 14.Lanz TA, Hosley JD, Adams WJ, Merchant KM (2004) Studies of Aβ pharmacodynamics in the brain, cerebrospinal fluid, and plasma in young (plaque-free) Tg2576 mice using the γ-secretase inhibitor N2-[(2S)-2-(3,5-difluorophenyl)-2-hydroxyethanoyl]-N1-[(7S)-5-methyl-6-oxo-6,7-dihydro-5H-dibenzo[b,d]azepin-7-yl]-l-alaninamide (LY-411575). J Pharmacol Exp Ther 309:49–55Google Scholar
- 15.Lanz TA, Karmilowicz MJ, Wood KM, Pozdnyakov N, Du P, Piotrowski MA et al (2006) Concentration-dependent modulation of amyloid-beta in vivo and in vitro using the gamma-secretase inhibitor, LY-450139. J Pharmacol Exp Ther 319:924–933Google Scholar
- 16.Säemann MD, Böhmig GA, Osterreicher CH, Burtscher H, Parolini O, Diakos C et al (2000) Anti-inflammatory effects of sodium butyrate on human monocytes: potent inhibition of IL-12 and up-regulation of IL-10 production. FASEB J 14:2380–2382Google Scholar
- 17.Tanaka A, Hase S, Miyazawa T, Takeuchi K (2002) Up-regulation of cyclooxygenase-2 by inhibition of cyclooxygenase-1: a key to nonsteroidal anti-inflammatory drug-induced intestinal damage. J Pharmacol Exp Ther 300:754–761Google Scholar
- 18.García-Ayllón MS, Silveyra MX, Andreasen N, Brimijoin S, Blennow K, Sáez-Valero J (2007) Cerebrospinal fluid acetylcholinesterase changes after treatment with donepezil in patients with Alzheimer’s disease. J Neurochem 101:1701–1711Google Scholar
- 19.Xu L, Kappler CS, Mani SK, Shepherd NR, Renaud L, Snider P et al (2009) Chronic administration of KB-R7943 induces up-regulation of cardiac NCX1. J Biol Chem 284:27265–27272Google Scholar
- 20.Serfőző Z, Lontay B, Kukor Z, Erdődi F (2012) Chronic inhibition of nitric oxide synthase activity by N(G)-nitro-L-arginine induces nitric oxide synthase expression in the developing rat cerebellum. Neurochem Int 60:605–615Google Scholar
- 21.Campanari ML, García-Ayllón MS, Belbin O, Galcerán J, Lleó A, Sáez-Valero J (2014) Acetylcholinesterase modulates presenilin-1 levels and γ-secretase activity. J Alzheimers Dis 41:911–924Google Scholar
- 22.Sáez-Valero J, de Gracia JA, Lockridge O (2005) Intraperitoneal administration of 340 kDa human plasma butyrylcholinesterase increases the level of the enzyme in the cerebrospinal fluid of rats. Neurosci Lett 383:93–98Google Scholar
- 23.Sáez-Valero J, Tornel PL, Muñoz-Delgado E, Vidal CJ (1993) Amphiphilic and hydrophilic forms of acetyl- and butyrylcholinesterase in human brain. J Neurosci Res 35:678–689Google Scholar
- 24.Evin G, Sharples RA, Weidemann A, Reinhard FB, Carbone V, Culvenor JG et al (2001) Aspartyl protease inhibitor pepstatin binds to the presenilins of Alzheimer’s disease. Biochemistry 40:8359–8368Google Scholar
- 25.Basavarajappa BS, Subbanna S (2014) CB1 receptor-mediated signaling underlies the hippocampal synaptic, learning, and memory deficits following treatment with JWH-081, a new component of spice/K2 preparations. Hippocampus 24:178–188Google Scholar
- 26.García-Ayllón MS, Cauli O, Silveyra MX, Rodrigo R, Candela A, Compañ A et al (2008) Brain cholinergic impairment in liver failure. Brain 131:2946–2956Google Scholar
- 27.Goldstein LB, Davis JN (1990) Beam-walking in rats: studies towards developing an animal model of functional recovery after brain injury. J Neurosci Methods 31:101–107Google Scholar
- 28.Hånell A, Marklund N (2014) Structured evaluation of rodent behavioral tests used in drug discovery research. Front Behav Neurosci 8:252Google Scholar
- 29.Gonzalez-Usano A, Cauli O, Agusti A, Felipo V (2014) Pregnenolone sulphate restores the glutamate-nitric oxide-cGMP pathway and extracellular GABA in cerebellum and learning and motor coordination in hyperammonemic rats. ACS Chem Neurosci 5:100–105Google Scholar
- 30.Dovey HF, John V, Anderson JP, Chen LZ, de Saint AP, Fang LY et al (2001) Functional gamma-secretase inhibitors reduce beta-amyloid peptide levels in brain. J Neurochem 76:173–181Google Scholar
- 31.Balmaceda V, Cuchillo-Ibáñez I, Pujadas L, García-Ayllón MS, Saura CA, Nimpf J et al (2014) ApoER2 processing by presenilin-1 modulates reelin expression. FASEB J 28:1543–1554Google Scholar
- 32.Thinakaran G, Borchelt DR, Lee MK, Slunt HH, Spitzer L, Kim G et al (1996) Endoproteolysis of presenilin 1 and accumulation of processed derivatives in vivo. Neuron 17:181–190Google Scholar
- 33.Podlisny MB, Citron M, Amarante P, Sherrington R, Xia W, Zhang J et al (1997) Presenilin proteins undergo heterogeneous endoproteolysis between Thr291 and Ala299 and occur as stable N- and C-terminal fragments in normal and Alzheimer brain tissue. Neurobiol Dis 3:325–337Google Scholar
- 34.Tong G, Castaneda L, Wang JS, Sverdlov O, Huang SP, Slemmon R et al (2012) Multicenter, randomized, double-blind, placebo-controlled, single-ascending dose study of the oral γ-secretase inhibitor BMS-708163 (avagacestat): tolerability profile, pharmacokinetic parameters, and pharmacodynamic markers. Clin Ther 34:654–667Google Scholar
- 35.Coric V, van Dyck CH, Salloway S, Andreasen N, Brody M, Richter RW et al (2012) Safety and tolerability of the γ-secretase inhibitor avagacestat in a phase 2 study of mild to moderate Alzheimer disease. Arch Neurol 69:1430–1440Google Scholar
- 36.Coric V, Salloway S, van Dyck CH, Dubois B, Andreasen N, Brody M et al (2015) Targeting prodromal Alzheimer disease with avagacestat: a randomized clinical trial. JAMA Neurol 72:1324–1333Google Scholar
- 37.Gillman KW, Starrett JE Jr, Parker MF, Xie K, Bronson JJ, Marcin LR et al (2010) Discovery and evaluation of BMS-708163, a potent, selective and orally bioavailable γ-secretase inhibitor. ACS Med Chem Lett 1:120–124Google Scholar
- 38.Crump CJ, Castro SV, Wang F, Pozdnyakov N, Ballard TE, Sisodia SS et al (2012) BMS-708, 163 targets presenilin and lacks notch-sparing activity. Biochemistry 51:7209–7211Google Scholar
- 39.Albright CF, Dockens RC, Meredith JE Jr, Olson RE, Slemmon R, Lentz KA et al (2013) Pharmacodynamics of selective inhibition of γ-secretase by avagacestat. J Pharmacol Exp Ther 344:686–695Google Scholar
- 40.García-Ayllón MS, Campanari ML, Brinkmalm G, Rábano A, Alom J, Saura CA et al (2013) CSF presenilin-1 complexes are increased in Alzheimer’s disease. Acta Neuropathol Commun 1:46Google Scholar
- 41.Sogorb-Esteve A, García-Ayllón MS, Fortea J, Sánchez-Valle R, Lleó A, Molinuevo JL, Sáez-Valero J (2016) Cerebrospinal fluid presenilin-1 increases at asymptomatic stage in genetically determined Alzheimer’s disease. Mol Neurodegener 11:66Google Scholar
- 42.Zahniser NR, Doolen S (2001) Chronic and acute regulation of Na+/Cl−-dependent neurotransmitter transporters: drugs, substrates, presynaptic receptors, and signaling systems. Pharmacol Ther 92:21–55Google Scholar
- 43.Lauritzen I et al (2012) The β-secretase-derived C-terminal fragment of βAPP, C99, but not Aβ, is a key contributor to early intraneuronal lesions in triple-transgenic mouse hippocampus. J Neurosci 32:16243–16255Google Scholar
- 44.Lauritzen I, Pardossi-Piquard R, Bauer C, Brigham E, Abraham JD, Ranaldi S et al (2016) Intraneuronal aggregation of the β-CTF fragment of APP (C99) induces Aβ-independent lysosomal-autophagic pathology. Acta Neuropathol 132:257–276Google Scholar
- 45.Li J et al (2002) SEL-10 interacts with presenilin 1, facilitates its ubiquitination, and alters A-beta peptide production. J Neurochem 82:1540–1548Google Scholar
- 46.Massey LK, Mah AL, Monteiro MJ (2005) Ubiquilin regulates presenilin endoproteolysis and modulates gamma-secretase components, Pen-2 and nicastrin. Biochem J 391:513–525Google Scholar
- 47.Chen F, Yang DS, Petanceska S, Yang A, Tandon A, Yu G et al (2000) Carboxyl-terminal fragments of Alzheimer beta-amyloid precursor protein accumulate in restricted and unpredicted intracellular compartments in presenilin 1-deficient cells. J Biol Chem 275:36794–36802Google Scholar
- 48.Sannerud R, Esselens C, Ejsmont P, Mattera R, Rochin L, Tharkeshwar AK et al (2016) Restricted location of PSEN2/γ-secretase determines substrate specificity and generates an intracellular Aβ pool. Cell 166:193–208Google Scholar
- 49.Mamada N, Tanokashira D, Hosaka A, Kametani F, Tamaoka A, Araki W (2015) Amyloid β-protein oligomers upregulate the β-secretase, BACE1, through a post-translational mechanism involving its altered subcellular distribution in neurons. Mol Brain 8:73Google Scholar
- 50.Searfoss GH, Jordan WH, Calligaro DO, Galbreath EJ, Schirtzinger LM, Berridge BR et al (2003) Adipsin, a biomarker of gastrointestinal toxicity mediated by a functional gamma-secretase inhibitor. J Biol Chem 278:46107–46116Google Scholar
- 51.Wong GT, Manfra D, Poulet FM, Zhang Q, Josien H, Bara T et al (2004) Chronic treatment with the gamma-secretase inhibitor LY-411, 575 inhibits beta-amyloid peptide production and alters lymphopoiesis and intestinal cell differentiation. J Biol Chem 279:12876–12882Google Scholar
- 52.Wolfe MS (2012) γ-Secretase as a target for Alzheimer’s disease. Adv Pharmacol 64:127–153Google Scholar
- 53.Wakabayashi T, De Strooper B (2008) Presenilins: members of the gamma-secretase quartets, but part-time soloists too. Physiology (Bethesda) 23:194–204Google Scholar
- 54.Lleó A, Saura CA (2011) γ-secretase substrates and their implications for drug development in Alzheimer’s disease. Curr Top Med Chem 11:1513–1527Google Scholar
- 55.Martone RL, Zhou H, Atchison K, Comery T, Xu JZ, Huang X et al (2009) Begacestat (GSI-953): a novel, selective thiophene sulfonamide inhibitor of amyloid precursor protein gamma-secretase for the treatment of Alzheimer’s disease. J Pharmacol Exp Ther 331:598–6087Google Scholar
- 56.De Strooper B, Chávez-Gutiérrez L (2015) Learning by failing: ideas and concepts to tackle γ-secretases in Alzheimer’s disease and beyond. Annu Rev Pharmacol Toxicol 55:419–437Google Scholar
- 57.Doody RS, Raman R, Farlow M, Iwatsubo T, Vellas B, Joffe S et al (2013) A phase 3 trial of semagacestat for treatment of Alzheimer’s disease. N Engl J Med 369:341–350Google Scholar
- 58.Mitani Y, Yarimizu J, Saita K, Uchino H, Akashiba H, Shitaka Y et al (2012) Differential effects between γ-secretase inhibitors and modulators on cognitive function in amyloid precursor protein-transgenic and nontransgenic mice. J Neurosci 32:2037–2050Google Scholar
- 59.Frisoni GB, Rozzini L, Gozzetti A, Binetti G, Zanetti O, Bianchetti A et al (1999) Behavioral syndromes in Alzheimer’s disease: description and correlates. Dement Geriatr Cogn Disord 10:130–138Google Scholar
- 60.Chung JA, Cummings JL (2000) Neurobehavioral and neuropsychiatric symptoms in Alzheimer’s disease. Neurol Clin 18:829–846Google Scholar
- 61.Yan L, Li L, Han W, Pan B, Xue X, Mei B (2013) Age-related neuropsychiatric symptoms in presenilins conditional double knockout mice. Brain Res Bull 97:104–111Google Scholar
- 62.Lalonde R, Qian S, Strazielle C (2003) Transgenic mice expressing the PS1-A246E mutation: effects on spatial learning, exploration, anxiety, and motor coordination. Behav Brain Res 138:71–79Google Scholar
- 63.Yuk DY, Lee YK, Nam SY, Yun YW, Hwang DY, Choi DY et al (2009) Reduced anxiety in the mice expressing mutant (N141I) presenilin 2. J Neurosci Res 87:522–531Google Scholar
- 64.Luciano M, Houlihan LM, Harris SE, Gow AJ, Hayward C, Starr JM et al (2010) Association of existing and new candidate genes for anxiety, depression and personality traits in older people. Behav Genet 40:518–532Google Scholar
- 65.Borghys H, Tuefferd M, Van Broeck B, Clessens E, Dillen L, Cools W et al (2012) A canine model to evaluate efficacy and safety of γ-secretase inhibitors and modulators. J Alzheimers Dis 28:809–822Google Scholar
- 66.Cook JJ, Wildsmith KR, Gilberto DB, Holahan MA, Kinney GG, Mathers PD et al (2010) Acute gamma-secretase inhibition of nonhuman primate CNS shifts amyloid precursor protein (APP) metabolism from amyloid-beta production to alternative APP fragments without amyloid-beta rebound. J Neurosci 30:6743–6750Google Scholar
- 67.Lleó A, Cavedo E, Parnetti L, Vanderstichele H, Herukka SK, Andreasen N et al (2015) Cerebrospinal fluid biomarkers in trials for Alzheimer and Parkinson diseases. Nat Rev Neurol 11:41–55Google Scholar
- 68.Tong G, Wang JS, Sverdlov O, Huang SP, Slemmon R, Croop R et al (2012) Effects of single doses of avagacestat (BMS-708163) on cerebrospinal fluid Aβ levels in healthy young men. Clin Drug Investig 32:761–769Google Scholar
- 69.Dockens R, Wang JS, Castaneda L, Sverdlov O, Huang SP, Slemmon R et al (2012) A placebo-controlled, multiple ascending dose study to evaluate the safety, pharmacokinetics and pharmacodynamics of avagacestat (BMS-708163) in healthy young and elderly subjects. Clin Pharmacokinet 51:681–693Google Scholar
- 70.Siemers ER, Dean RA, Friedrich S, Ferguson-Sells L, Gonzales C, Farlow MR et al (2007) Safety, tolerability, and effects on plasma and cerebrospinal fluid amyloid-beta after inhibition of gamma-secretase. Clin Neuropharmacol 30:317–325Google Scholar
- 71.Bateman RJ, Siemers ER, Mawuenyega KG, Wen G, Browning KR, Sigurdson WC et al (2009) A gamma-secretase inhibitor decreases amyloid-beta production in the central nervous system. Ann Neurol 66:48–54Google Scholar
- 72.Fleisher AS, Raman R, Siemers ER, Becerra L, Clark CM, Dean RA et al (2008) Phase 2 safety trial targeting amyloid beta production with a gamma-secretase inhibitor in Alzheimer disease. Arch Neurol 65:1031–1038Google Scholar
- 73.Portelius E, Dean RA, Gustavsson MK, Andreasson U, Zetterberg H, Siemers E et al (2010) A novel Abeta isoform pattern in CSF reflects gamma-secretase inhibition in Alzheimer’s disease. Alzheimers Res Ther 2:7Google Scholar
- 74.Svedružić ŽM, Popović K, Šendula-Jengić V (2013) Modulators of γ-secretase activity can facilitate the toxic side-effects and pathogenesis of Alzheimer’s disease. PLoS One 8:e50759Google Scholar
- 75.Pitsi D, Octave JN (2004) Presenilin 1 stabilizes the C-terminal fragment of the amyloid precursor protein independently of gamma-secretase activity. J Biol Chem 279:25333–25338Google Scholar
- 76.De Strooper B, Iwatsubo T, Wolfe MS (2012) Presenilins and γ-secretase: structure, function, and role in Alzheimer disease. Cold Spring Harb Perspect Med 2:a006304Google Scholar
- 77.Stiller I, Lizák B, Bánhegyi G (2014) Physiological functions of presenilins; beyond γ-secretase. Curr Pharm Biotechnol 15:1019–1025Google Scholar
- 78.Leem JY, Vijayan S, Han P, Cai D, Machura M, Lopes KO et al (2002) Presenilin 1 is required for maturation and cell surface accumulation of nicastrin. J Biol Chem 277:19236–19240Google Scholar
- 79.Hébert SS, Bourdages V, Godin C, Ferland M, Carreau M, Lévesque G (2003) Presenilin-1 interacts directly with the beta-site amyloid protein precursor cleaving enzyme (BACE1). Neurobiol Dis 13:238–245Google Scholar
- 80.Silveyra MX, Evin G, Montenegro MF, Vidal CJ, Martínez S, Culvenor JG (2008) Presenilin 1 interacts with acetylcholinesterase and alters its enzymatic activity and glycosylation. Mol Cell Biol 28:2908–2919Google Scholar
- 81.Leem JY, Saura CA, Pietrzik C, Christianson J, Wanamaker C, King LT (2002) A role for presenilin 1 in regulating the delivery of amyloid precursor protein to the cell surface. Neurobiol Dis 11:64–82Google Scholar
- 82.Farquhar MJ, Gray CW, Breen KC (2003) The over-expression of the wild type or mutant forms of the presenilin-1 protein alters glycoprotein processing in a human neuroblastoma cell line. Neurosci Lett 346:53–56Google Scholar
- 83.De Strooper B (2014) Lessons from a failed γ-secretase Alzheimer trial. Cell 159:721–726Google Scholar
- 84.Xia W, Wong ST, Hanlon E, Morin P (2012) γ-Secretase modulator in Alzheimer’s disease: shifting the end. J Alzheimers Dis 31:685–696Google Scholar
- 85.Yan R, Vassar R (2014) Targeting the β secretase BACE1 for Alzheimer’s disease therapy. Lancet Neurol 13:319–329Google Scholar
- 86.Lahiri DK, Maloney B, Long JM, Greig NH (2014) Lessons from a BACE1 inhibitor trial: off-site but not off base. Alzheimers Dement 10:S411–S419Google Scholar
- 87.Gertsik N, Chiu D, Li YM (2015) Complex regulation of γ-secretase: from obligatory to modulatory subunits. Front Aging Neurosci 6:342Google Scholar
- 88.Naoko Takebe, Dat Nguyen, Sherry X. Yang, (2014) Targeting Notch signaling pathway in cancer: Clinical development advances and challenges. Pharmacol Ther 141(2):140–149Google 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.