Amyloid beta and its naturally occurring N-terminal variants are potent activators of human and mouse formyl peptide receptor 1

Abstract

Formyl peptide receptors (FPRs) may contribute to inflammation in Alzheimer's disease through interactions with neuropathological Amyloid beta (Aβ) peptides. Previous studies reported activation of FPR2 by Aβ_1-42, but further investigation of other FPRs and Aβ variants is needed. This study provides a comprehensive overview of the interactions of mouse and human FPRs with different physiologically relevant Aβ-peptides using transiently transfected cells in combination with calcium imaging. We observed that, in addition to hFPR2, all other hFPRs also responded to Aβ_1-42, Aβ_1-40, and the naturally occurring variants Aβ_11-40 and Aβ_17-40. Notably, Aβ_11-40 and Aβ_17-40 are very potent activators of mouse and human FPR1, acting at nanomolar concentrations. Buffer composition and aggregation state are extremely crucial factors that critically affect the interaction of Aβ with different FPR subtypes. To investigate the physiological relevance of these findings, we examined the effects of Aβ_11-40 and Aβ_17-40 on the human glial cell line U87. Both peptides induced a strong calcium flux at concentrations that are very similar to those obtained in experiments for hFPR1 in HEK cells. Further immunocytochemistry, qPCR, and pharmacological experiments verified that these responses were primarily mediated through hFPR1. Chemotaxis experiments revealed that Aβ_11-40 but not Aβ_17-40 evoked cell migration, which argues for a functional selectivity of different Aβ peptides. Together, these findings provide the first evidence that not only hFPR2 but also hFPR1 and hFPR3 may contribute to neuroinflammation in Alzheimer's disease through an interaction with different Aβ variants.

Keywords: Alzheimer's disease; amyloid beta; formyl peptide receptors; glia; neuroinflammation.

Figure 1

Aβ_1-42activates all human FPRs.A, representative Ca²⁺ traces of HEK293T cells transiently transfected with FPR plasmids or an empty vector (mock) after stimulation with Aβ_1-42 obtained from P&E. B, mean peak Ca²⁺ responses of human (red) and mouse (blue) FPRs upon stimulation with different concentrations of Aβ_1-42. Buffer (gray) denotes responses to the assay buffer without Aβ_1-42. Bars represent mean values of three independent experiments (n = 3) carried out as technical duplicates (N = 2). All Error bars, S.D.; One-way ANOVA test, Dunnett post hoc test; ∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001; ns, no significance. Aβ, amyloid beta; FPR, Formyl peptide receptor.

Figure 2

Manufacturer- and solvent-effects on FPR activation by Aβ_1-42.A, left: Schematic depiction of ThT aggregation assay. Right: Mean fluorescence of 22.5 μM Aβ_1-42 in C1 buffers from different manufacturers in a ThT aggregation assay during the first 10 min (clear bars) versus fluorescence after 120 min (striped bars). Buffer refers to ThT fluorescence without addition of peptides (gray bars). All n = 3, N = 3, except for Sigma and Anaspec with n = 2, N = 3; One-way ANOVA test, Dunnett post hoc test. B, mean Ca²⁺ peak responses of human (red) or mouse (blue) FPRs to 10 μM of Aβ_1-42 peptides obtained from Peptides & Elephants (P&E) and Synpeptide; n = 3, N = 2, One-way ANOVA test, Dunnett post hoc test in comparison to respective buffer controls. C, heat map of mean Ca²⁺ responses of FPRs elicited by Aβ_1-42 peptides obtained from five different manufacturers. The scale ranges from white (no response) to deep orange (ΔF/F0 ≥ 0.4). Responses are shown in Figure S1B. D, secondary structure composition of four Aβ_1-42 peptides analyzed by circular dichroism (CD) spectroscopy; n = 3, N = 1; One-way ANOVA test, Tukey post hoc test. E, mean Ca²⁺ peak responses of cells transfected with human FPRs (red) or mock (gray) towards 5 μM Aβ_1-42 dissolved in the respective buffers; n = 3, N = 1, One-way ANOVA test, Dunnett post hoc test. F, comparison of ThT fluorescence of Aβ_1-42 (P&E) dissolved in either C1, dimethyl sulfoxide (DMSO), Tris–NaCl, or HBSS. All assays were performed in the respective buffers. For experiments with peptides predissolved in DMSO, assays were conducted in C1 with a final concentration of DMSO: 0.2% (V/V). One-way ANOVA test, Dunnett post hoc test. All Error bars, S.D.; ∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001; ns, no significance. Aβ, amyloid beta; FPR, Formyl peptide receptor; ThT, thioflavin T.

Figure 3

Naturally occurring N-abridged Aβ fragments activate FPR1 tenfold better than Aβ_1-42.A, comparison of mean Ca²⁺ peak responses of human (red) or mouse (blue) FPRs to a stimulation with 10 μM Aβ_1-40 and Aβ_1-42 or 5 μM of the natural occurring N-abridged variants Aβ_11-40 and Aβ_17-40 or with 10 μM of the C-abridged variants Aβ_1-10 and Aβ_1-16. Colored bars indicate responses of human (red) or mouse (blue) FPRs, n = 3, N = 1. B, concentration response curves of selected variants, n = 3, N = 1. C, left: scheme indicating the size and location of the different Aβ variants. Right: Table depicting the proposed 3D-structures and thresholds for minimal detectable activation during Ca²⁺ imaging of the responding Aβ variants. All Error bars, S.D. One-way ANOVA test, Dunnett post hoc test; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001; ns, no significance. Aβ, amyloid beta; FPR, Formyl peptide receptor.

Figure 4

N-abridged Aβ peptides induce hFPR1-dependent responses in glial U87 cells.A, mean Ca²⁺ peak responses of U87 cells after stimulation with different concentrations of Aβ_11-40 or Aβ_17-40.Striped bars indicate the response towards 10 μM of the positive control WKWVm-NH₂, light gray and dark gray bars indicate negative controls; n = 3, N = 1. B, comparison of the Ca²⁺ responses upon stimulation with either 5 μM Aβ_11-40 or Aβ_17-40 alone (green bars) or in the presence of 10 μM of the competitive FPR-antagonist tBoc2 (black bars); n = 3, N = 1. C, dose-dependent chemotaxis of U87 cells upon stimulation with either Aβ_11-40 or Aβ_17-40. Each bar represents the number of cells that migrated through a porous membrane towards the respective stimuli. Green bars indicate migration towards N-abridged fragments, striped bars display migration towards the positive control 1 μM fMLF, light gray bars show migration without stimuli, and dark gray bars represent the response to 0.1% DMSO, n = 3, N = 1. D, migration of U87 cells that were either untreated (green bars) or treated with 10 μM tBoc2 (black bars) towards 1 μM of Aβ_11-40 or Aβ_17-40; n = 3, N = 1. All Error bars, S.D. One-way ANOVA test, Dunnett post hoc test for A and C and t test for B and D; ∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001; ns, no significance. Aβ, amyloid beta; FPR, Formyl peptide receptor.

Figure 5

The activation of U87 cells by N-abridged Aβ peptides depends on hFPR1.A, PCR experiments show that U87 cells contain high mRNA levels of hFPR1 but only low amounts of FPR2 and FPR3. Left: Representative image of gel electrophoresis after reverse transcription polymerase chain reaction with primers for all hFPRs. Right: Quantification of complementary DNA for all FPRs obtained through reverse transcription quantitative polymerase chain reaction, n = 5, N = 2. Details on the primer efficiency, specificity, and linearity are given in Figures S6 and S7. One-way ANOVA test, Dunnett post hoc test. B, left: representative immunocytochemistry staining of U87 cells and transfected HEK293T cells with FPR subtype–specific antibodies (red) and nuclei staining (blue). For visibility, brightness and contrast were adjusted for U87 cells and HEK293T cells differently; for absolute intensity comparison pictures with equal settings for exposure time, brightness and contrast pictures are shown in Figure S5. Scale bars indicate 100 μm. Right: Quantification of the mean FPR staining in U87 cells (green) in comparison to FPR-transfected HEK293T cells (red). Analysis was performed on images acquired with the same settings. n = 2, N = 5 for U87 cells and n = 2, N = 2 for HEK293T cells. Evidence for closely similar staining intensities of FPR1 and FPR2 by the used protocols is shown in Figure S5. One-way ANOVA test, Tukey post hoc test. C, the concentration-dependent Ca²⁺ responses of HEK293 cells and U87 cells towards the N-abridged fragments Aβ_11-40 and Aβ_17-40 reveal a clear correlation with the hFPR1 response, n = 3, N = 1. D, U87 cells respond to the potent hFPR1 activator SP6 (10 nM) but not to a potent hFPR2 agonist SP4 (10 nM) in calcium imaging experiments, n = 3, N = 3. One-way ANOVA test, Dunnett post hoc test. E, Cross-desensitization experiment of U87 cells show that the Ca²⁺ responses towards Aβ_17-40 are abolished by the FPR1 activator SP6 but not by the hFPR2 agonist SP4. Representative Ca⁺ traces (left) and mean Ca²⁺ peak responses (right) to a secondary Aβ_17-40 stimulus after pre-application of SP4 or SP6 as a first stimulus, n = 1, N = 3; t test. F, U87 migrate towards the hFPR1-stimulus SP6 (10 nM) but do not respond to the hFPR2 agonist SP4 (10 nM), n = 3, N = 1. One-way ANOVA test, Tukey post hoc test. All Error bars, S.D. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001; ns, no significance. Aβ, amyloid beta; FPR, Formyl peptide receptor.

Figure 6

Minimal requirements for the investigation of Aβ interactions with FPRs. Aβ, amyloid beta; FPR, Formyl peptide receptor.

Supplemental Figure S1

Manufacturer-dependent FPR activation by Aβ1-42 A, representative TEM images of Aβ1-42 obtained from either Peptides&Elepehants (P&E), Synpeptide, or Anaspec show different types of aggregates. AS-HFIP refers to a commercially available HFIP-treated peptide from Anaspec. Scale bars indicate 0.5 μm. B, mean Ca2+ peak responses of FPR-transfected HEK293T cells after stimulation with 10 μM Aβ1-42 obtained from different manufacturers. Colored bars indicate the response of human (red) or mouse (blue) FPRs, n=3, N=2. Data correspond to the heat map shown in Fig 2B. Bar graphs for P&E and Synpeptide are shown again for the sake of comparability. C, complete CD spectra of all four Aβ1-42 displayed in figure 2. N=3. All Error bars, S.D.; One-way ANOVA test, Dunnett post hoc test; ∗, p ≤ 0.05; ∗∗, p ≤ 0.01; ∗∗∗, p ≤ 0.001; ns, no significance

Supplemental Figure S2

Alterations of FPR responses to Aβ1-42 after extended storage. Comparison of the Ca2+ responses of human (red) or mouse (blue) FPRs to 10 μM Aβ1-42 (Synpeptide) that were either freshly dissolved in the assay buffer C1 (clear bars) or had been stored in C1 at -20° C for 6 months without freeze-thaw cycles (striped bars), n=1, N=2. All Error bars, S.D.; ∗, p ≤ 0.05; ns, no significance.

Supplemental Figure S3

DMSO influences FPR activation by Aβ1-42. Mean Ca2+ peak responses of human (red) or mouse (blue) FPRs to 5 μM of Aβ1-42 peptides obtained from P&E, Synpeptide or Anaspec dissolved in either DMSO (striped bars) or C1 (clear bars), n=3, N=2. All Error bars, S.D.; ∗, p ≤ 0.05; ns, no significance.

Supplemental Figure S4

Aβ17-40 is less susceptible to manufacturer-dependent effects than full length Aβ. Mean Ca2+ peak responses of FPR-transfected HEK293T cells to 5 μM of Aβ17-40 from three different manufacturers. Clear bars, Aβ17-40 from Anaspec. Striped bars, Aβ17-40 from Synpeptide. Dotted bars: Aβ17-40 from Sigma-Aldrich, n=3, N=1. All Error bars, S.D.; One-way ANOVA test, Dunnett post hoc test; ∗, p ≤ 0.05; ns, no significance.

Supplemental Figure S5

Validation of immunofluorescence staining with FPR subtype-specific antibodies. Top: Comparison of immunofluorescence staining intensity of FPR-transfected HEK293T and U87 cells with receptor subtype specific antibodies. The hFPR1 (1 μg/ml), hFPR2 (0.2 μg/ml) and hFPR3 (1 μg/ml) antibodies were used. Note that the absolute staining intensities of all hFPRs in HEK293T cells were highly similar, however a strong reduction of hFPR2 staining in U87 cells is observed. No cross reactivity was detected. All images were acquired with the same settings and are displayed with the same illumination, brightness and contrast. Evidence for similar cell surface expression and production rates for hFPR1 and hFPR2 can be found in supporting Fig. S8. Bottom: corresponding nuclei staining with Hoechst 33342 demonstrating a comparable amount of total cell number in all visual fields.

Supplemental Figure S6

FPR1 is the dominant receptor in U87 cells. A, Ct values of all human FPRs from the RNA of U87 cells obtained through RT-qPCR, n=5, N=2. B, Ca2+ responses of U87 cells (green) or FPR-transfected HEK293T cells (red) towards 5 μM Aβ17-40 with (clear bars) or without (striped bars) co-application of 10 μM tBoc2. Signals were normalized to responses of the respective cells towards 10 μM WKYMVm, n=1, N=2 for HEK293T cells and N=3 for U87 cells. C, mean Ca2+ peak responses of U87 cells (green) or FPR-transfected HEK293T cells (red) that were treated with 10 nM SP4 (left) or 10 nM SP6 (right), n=1, N=3. All Error bars, S.D. One-way ANOVA test, Dunnett post hoc test; ∗, p ≤ 0.05, ∗∗, p ≤ 0.01, ∗∗∗, p ≤ 0.001; ns, no significance

Supplemental Figure S7

Validation of RT-qPCR primers for all human FPRs. Left: mean standard curves, regression line, and amplification efficiency (n =3, N=3) generated by using a 10-fold serial dilution of a target DNA template starting with 0,1 ng of a purified and sequenced PCR product in t-RNA. Ct values were plotted against the log of template quantity for each dilution. Representative PCR products for all primer sets were controlled by gel electrophoresis and sequencing for their specificity. Right: A representative data set of an amplification curve from these dilution series for each primer.

Supplemental Figure S8

Evidence for similar cell surface expression and calcium responses of FPR1 and FPR2 in HEK293T cells. A, HEK293T cells were transfected with varying amounts of plasmid DNA for hFPR1 or hFPR2 and subsequently treated with 1 μM of FITC-labeled WKWVm-NH2 to monitor their cell surface expression. Left: representative images of treated cells transfected with different DNA amounts Right: quantitative analysis of positively stained cells. Note that cell surface expression of both receptors upon dilution is reduced in a similar manure. Error bars, S.D. n=1, N=3 B, Mean Ca2+ peak responses of HEK293T cells transfected with decreasing amounts of human (red) or mouse (blue) FPR plasmid DNA. FPR1 and 2 of each species were treated with 1 μM of WKWVm-NH2 and FPR3-transfected cells with 30 μM of WKWVm-CHO. Note that the amount of FPR 1 and 2 DNA can be diminished by more than 10fold without a strong reduction of the signal amplitudes.