Aggregated Tau-PHF6 (VQIVYK) Potentiates NLRP3 Inflammasome Expression and Autophagy in Human Microglial Cells

The human microglial clone 3 (HMC3) cell line was purchased from ATCC^® (Product Number: CRL-3304™). The HMC3 cell line is derived from human primary microglial cells and is widely considered to be a unique experimental model for glial experiments [38]. Cells were grown in a humidified setting in an incubator at 37 °C with a 5% CO₂ supply. The culture medium Gibco™ Dulbecco’s modified Eagle medium (DMEM) was supplemented with 10% fetal bovine serum (FBS) and 0.5% penicillin-streptomycin (P/S) (all from Life Technologies, Waltham, MA, USA). For experiments, 25,000 cells/cm² were seeded on poly-L-ornithine (PLO) coated experimental plates in phenol-red free Gibco^TM Roswell Park Memorial Institute (RPMI) 1640 medium (Thermo Fisher Scientific, Waltham, MA, USA) containing 5% FBS and 0.5% P/S. The stimulation of cells with aggregated or unaggregated PHF6 was performed 24 h after seeding (Figure 1).

Tau-PHF6 (Ac-VQIVYK-NH₂) was purchased from Bachem (Bubendorf, Switzerland). One mg of the peptide was dissolved in 1 mL of ultrapure water. The stock solutions of 55 μM Heparin and 3.14 mM Thioflavin T dye (ThT) (both from Sigma–Aldrich, Darmstadt, Germany) were also prepared in ultrapure water. For the aggregation experiment, as detailed by KrishnaKumar et al. 2018, stock solutions of peptide, heparin, and ThT were diluted in 20 mM MOPS buffer (pH 7.2) so that the final working concentrations in a 96-well Corning^® dark flat bottom plate were 50 μM, 10 μM, and 100 μM, respectively [7]. For each well, 100 μL of aggregation mixture was used. Control wells were set up, having water in place of peptide or with the buffer only. Kinetic fluorescent measurement was performed in quadruplicate every 20 s at λ_ex (excitation) = 440 nm and λ_em (emission) = 485 nm using the Infinite^® M200 plate reader (Tecan, Zurich, Switzerland) for over an hour. For cell culture experiments, the aggregation mixture contained peptide and heparin, but no ThT because it could be toxic to the cells. The aggregation mixture was later diluted with the cell culture medium to match the final required concentration on cells. Apart from the kinetic measurement assay, a turbidity assay was also performed at 350 nm (without ThT) using the Tecan plate reader, which helps to observe the difficulty of the light to pass through the aggregate formation over time.

An appropriate number of cells were seeded on 24-well plates in RPMI 1640 media for 24 h before stimulation with the appropriate concentration of aggregated and unaggregated PHF6. After 6 and 24 h stimulation, protein isolation for WB, cell viability assay, immunocytochemistry, RNA isolation and qRT-PCR, and autophagy assay were performed.

After the ThT assay, fibrils were visualized on a glass-slide by Leica DM6000 B (Leica Microsystems, Wetzlar, Germany) microscope using the green channel at a 63X (oil immersion) objective magnification.

CD spectra were measured on an AVIV 62D CD spectrometer. All measurements were performed using a 1 mm glass cuvette. The concentration of the assembled tau-derived PHF6 probe was adjusted to 50 μM. Wavelength-dependent circular dichroism of the probe was recorded in the range of 190–250 nm and plotted against the applied wavelength λ. Further, using the webserver BeStSel [39], a secondary structure prediction was performed from the CD spectroscopy data.

To explain the short term and long-term effects of different concentrations of the tau-PHF6 peptide (aggregated and unaggregated) on human microglial HMC3 cells, a dose-dependent study was carried out with increasing concentrations (1, 5, 10, and 20 μM) of the peptide at two different time points (6 and 24 h).

After 6 and 24 h of stimulation, a phenol-chloroform-based extraction method was used for the RNA isolation [40]. The growth medium was removed and replaced with the peqGOLD TriFast™ (Peqlab, Darmstadt, Germany) reagent according to the manufacturer’s protocol. NanoDrop™ 1000 (Thermo Fisher Scientific, Waltham, MA, USA) reported the RNA purity and concentration. RNA samples were regarded as pure if the 260/280 ratio indicated 2.0 ± 0.1. For the reverse transcription, 1 μg/mL of total RNA was used to produce cDNA using the Moloney murine leukemia virus (M-MLV) reverse transcription kit and random primers (both from Invitrogen™, Waltham, MA, USA). Semi-quantitative polymerase chain reaction (PCR) was performed with reference housekeeping gene primers B2M (β2-microglobulin) and HPRT (hypoxanthine phosphoribosyltransferase 1) (Table 1). Mastercycler gradient S (Eppendorf, Hamburg, Germany) was used for the reaction. The PCR run protocol was as follows: 95 °C for 3 min, at 95 °C for 40 s (denaturation), at the annealing temperature of the primer (Table 2) for 40 s, at 72 °C for 45 s, at 72 °C for 45 s (extension), followed by 72 °C for 2 min and final hold at 4 °C. The steps from denaturation up to extension were repeated 40 times. Nucleic acid bands were separated by gel electrophoresis (constant voltage of 160 V for 20 min) using 2% agarose gels containing Midori Green Advance DNA staining dye (Nippon Genetics, Duren, Germany). Gels were finally visualized in Vilber E-box VX2 (Peqlab, Darmstadt, Germany).

For quantitative real-time PCR (qRT-PCR), the samples were diluted 10-fold with ultrapure water. Appropriate amounts of AceQ qPCR SYBR^® Green (CliniSciences, Nanterre, France), primers (Table 1), and ultrapure water were used for the master mixture. CFX Connect™ Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) was used with the following configuration: at 95 °C for 10 min (enzyme activation), at 95 °C for 15 s (denaturation), at the annealing temperature of the primers (Table 1) for 30 s, at 72 °C for 30 s (amplification), followed by a melting curve analysis. The cycle of denaturation up to amplification was repeated 40 times. All target gene expressions were normalized to the expression levels of the housekeeping genes (B2M and HPRT). Apart from the analysis using CFX Maestro software (Bio-Rad, Hercules, CA, USA), gel electrophoresis was also performed to evaluate the specificity of amplified products.

After stimulation of microglial HMC3 cells for 6 and 24 h, protein isolation was performed using the radioimmunoprecipitation assay buffer (RIPA), supplemented with protease inhibitor cOmplete mini™ (Sigma–Aldrich, Darmstadt, Germany). For measuring the protein concentration, Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) was used with absorbance recorded at λ = 562 nm using the Tecan Infinite^® M200 plate reader. After heating the samples at 95 °C for 8 min, they were loaded (20 μg protein per lane) onto the gels, and electrophoresis was performed at 80 V for 15 min followed by 140 V for 1 h. Blotting of the gel onto methanol-activated polyvinylidene difluoride (PVDF) membranes (Trans-Blot^® Turbo™ RTA Mini PVDF Transfer Kit, Bio-Rad, Hercules, CA, USA) was performed using the Trans-Blot^® Turbo™ Transfer System (Bio-Rad, Hercules, CA, USA) at 25 V for 50 min. Ponceau S was used to verify the success of the blotting. After re-activation of the membrane with methanol and subsequent blocking for 1 h with 5% milk in TBS, membranes were incubated with the primary antibody overnight at 4 °C in the blocking solution. On the next day, after washing, the secondary antibody was added to visualize the immunoreactive signals using the Vilber Fusion Solo detection system (Peqlab, Darmstadt, Germany). Table 2 lists the antibodies used for Western blot. For chemiluminescence detection of bands, the Westar Supernova (Cyanagen, Bologna, Italy) substrate reagent was used. Using ImageJ software (NIH, Bethesda, MD, USA), densitometric analysis of bands was performed and later normalized to the β-actin housekeeper protein control bands.

Microglial HMC3 cells were seeded on PLO-coated coverslips, and after stimulation and proper incubation time, cells were 3.7% PFA-fixed. Then, cells were washed with phosphate-buffered saline (PBS) and permeabilized with Triton-X. After blocking, incubation with the primary antibody was done overnight at 4 °C. After washing, a fluorescent-labeled secondary antibody was added for 2 h. The nuclei were counterstained with Hoechst 33342, Trihydrochloride, Trihydrate (Invitrogen™, Waltham, MA, USA). Table 3 lists all antibodies used for ICC. The Leica DM6000 B microscope was used to capture fluorescence images, with identical microscope settings maintained for each experiment.

Fluorescence intensity was evaluated by FIJI software [41]. The mean grey value was determined per sample and the background was subtracted. Afterwards, relative quantity to the control group was calculated. A total of four pictures were used per group and experiment. The fluorescence intensity is a linear grade and can be used for determination of protein levels when the same exposure settings are used for acquisition [42].

The CASP1 activity for the determination of inflammasome activation was determined by using the Caspase-Glo^® 1 Inflammasome Assay (Promega, Madison, WI, USA), following the manufacture’s protocol. In brief, half of the supernatant was removed and replaced with the 1:2 dilution of the ready-to-use Glo1 reagent and incubated for 60 min at room temperature. For NLRP3 inhibition, the cells were pre-incubated for 1 h with the NLRP3-specific inhibitor MCC950 (1 µM; AdipoGen Life Science, Liestal, Switzerland). Finally, luminescence signals were read out with the Tecan Infinite^® M200 plate reader.

After recovering the supernatants from the culture plates, protein concentration was measured by the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) with absorbance recorded at λ = 562 nm using the Tecan Infinite^® M200 plate reader. Human IL1β/IL1F2 DuoSet ELISA (R&D Systems, Minneapolis, MN, USA) was performed as per the manufacturer’s protocol. Specialized ELISA 96-well plate was incubated with the captured antibody overnight, followed by 1 h of blocking and then the addition of standard/samples overnight. Next, a detection antibody was added for 2 h at room temperature. Then, Streptavidin-HRP conjugate was added for 20 min, followed by the addition of substrate solution for 20 min. Nigericin (20 µM, Sigma–Aldrich, Darmstadt, Germany), LPS (1 µg/mL, Sigma–Aldrich, Darmstadt, Germany), or the combination of both reagents, were used as positive controls. The Tecan Infinite^® M200 plate reader was then used to measure the color development at 405 nm.

Because autophagy association has previously been reported in the case of Aβ [29], the formation of autophagosomes due to tau-PHF6 stimulation was tracked using the CYTO-ID^® autophagy detection kit 2.0 (Enzo Life Sciences, Farmingdale, NY, USA). A mixture of rapamycin (0.5 μM) and chloroquine (10 μM) was used as a positive control, as specified by the manufacturer. For microscopic analysis, Hoechst and CYTO-ID^® detection reagent 2 (1:500 final dilution) were directly added to the cells growing in 24-well plates. Cells were imaged using the Leica DM6000 B microscope with a 63X oil immersion objective. For the plate reader measurement, the detection reagent 2 (1:1000 final dilution) and Hoechst were added to cells growing in 96-well plates, and a fluorescence measurement was recorded at λ_ex = 480 nm and λ_em = 530 nm using the M200 Tecan plate reader.

Experiments were conducted in triplicates (gene expression) or quadruplets (all other experiments), and data are represented as the arithmetic means ± SEM. In prior experiments, the needed n values were calculated using G*Power [43] according to the first gene expression data with the following settings: f = 1.6; α = 0.05; β = 0.9; 3 groups. G*Power calculated, for the gene expression analysis, that 3 n per group were needed. Because we used gene expression data for the calculation for the further experiment, we added one extra n for the other experiments, due to lower effects being expected. Before analysis, using the Shapiro–Wilk normality test, a Gaussian distribution check of residual data points was performed. Further, using the Bartlett test, an equal variance was evaluated. If one of the tests returns significance, data transformation was performed using the Box-Cox method, and reevaluation of the tests was done for the newly generated transformed data points. Finally, a one-way or two-way (only for grouped analyses) ANOVA, followed by Tukey’s post-hoc test for inter-group differences, was performed. Statistical significance was set at 95% confidence interval with p values set as */# ≤ 0.05, **/## ≤ 0.01 ***/### ≤ 0.001.

When incubated in the presence of heparin, PHF6 peptide solution containing ThT reported a sharp increase in fluorescence in 10–15 min, attaining a sigmoid curve with a distinct lag phase and a plateau (Figure A1a). This effect strongly suggests the presence of high amyloidogenic aggregated species in the mixture [44,45]. The turbidity assay also supported the aggregated species formation (Figure A1b). CD spectroscopy data depicted a strong positive peak at 196 nm and a negative curve between 210–215 nm, which are the characteristics of β-sheet structures [46,47]. Predictions from the BeStSel server also indicated a high percentage of β-sheet along with random coil. Finally, for the cell experiments, two different forms of PHF6 were used, one being the aggregated sample with β-sheet structures (denoted as aPHF6) and the other being unaggregated monomeric PHF6 (denoted as PHF6).

The HMC3 cells were stimulated with aPHF6 for 6 and 24 h to identify the short- and long-term effects on cell viability. Cell toxicity measured by the LDH release after 6 h of stimulation with 20 μM PHF6 reached 10% compared to the lysed cells (positive control). After 24 h of exposure, the LDH release was around 20% of the positive control (Figure 2a). There was no significant dose-dependent cytotoxic effect of aPHF6 or PHF6 on microglial cells across the used concentrations of 1–20 μM (Figure A2a). As indicated with the CTB assay, after 6 h of stimulation with 20 μM aPHF6 or PHF6, the metabolic activity almost doubled compared to the untreated control (Figure 2b), whereas with lesser concentrations of aPHF6 and PHF6 (1–10 μM), the metabolic activity after 6 h also increased but remained around 1.5 times that of the control (Figure A2b). After 24 h of stimulation, however, the metabolic rate decreased to the control levels across all different groups with no dose-dependency being observed.

In the first step, the expression of NLRP3 after aPHF6 stimulation in HMC3 cells was evaluated. For this, a trial was done with four different concentrations of the aPHF6 (1, 5, 10, and 20 μM) and two different concentrations of PHF6 (10 and 20 μM) across two different time points (6 and 24 h). A sharp increase in the mRNA level of NLRP3 was observed with 20 μM aPHF6 stimulation at 6 h compared to the control, which eventually returned to the basal level after 24 h (Figure A3). With the lesser concentrations of aPHF6 or PHF6 (1–10 μM), the gene expressions were comparable to the control levels.

Hence, 20 μM aPHF6 and 20 μM PHF6 samples were finally selected for further evaluation. With the aPHF6 stimulation, a seven-fold increase in the NLRP3 gene expression was observed after 6 h compared to the control (Figure 3a). In contrast, after 24 h, the gene expression level indicated only a 1.5-time increase (Figure 3d). For the 20 μM PHF6, however, no such increase was observed. Densitometric analysis of Western blots also depicted an almost two-fold increase in the protein expression levels after 6 h of aPHF6 stimulation in HMC3 cells compared to the control group (Figure 3b), whereas after 24 h, no shifts were detected (Figure 3e). With the 20 μM PHF6, however, no significant increase in protein expression was observed at both time points. Immunofluorescence staining of cells was performed using Iba1, NLRP3, and Hoechst (nuclear staining). Around a 1.5-fold increase in NLRP3 fluorescence intensity after 6 h of 20 μM aPHF6 stimulation was observed, as quantified by ImageJ (Figure 3c). After 24 h, however, NLRP3 intensity was found comparable to the control levels across all different groups, with no significant increase in protein expression being observed (Figure 3f).

The adapter ASC (PYCARD) is an active component of the active NLRP3 inflammasome. Similar to NLRP3, a time-dependent increase in the expression levels of mRNA and protein was also witnessed with PYCARD/ASC in aPHF6-induced HMC3 cells. After 6 h of aPHF6 stimulation, the gene expression increased by 15-fold compared to the control (Figure 4a). However, after 24 h of stimulation, the gene expression significantly abated to the basal levels of the untreated control. Western blot analysis also depicted a similar trend of about a 2.5-fold increase in the ASC protein expression after 6 h of aPHF6 stimulation (Figure 4b), whereas, after 24 h, the increase was only two-fold with respect to the control. With the 20 μM PHF6, however, no significant increase was observed at both time points. To evaluate the effect of aPHF6 on microglial inflammation, gene expression levels of several inflammation markers such as NLRP1, IL12A, IL18, and TLR4 were also checked (Figure A4). Compared to the control, NLRP1 and IL12A registered about a 5- and 15-fold increase in the gene expression, respectively, after 6 h of aPHF6 stimulation. Similarly, IL18 and TLR4 have shown a 20- and 30-fold increase, respectively. After 24 h, however, the inflammation marker gene expressions returned to the basal levels.

Because NLRP3 inflammasome complex upregulation could lead to the secretion of cytokines to make neighboring cells aware of the inflammation, besides the expression of IL1B, the CASP1 activity was also analyzed. After 6 h, a significant increase in the CASP1 activity was measured in the cells only after aPHF6 stimulation (Figure 5a). The unaggregated PHF6 showed no effect on CASP1. In the MCC950 pre-incubated cells, no increasedCASP1 activity was measured (Figure 5b), indicating a strong NLRP3 participation in CASP1-mediated cytokine maturation. After 24 h of aPHF6 stimulation, a CASP1 activity increase was absent and not further modulated by MCC950 pre-incubation (Figure 5c,d). Regarding the IL1B gene expression, a six-fold increase was observed in 20 μM aPHF6 stimulated cells compared to the PHF6-stimulated and untreated cells after 6 h of stimulation (Figure 6a). However, after 24 h, no changes in the IL1B gene expression were detected in either of the treatments (Figure 6d). Western blot observations for the protein expression of the mature IL1β form (17 kDa) revealed a four-fold increase in the expression after 6 h of aPHF6 stimulation, whereas that after 24 h remained comparable to the control group (Figure 6c,f). Note that the IL1β pro-form (31 kDa) was also increased after 6 h but decreased after 24 h in comparison to the control groups (no quantification). With the 20 μM PHF6, however, no significant increase in gene or protein expression was observed at both time points. Moreover, using sandwich ELISA (Figure A6), no color change was observed in comparison to the used positive controls (Nigericin or LPS), indicating no release of IL1β into the supernatant. Then the gene expression of M2 markers such as IL10, MRC1, CD163, and CD200R1 was measured. Compared to the control, IL10 and MRC1 registered an almost 300 times increase in the gene expression after 6 h of aPHF6 stimulation. Similarly, CD163 and CD200R1 reported 500-, and 800-fold increases, respectively. After 24 h, however, the marker gene expressions returned to the basal levels (Figure A5). A substantial increase in gene expression of M2 markers was observed compared to M1 markers after 6 h of aPHF6 stimulation.

After 6 h of stimulation with aPHF6, the autophagy gene marker Beclin-1 (BECN1) was observed to be significantly increased, whereas, after 24 h of stimulation, no alteration in any group was seen (Figure 7a,d). Similar results were measured for Beclin-1 protein levels, which were only significantly elevated after 6 h aPHF6 stimulation (Figure 8b,h). Measurement of additional autophagy markers, such as p62 and LC3I/II, underlined the activation after 6 h (Figure 8d–f). Note that, in contrast to LC3I, the protein levels of LC3II were unchanged at this time point (Figure 8f). No changes were detected after 24 h of aPHF6 stimulation (Figure 8k–m). Further, to verify the autophagy effect, CYTO-ID^® 2.0 based autophagic vesicle detection assay was performed, which depicted an almost 1.5 times increase in fluorescence (autophagy vesicle induction) after 6 h of aPHF6 stimulation, as quantified by ImageJ compared to the control group (Figure 7b,c). After 24 h of aPHF6 stimulation, however, the autophagy vesicle induction was comparable to the control (Figure 7e,f). The autophagy induction with the 20 μM PHF6 remained insignificant at both time points.

The primary data can be requested by mail from the corresponding author.