Six DNA constructs, each encoding an FAD-mutant γ-secretase, were generated, with all PSEN1 mutations corresponding to those under study by the Dominantly Inherited Alzheimer Network (DIAN). The selected mutations were S169L, S170F, G378E, F386S, A431E, A434T (Figure 1B). PSEN1 mutations S169L and S170F are in transmembrane domain 3 (TMD3), near the hinge region with TMD2. Mutations G378E and F386S are situated in TMD7, close to the catalytic aspartate residue D385 in the conserved GxGD motif. Mutations A431E and A434T are found in a highly conserved region that includes the PALP motif, which is essential for enzymatic activity (Wang et al., 2004; Tomita et al., 2001). The age of onset for these mutations varies as follows: S169L (29–31 years) (Taddei et al., 1998), S170F (mean 27 years) (Snider et al., 2005), G378E (38–44 years) (Lanoiselée et al., 2017), F386S (37–58 years) (Raux et al., 2005), A431E (mean 43 years) (Dumois-Petersen et al., 2020), and A434T (35 years) (Jiao et al., 2014). Monocistronic pMLINK plasmids encoding human PSEN1, each harboring one of these mutations, were individually prepared via mutagenesis of the wild-type (WT) construct. Each mutant PSEN1 DNA insert was cut out with restriction enzymes from this monocistronic plasmid and inserted into a tricistronic construct encoding the other three components of the γ-secretase complex (nicastrin, Aph-1, and Pen-2) through ligation-independent cloning (LIC). The resulting tetracistronic constructs encode all four components of the protease complex (Figure 1—figure supplement 1; Lu et al., 2014).

The tetracistronic constructs, each encoding either WT or one of the six FAD-mutant forms of the γ-secretase complex, were transiently transfected into human embryonic kidney (HEK)293F cells, and the expressed protease complexes were subsequently purified (Devkota et al., 2021; Devkota et al., 2024). Concurrently, two versions of the recombinant FLAG epitope-tagged version of C99 substrate (C100-Flag) were expressed in E. coli: one in normal media, and the other in M9 minimal media containing ¹⁵NH₄Cl and ¹³C-glucose as the sole sources of nitrogen and carbon, respectively. Purification provides light and heavy isotope-labeled C100-Flag (Figure 1—figure supplement 2A). Enzyme purity was verified by western blot analysis, which demonstrated the presence of all four components of γ-secretase, with PSEN1 autoproteolyzed into N-terminal and C-terminal fragments (NTF and CTF), indicative of maturation of the complex to the catalytically active protease (Figure 1—figure supplement 2B). Additionally, the quality of the substrates was assessed using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and silver staining (Figure 1—figure supplement 2C).

Each purified mutant protease (30 nM) was incubated with saturating levels of the heavy-isotope-labeled substrate (5 µM) at 37°C for 16 hr. In parallel, the WT enzyme was incubated with light-isotope-labeled C100-Flag substrate. Due to the slow rate for this reaction (WT enzyme k_cat = 2 hr^–1), this long incubation period is still within the linear range for the formation of products (i.e. the enzyme remains saturated with substrate throughout the incubation period) (Devkota et al., 2021). The resulting products for all proteolytic events for each of these samples were then analyzed using MS techniques as described below. Due to the hydrophobicity and insolubility of the Aβ products, quantification by direct MS posed significant challenges. Consequently, we focused on analysis of the coproducts (AICDs and small peptides) and indirectly calculation of the concentration of the Aβ products.

Rates for the initial endoproteolytic step (ε cleavage) were determined by quantifying AICD species (Figure 1A) using MALDI-TOF MS. We utilized synthetic AICD 50–99-Flag and AICD 49–99-Flag peptides to construct standard concentration curves by MALDI-TOF MS (Figure 1C). This approach enabled accurate measurements of the production of AICD 50–99 (coproduct of Aβ49) and AICD 49–99 (coproduct of Aβ48) for both WT and mutant enzymes. For all standards and experimental samples, equal concentrations of insulin were added as an internal standard. The standard curves were generated based on the intensity ratio of the AICD parent ion to that of the internal standard (Figure 1C).

Using these standard curves, we calculated concentrations of AICD species in our test samples (Figure 1D). Equal concentrations of both AICD 50–99 and AICD 49–99 were produced when comparing WT protease incubated with light-isotope C100 to WT protease incubated with heavy-isotope C100. This demonstrates that use of heavy C100 does not affect the concentrations of products formed during the ε cleavage step. Quantification of AICD species was possible for four mutant proteases but not for A431E and A434T, as product concentrations from these two mutant enzymes were below the detection limit (Figure 1D, Figure 1—figure supplement 3), consistent with the known essential role of the PALP motif in proteolytic activity (Wang et al., 2004; Tomita et al., 2001). All FAD-mutant PSEN1-containing enzymes except for F386S exhibited a significant reduction in the production of both AICD 50–99 and AICD 49–99, which also measures formation of coproducts Aβ49 and Aβ48, respectively. Reduced ε cleavage of APP and Notch substrate has been observed with other PSEN1 FAD mutations (Devkota et al., 2024; Song et al., 1999; Bentahir et al., 2006). Uniquely, the PSEN1 F386S mutant protease resulted in a significant increase in production of AICD 49–99 and similar levels of AICD 50–99 compared to those observed with WT protease. Three other mutant enzymes (S169L, G378E, and F386S) likewise biased ε cleavage toward AICD 49–99, thereby favoring the Aβ48 → Aβ42 pathway, consistent with previous reports on PSEN1 FAD mutations (Sato et al., 2003).

Equal volumes of WT enzyme incubated with light-isotope C100 and PSEN1 FAD-mutant enzymes incubated separately with heavy-isotope C100 were combined in a 1:1 ratio after quenching the reactions (Figure 2A). The 1:1 mixtures were then used to quantify the coproducts of the trimming steps along the two canonical pathways Aβ49→Aβ46→Aβ43→Aβ40 and Aβ48→Aβ45→Aβ42 →Aβ38 via liquid chromatography coupled with tandem MS (LC-MS/MS) (Takami et al., 2009; Devkota et al., 2021), as we previously described for six other FAD-mutant enzymes (Devkota et al., 2024). Mixing of the enzyme reactions after quenching but before analysis allows quantification of each small peptide coproduct formed from WT enzyme and mutant enzyme in the same LC-MS/MS run: Products from WT enzyme incubated with light-isotope C100 have lower mass than products from mutant enzyme incubated with heavy-isotope C100 (Figure 2A). Standard curves were generated for the tri- and tetrapeptide coproducts of each trimming step, using synthetic ITL, VIV, IAT, VIT, TVI, and VVIA peptides, allowing accurate determination of the concentrations of each coproduct generated from all WT and PSEN1 FAD-mutant enzyme reaction mixtures. Equal concentrations of each tri- and tetrapeptide were produced when comparing WT protease incubated with light-isotope C100 to WT protease incubated with heavy-isotope C100, demonstrating that the use of heavy-isotope C100 does not affect the concentrations of products formed during these cleavage steps (Figure 2B).

Quantification of small peptide coproducts from WT vs. PSEN1 FAD-mutant enzyme reactions revealed that each of the six mutations leads to a specific profile of effects on carboxypeptidase trimming. Mutations located within or near the conserved PALP motif (Wang et al., 2004; Tomita et al., 2001), A431E and A434T, exhibited significantly lower levels of trimming coproducts compared to WT and other mutants (Figure 2C, Figure 2—figure supplement 1). Although AICD levels from A431E and A434T PSEN1 mutant proteases were below those of the lowest standards, very low levels of certain small peptide coproducts were detected within the range of their standard curves. In strong contrast, F386S PSEN1 mutant enzyme produced equal levels of ITL (coproduct of Aβ49→Aβ46) and VIV (coproduct of Aβ46→Aβ43) compared to WT. However, levels of IAT (coproduct of Aβ43→Aβ40) were substantially lower for this mutant enzyme, indicating a buildup of Aβ43 (Figure 2C, Figure 2—figure supplement 1). Along the Aβ42 pathway, F386S PSEN1 enzyme produced levels of VIT (coproduct of Aβ48→Aβ45) only slightly lower than WT enzyme; however, because F386S generated increased levels of AICD 49–99 (coproduct of C99→Aβ48), this mutant also led to a buildup Aβ48. For mutants S169L, S170F, and G378E, small peptide coproducts were generally lower than those generated from WT (Figure 2C, Figure 2—figure supplement 1), consistent with their reduced initial ε proteolysis (Figure 1D).

Quantification of all coproducts (AICDs, small peptides) using MS techniques enabled calculation of the percent cleavage for each trimming step in the processive proteolysis of APP by γ-secretase (Figure 2D). That is, given the level of Aβ precursor produced, how much of this precursor was cleaved in the next step? This analysis revealed that none of the mutant enzymes exhibited deficiencies in the Aβ49→Aβ46 or Aβ46→Aβ43 cleavage steps compared to WT enzyme, although percent cleavage could not be determined for A431E and A434T, as these mutant enzymes produced coproducts AICD 50–99 and ITL below the limits of detection. The S169L and F386S mutations demonstrated significant changes in the Aβ43→Aβ40 step, with S169L showing a marked increase and F386S showing a decrease compared to WT. All mutations were deficient in the Aβ48→Aβ45 cleavage step, except for S170F, which exhibited an increase in this trimming step. It should be pointed out, however, that this increased percent efficiency of S170F—over 200%—is not possible, suggesting either under-detection of AICD 49–99 (production of Aβ48) or over-detection of VIT (degradation of Aβ48) for this mutant enzyme. The reason for this discrepancy is not clear. All mutations except for S169L and A434T were also deficient in the Aβ45→Aβ42 trimming step. Furthermore, all mutations were deficient in the Aβ42 →Aβ38 cleavage process, except for S169L and A431E, which showed increases compared to WT enzyme. We should point out, however, that detected products from A431E and A434T were very low and therefore likely making calculations of percent cleavage unreliable. This detailed analysis highlights specific deficiencies and alterations in the cleavage patterns associated with each mutation, providing a comprehensive and quantitative understanding of how these mutations affect the proteolytic processing of APP by γ-secretase. Overall, calculation of percent efficiency for each trimming step, along with analysis of effects on initial ε cleavage, revealed that all PSEN1 FAD-mutant proteases are deficient in multiple cleavage steps in processive proteolysis of APP substrate, consistent with the previous findings for the other six PSEN1 FAD mutations (Devkota et al., 2024).

By analyzing the extent of degradation and production of each Aβ peptide, we were able to calculate the levels of each Aβ peptide in the final quenched enzyme reaction mixtures (Table 1). We observed a negative value for the concentration of Aβ48 produced from the S170F enzyme, due to the discrepancy pointed out above in quantifying the coproducts of Aβ48 production and degradation. To validate the calculated concentrations of Aβ peptides obtained by LC-MS/MS, we used Aβ40- and Aβ42-specific ELISAs (Figure 3A). The ELISA results were consistent with the calculations from the LC-MS/MS data for all mutations except for F386S. This mutation exhibited higher concentrations of both Aβ40 and Aβ42 when measured by ELISA compared to the calculations derived from MS data (Figure 3A vs. B). Because the PSEN1 F386S mutant enzyme produced high levels of Aβ43 (Table 1), as previously reported (Liu et al., 2023), we tested for cross-reactivity in the ELISAs with synthetic Aβ43 peptide. We found that the putatively Aβ42- and Aβ40-specific ELISAs both cross-reacted with Aβ43 (Supplementary file 1, Supplementary file 2), which can explain the observed discrepancies with F386S enzyme. According to the ELISA results, all six PSEN1 FAD-mutant enzymes produced lower levels of Aβ40 and Aβ42 compared to WT, consistent with our previous findings with six other PSEN1 FAD mutations (Devkota et al., 2024). Mutants S169L, G378E, and F386S showed an increased Aβ42/Aβ40 ratio by both ELISA and LC-MS/MS (Figure 3). In contrast, Aβ42/Aβ40 produced from the S170F mutant enzyme was equivalent to that of WT by both methods. Mutants A431E and A434T appear to increase Aβ42/Aβ40 by LC-MS/MS; however, these two FAD-mutant enzymes produce such low levels of peptide coproducts that the calculated ratio is not reliable. By ELISA, only Aβ42 was detectable from A431E and A434T mutant enzymes, precluding calculation of Aβ42/Aβ40 ratios.

All data describe biological replicates, and experiments were replicated in two independent experiments.

To test the effects of FAD mutations on the stability of γ-secretase E-S complexes, we conducted FLIM in intact cells (Elangovan et al., 2002). WT or FAD-mutant PSEN1 was co-expressed with C99 substrate in HEK293 cells in which endogenous PSEN1 and PSEN2 were knocked out through CRISPR/Cas9 gene editing (Liu et al., 2021). The C99 substrate construct consisted of human APP C99 flanked by an N-terminal signal peptide and a C-terminal near-infrared fluorescence protein: miRFP720 (C99-720) (Figure 4A). Transfected cells were fixed and permeabilized and treated with primary antibodies that bind to the N-terminal region of C99/Aβ (mouse antibody 6E10) and with an epitope on the nicastrin component of γ-secretase (rabbit antibody NBP2-57365) that lies in close proximity to the N-terminal region of bound C99/Aβ. Secondary antibodies conjugated to fluorophores (anti-mouse IgG antibody conjugated with Alexa Fluor 488 and anti-rabbit IgG antibody conjugated with Cy3) were then added. In this experimental design, reduction in the fluorescence lifetime of the Alexa Fluor 488, through fluorescence resonance energy transfer to Cy3, indicates E-S complex detection. Importantly, use of rabbit antibody toward a distal region in the γ-secretase complex results in no change in Alexa Fluor 488 fluorescence lifetime (Devkota et al., 2024).

For each but one PSEN1 FAD mutant tested, Alexa Fluor 488 fluorescence lifetime is reduced compared to that of WT PSEN1 (Figure 4B and C), consistent with increased stability of E-S complexes. The miRFP720 fusion to the C-terminus of C99 provides a means of distinguishing between C99-rich regions in the cells (low ratio of 6E10 to C99-720) from Aβ-rich regions (high ratio of 6E10 to C99-720) (Figure 4A; Devkota et al., 2024; Maesako et al., 2022; McKendell et al., 2022). Although the PSEN1 F386S mutant did not show an overall fluorescence lifetime reduction compared to WT, when differentiating C99-rich regions vs. Aβ-rich regions (Figure 4D), fluorescence lifetime was significantly reduced in Aβ-rich regions and not C99-rich regions (Figure 4E vs. F). This finding is consistent with the results of proteolytic analysis, as PSEN1 F386S is the only mutant that was not deficient in ε cleavage but was deficient in specific trimming steps, especially Aβ43→Aβ40 and Aβ48→Aβ45. We should point out that the observed decrease in donor lifetime with the PSEN1 FAD mutants might also be due to lower levels of C99-720 expression or higher levels of PSEN1 CTF (i.e. mature γ-secretase complexes). However, C99-720 intensities are not significantly different between cells transfected with WT and those with FAD PSEN1 (Figure 4—figure supplement 1A). Furthermore, western blot analysis shows that levels of C99-720 are not significantly low and those of PSEN1 CTF are not high in FAD PSEN1 compared to WT PSEN1 expressing cells (Figure 4—figure supplement 1B and C). Although PSEN1 CTF levels trend low for PSEN1 F386S, this mutant resulted in decreased FLIM only in Aβ-rich regions. Thus, the reduced Alexa 488 fluorescence lifetime apparently reflects effects of FAD mutation on E-S complex stability.

While detailed and quantitative studies of the impact of FAD mutations on all proteolytic stages of γ-secretase processing of C99 have been carried out here, these experiments utilized purified enzymes and substrates in a detergent-solubilized system. This system differs significantly from the cell membrane environment, which is characterized by complex lipid and protein compositions and dynamic microdomains that can affect protein folding, stability, and function. Therefore, the effects of these mutations on γ-secretase processing of C99 may vary considerably when studied in detergent-solubilized systems compared to their natural membrane environment. However, we previously noted that detergent-solubilized and reconstituted proteoliposome systems give similar ratios of AICD 49–99/AICD 50–99 and Aβ42/Aβ40 (Devkota et al., 2021). While using MALDI-TOF instead of western blotting provided more accurate quantification of AICDs, we encountered challenges with lower concentrations of AICDs produced by mutations such as A431E and A434T. The concentrations of AICDs in these mostly deficient mutations fell below the detection limit of our instrument, limiting our ability to assess their levels accurately. In addition, FLIM studies relied on overexpression of PSEN1 and C99-720, which could lead to artifacts.

Expi293F cells, a HEK cell line purchased from Thermo Fisher, were utilized to produce γ-secretase complexes protein. The stock received from Invitrogen was grown and aliquoted for direct use in these studies without authentication or testing for mycoplasma contamination. HEK293 cells with PSEN1/2 double knockout by CRISPR (Liu et al., 2021), a gift of Dr. Lei Liu (Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA), were used for transfection and FLIM. The cells were authenticated using STR profiling and monitored for mycoplasma contamination every 2 months using MycoAlert Mycoplasma Detection Kit (LONZA, Basel, Switzerland).

Expi293F cells were cultured in Expi293 expression medium (Thermo Fisher Scientific, A1435101). Transient transfection occurred once the cell density reached 3×10⁶ cells/mL. The cells were kept at 37°C, shaken at 125 rpm, and under 8% CO₂. Harvesting took place when cell viability decreased to 75%.

E. coli DH5α was employed for molecular cloning and plasmid preparation and was cultured in LB medium at 37°C with constant shaking. E. coli BL21 (DE3) was used for the transduction and expression of the γ-secretase substrate C100-Flag (Li et al., 2000), grown in LB medium at 37°C with continuous shaking until OD₆₀₀ reached 0.8, at which point expression was induced using 0.5 mM IPTG.

γ-Secretase was produced and purified from Expi293F cells as described in earlier studies (Bolduc et al., 2017; Fraering et al., 2004; Osenkowski et al., 2009). Briefly, Expi293F cells were grown in Expi293 expression medium supplemented with penicillin-streptomycin until they reached a density of 3×10⁶ cells/mL. Before transfection, the medium was exchanged for fresh Expi293 expression medium (without penicillin-streptomycin). For the transfection process, 100 µg of the pMLINK tetracistronic vector and 320 µL of ExpiFectamine 293 Reagent were mixed in 12 mL of Opti-MEM I Reduced Serum Medium and incubated at room temperature for 20 min. The resulting ExpiFectamine 293/plasmid DNA complexes were then added to the cell culture. After 20 hr, ExpiFectamine 293 Transfection Enhancer 1 and Enhancer 2 were added, and the cells were cultured until their viability dropped to 75%. The cells were then harvested by centrifugation at 300 × g for 5 min and resuspended in a buffer containing 50 mM MES (pH 6.0), 150 mM NaCl, 5 mM CaCl₂, and 5 mM MgCl₂. They were lysed by passing twice through a French press. Unbroken cells and debris were removed by centrifugation at 3000 × g for 10 min, and the supernatant was ultracentrifuged at 100,000 × g for 1 hr to obtain the membrane pellet. The membrane pellet was washed with 0.1 M sodium bicarbonate (pH 11.3) by sequentially passing through syringes with 18-, 22-, 25-, and 27-gauge needles, followed by another ultracentrifugation at 100,000 × g for 1 hr. The pellet was resuspended in 50 mM HEPES (pH 7), 150 mM NaCl, and 1% CHAPSO, and incubated at 4°C with shaking for 1 hr. This mixture was ultracentrifuged again at 100,000 × g. The supernatant was incubated with anti-FLAG M2-agarose beads in TBS containing 0.1% digitonin for 16 hr at 4°C with shaking. The beads were washed three times with TBS/0.1% digitonin before eluting the γ-secretase complex using a buffer with 0.2 mg/mL FLAG peptide in TBS/0.1% digitonin. The eluate was stored at –80°C until needed.

Two versions of C100-Flag, light- and heavy-isotope-labeled C100-Flag were prepared. E. coli BL21 cells transformed with C100-FLAG construct in pET22b vector (Li et al., 2000) were grown in media at 37°C with 180 rpm shaking until OD₆₀₀ reached 0.8. The standard media and M9 minimal media with 20% ¹³C-glucose (Cambridge Isotope Laboratories) and ¹⁵NH₄Cl (Cambridge Isotope Laboratories) were used to produce light-isotope labeling and heavy-isotope-labeled C100-Flag substrate respectively. M9 Minimal media composition: 3.4 g of anhydrous N₂HPO₄, 8.794 g of KH₂PO₄, 0.25 g of NaCl, 0.5 g of ¹⁵NH₄Cl, 10 mL of 20% ¹³C glucose, 1 mL of 1 M MgSO₄-7H₂O, 10 µL of 1 M CaCl₂-2H₂O, 500 µL of 0.5% thiamine-HCl, 5 mL of BME vitamin solution (Sigma-Aldrich), and 10 µL of 1 M FeSO₄-7H₂O were dissolved into 500 mL sterilized and deionized water. Cells were induced with 0.5 mM IPTG and cultured for 3 hr. After harvesting by centrifugation, the cells were resuspended in lysis buffer (25 mM Tris, pH 8, and 1% Triton X-100) and lysed by passing through a French press three times. The cleared lysate was incubated with anti-FLAG M2-agarose beads (Sigma-Aldrich) at 4°C for 16 hr with shaking. The beads were washed three times with lysis buffer, and the C100-FLAG protein was eluted using a buffer containing 100 mM glycine at pH 2.7 and 0.25% NP-40, followed by neutralization with Tris buffer at pH 8 and stored at –80°C. The composition and integrity of C100-FLAG were confirmed by SDS-PAGE with western blotting and MALDI-TOF MS.

The following antibodies were used: Anti-Nicastrin (Novus Biologicals, NBP2-57365), Anti-PSEN-1-NTF (Bio-Legend, 823401), anti-PSEN-1-CTF (Cell Signaling, 5643), Anti-Aph-1 (Bio-Legend, 823101), and Anti-Flag M2 (Sigma-Aldrich, F1804). Western blot analysis was carried out using standard procedures. Protein levels of the purified γ-secretase enzyme and the C100 substrate were quantified using a BCA assay (Thermo Fisher Scientific, USA) and normalized for different expression levels based on band intensity. These proteins were separated by SDS-PAGE and transferred onto PVDF membranes. Immunoblotting was then performed and visualized using the enhanced chemiluminescence method.

The in vitro γ-secretase assay was conducted as previously detailed (Devkota et al., 2021). In summary, 30 nM of either WT or FAD-mutant γ-secretase was preincubated for 30 min at 37°C in an assay buffer containing 750 µL of 50 mM HEPES (pH 7.0), 150 mM NaCl, 125 µL of DOPC (1,2-dioleoyl-sn-glycero-3-phosphocholine), and 125 µL of DOPE (1,2-dioleoyl-sn-glycero-3-phosphoethanolamine). The reactions were initiated by adding the light- or heavy-isotope form of the C100-FLAG substrate to a final concentration of 5 mM and incubated at 37°C for 16 hr. The reactions were terminated by flash freezing in liquid nitrogen and stored at –20°C.

As previously mentioned, small peptides were examined by LC-MS/MS utilizing an ESI quadrupole time-of-flight (Q-TOF) mass spectrometer (Q-TOF Premier, Waters) (Devkota et al., 2021). Before being injected into a C18 analytical chromatography column, the reaction mixtures containing light C100-FLAG/WT γ-secretase and heavy C100-Flag/FAD γ-secretase were mixed 1:1. The reaction mixtures were then eluted using a step gradient of 0.08% aqueous formic acid, acetonitrile, isopropanol, and a 1:1 acetone/dioxane mixture. Tandem MS was used to determine which three collision-induced dissociation fragments were the most prevalent for each small peptide. A C18 analytical chromatography column was loaded with different amounts of the synthetic peptide standards (>98% purity, New England Peptide) after they had been dissolved in an assay buffer for LC-MS/MS analysis. Employing an ion mass width of 0.02 unit, the signals from the three most abundant ions were added to produce a peptide chromatographic area. The ‘V’ mode was used for data acquisition.

AICD-FLAG in the reaction mixture was immunoprecipitated with anti-FLAG M2 beads (Sigma-Aldrich) in 10 mM MES pH 6.5, 10 mM NaCl, 0.05% DDM detergent for 16 hr at 4°C. AICD products were eluted from the anti-FLAG beads with acetonitrile: water (1:1) with 0.1% trifluoroacetic acid. In parallel, different concentrations of AICD standards (2500 nM, 2000 nM, 1000 nM, 500 nM, 250 nM, 125 nM, and 62.5 nM) were prepared using synthetic peptides. To all samples and standards, 5 nM of ProteoMass Insulin MALDI-MS was added as an internal standard. Finally, the elutes were analyzed on a Bruker Autoflex MALDI-TOF MS.

To quantify Aβ peptides, we analyzed reactions from in vitro cleavage assays involving purified γ-secretase complexes and the C100-FLAG substrate using specific ELISA kits from Invitrogen, in accordance with the manufacturer’s instructions.

Fluorescence emissions were collected using the ET525/50m-2p filter (Chroma Technology Corp, Bellows Falls, VT, USA). The donor Alexa Fluor 488 lifetime was recorded using a high-speed photomultiplier tube (MCP R3809; Hamamatsu photonics, Hamamatsu City, Japan) and a time-correlated single-photon counting acquisition board (SPC-830; Becker & Hickl GmbH, Berlin, Germany). The acquired FLIM data were analyzed using SPC Image software (Becker & Hickl GmbH). Pseudo-colored images corresponding to the ratios of 6E10-Alexa Fluor 488 over C99-720 emission were generated in MATLAB (MathWorks, Natick, MA, USA).

Materials are available from the authors upon request.