Materials

A hypoxia modular incubator chamber (MIC-101) was acquired from Billups-Rothenberg, Inc. (San Diego, CA, USA). We used the following CO2 incubators; New Brunswick Galaxy 170 R Stainless Steel CO2 Incubator and New Brunswick Innova CO-170 CO2 Incubator, both acquired from New Brunswick Scientific (Edison, New Jersey, USA). Phosphate-buffered saline (PBS – MS01P01003) was purchased from Biowest (Nuaillé, France). Fluoro-Jade C Dry Powder (TR-160-FJC) was obtained from Biosensis (Thebarton, Australia). JC-1 (#30001) was obtained from Biotium, Inc. (Hayward, CA, USA). B27 (10889–038), Neurobasal medium (with glucose: 12348–017; without glucose: 12015621), and DMEM (with glucose: 31053–028; without glucose: A1443001) were obtained from Gibco (Grand Island, NY, USA). The oxygen analyser (GOX100 Cat. No. 600437) was from Greisinger (Regenstauf, Germany). The Bradford assay (5000006) was obtained from Bio-Rad Laboratories (Munchen, Germany). ELISA kits for IL-1β (E0143Hu), IL-10 (E0102Hu), PPARγ (E1511Hu), PGC1α (E3509Hu), and BCL2 (BPE040) were purchased from Bioassay Technology Laboratory (Shanghai, China). The culture plates (6-well: 92006, 24-well: 92024, 96-wells: 92096) were obtained from TPP Techno Plastic Products AG (Trasadingen, Switzerland), and 75 cm2 U-shaped canted neck cell culture flasks (353136) with a vent cap was obtained from Corning (New York, USA). HMC3 cells (CRL-3304™, Lot number: 70046457), EMEM (30–2003), FBS (30–2020), and trypsin/EDTA (30–2101) were purchased from American Type Culture Collection (ATCC—Virginia, USA). Fast Probe qPCR Master Mix (E0422-03) was from EurX (Gdansk, Poland). The Cytotoxicity Detection Kit (11644793001) was purchased from Roche Diagnostics GmbH (Mannheim, Germany). The primary anti-IBA1 (sc-32725) antibody was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Amorfrutin B (SMB00532), GW9662 (M6191), L-glutamine (G-8540), FBS (F7524), DMSO (D8418), RIPA (89,901) buffer, protease inhibitor cocktail for mammalian tissues (P8340), penicillin–streptomycin antibiotics (P4333), caspase-1 substrate (SCP0070), MTT (M5655), BrdU cell proliferation assay (QIA58), calcein AM (C1359), and poly-L-ornithine hydrobromide (P4538) were obtained from Sigma‒Aldrich (St. Louis, MO, USA). The RNeasy Mini Kit (74106) was obtained from Qiagen (Hilden, Germany). Alexa Fluor 647 IgG (A21235), a high-capacity cDNA reverse transcription kit (436881), alamarBlue™ Cell Viability Reagent (DAL1101), Hoechst 33342 (62249), and TaqMan probes (4331182) for specific genes encoding ACTB, GAPDH, HPRT, IL1B, IL10, TNFA, PPARG, PGC1A, and BCL2 were obtained from Thermo Fisher Scientific (Waltham, MA, USA).

Human Microglial HMC3 Cell Line

For our experiments, the HMC3 cell line was obtained from American Type Culture Collection (ATCC, VA, USA; CRL-3304™, Lot number: 70046457), which is responsible for its distribution and confirmation of authenticity through a comprehensive identification procedure. Although rodent microglia has been extensively studied, the use of human microglia has been limited until human immortalized microglial cell lines have been developed and incorporated in ATCC. HMC3 cells differ from the primary microglia, mainly due to SV40-dependent immortalization of human embryonic microglial cells. According to Rai et al. (2020) HMC3 express microglia-specific and also typical myeloid markers with the exception of low or absent CD45 and CD11b expression. However, the cells have been characterized as responding to a pattern of chemokines and inflammatory stimuli and also regulating the expression of activated microglia markers (Dello Russo et al. 2018; Baek et al. 2021). In accordance with the manufacturer’s instructions, cells were seeded in T-75 flasks and cultivated in 15 ml of complete growth medium consisting of EMEM supplemented with FBS (10%). Trypsinization with 0.25% (w/v) trypsin and 0.53 mM EDTA solution was performed after the cells reached 90% confluence. For the purpose of obtaining a homogenous population for the experiments, cells were subcultured twice, analyzed to exclude the presence of mycoplasma and then frozen in aliquots at passage 3 (p3). After thawing, the cells were moved from cryovial to T-75 flasks and suspended in EMEM supplemented with 10% FBS (Fig. 1a). Due to experimental plans requiring two types of the same media, differing only in the presence or absence of glucose, the next day it was necessary to replace the culture EMEM medium with DMEM, which is commercially available with and without glucose and has a composition similar to EMEM. The day before experiment, the cells were seeded on multiwell dishes (2 × 104 viable cells per cm2) with reduced FBS concentration (5%) and allowed to recover overnight. The concentration of FBS was reduced in order to achieve more stable experimental conditions by controlling the proliferation rate, which is crucial for obtaining repeatable results.

Fig. 1

The timeline showing experimental conditions from culturing to the end of the experiment (a). Experimental models of hypoxia/ischemia and amorfrutin B post-treatment in human microglia cell line HMC3 (b and c) and primary neuronal cell cultures (d)

Primary neocortical neurons

Primary cultures of cortical neurons were obtained from mouse embryos at 15 days of gestation as previously described (Kajta et al. 2004). All animals were maintained according to the 3Rs principles and in compliance with European Union Legislation approved by the Bioethics Commission. Cortices from Swiss CD1 mouse embryos were dissected and digested with trypsin. After centrifugation, the isolated cells were suspended in Neurobasal medium supplemented with 10% FBS and placed in poly-L-ornithine-coated multiwell dishes at a density of ~ 2.0 × 105 viable cells/cm2. After 3 days, the culture medium was exchanged for supplemented medium without FBS, and the cells were cultivated in a humidified atmosphere (37°C with 5% (vol/vol) CO2) for the next 4 days prior to the experiment. The percentage of astrocytes (identified with GFAP) did not exceed 10%, as previously described (Kajta et al. 2004).

Experimental Models of Hypoxia/IschemiaHypoxia of HMC3 Cells and Primary Neurons

To simulate hypoxic conditions and obtain O2 levels close to zero, HMC3 cells and primary neurons were placed in a prewarmed and humidified hypoxia modular incubator chamber with 95% N2/5% CO2, as previously described (Wnuk et al. 2021a, Wnuk, Przepiórska et al. 2021). The O2 level in the chamber reached less than 0.5%, as measured with an oxygen analyzer. As for HMC3 cells, the experimental model consisted of the following steps: i. 6 h of hypoxia with a reduced concentration of FBS (1%) in DMEM (4.5 g/l glucose), ii. 18 h of reoxygenation with the restored concentration of FBS (5%) in the same DMEM (6 + 18 paradigm; Fig. 1b). To capture hard to observe effects related to caspase-1 activity, we decided to shorten the chosen paradigm to 2 h of hypoxia and 5 h of reoxygenation, but only in this one case (2 + 5 paradigm; Fig. 1c). For primary neuronal cell cultures, the experimental model consisted of the following steps: i. 6 h of hypoxia in neurobasal medium (4.5 g/l glucose), ii. 18 h of reoxygenation with replaced neurobasal medium (Fig. 1d).

Ischemia of HMC3 Cells and Primary Neurons

To mimic ischemic conditions and deprive cells of O2 and glucose, HMC3 cells and primary neurons were placed in a prewarmed and humidified hypoxia modular incubator chamber with 95% N2/5% CO2 in glucose-free medium. As for HMC3 the following experimental procedure was used: i. 6 h of ischemia with a reduced concentration of FBS (1%) in DMEM without glucose (0 g/l), ii. 18 h of reoxygenation with the restored concentration of FBS (5%) in DMEM supplemented with 4.5 g/l glucose (6 + 18 paradigm; Fig. 1b). Only in the case of caspase-1 activity assessment, we decided to shorten the chosen paradigm to 2 h of ischemia and 5 h of reoxygenation to capture hard to observe effects (2 + 5 paradigm; Fig. 1c). For primary neuronal cultures, the experimental model consisted of the following steps: i. 6 h of ischemia with neurobasal medium without glucose (0 g/l), ii. 18 h of reoxygenation with the medium replaced with standard neurobasal containing 4.5 g/l glucose (Fig. 1d). The control group included cultures that were not subjected to hypoxia or ischemia but were exposed to changes in FBS concentration analogous to those in the other groups (normoxic group).

Post-treatment with Amorfrutin BTreatment of HMC3 Cells Subjected to Hypoxia/Ischemia

In our study, we applied the post-treatment paradigm based on the application of amorfrutin B at the beginning of the reoxygenation period. This paradigm better reflects clinical standards due to the narrow therapeutic window of currently available therapies (up to 4.5 h). Taking into account our previous and present studies, we decided to apply 1 and 5 µM amorfrutin B, which induces neuroprotective effects in neurons subjected to hypoxia/ischemia. We wanted to observe specific effects on microglia to compare and follow-up effects obtained in our previous and present studies. Amorfrutin B was dissolved in DMSO and then in DMEM, resulting in DMSO concentrations less than 0.1%. After the experiments, the biological material was collected and used for further biochemical and molecular analyses.

Treatment of Primary Neurons Exposed to Hypoxia/Ischemia

To complement and compare the effects obtained in HMC3 cells, we decided to conduct additional experiments supporting our previous studies. Both concentration-dependent experiments (0.1, 1, 5, and 10 µM) and experiments with the most effective concentrations of amorfrutin B (1 and 5 µM) were performed. In pursuit of better clinical translation, the post-treatment paradigm was applied, i.e., amorfrutin B was added to the cell cultures at the beginning of the reoxygenation period for the next 18 h until the end of the experiments. Treated neurons were maintained in a humidified incubator (37 °C with 5% (vol/vol) CO2), and after 24 h of experiment, the biological material was collected for further biochemical or molecular analyses. After amorfrutin B was dissolved in DMSO, it was added to the neurobasal medium. The concentration of DMSO was less than 0.1%.

Treatment with a PPARγ Antagonist (GW9662)

The involvement of the PPARγ receptor in amorfrutin B-induced neuroprotection was verified with the receptor antagonist GW9662 (1 µM). GW9662 was added to the culture medium after 6 h of hypoxia/ischemia in a post-treatment paradigm. After 40 min, amorfrutin B (1 and 5 µM) was added to the cell cultures for the next 18 h of reoxygenation. Amorfrutin B and GW9662 were used at concentrations that did not affect the level of cell neurodegeneration, as determined by Fluoro-Jade C; moreover, we applied both effective concentrations of amorfrutin B (1 and 5 µM) to avoid nonspecific effects. Both compounds were dissolved in DMSO and subsequently in culture medium, leading to a DMSO concentration less than 0.1%.

Measurement of Cell Viability with AlamarBlue™ Reagent

After 6 h of hypoxia/ischemia and 18 h of amorfrutin B treatment, we diluted 10 × alamarBlue™ in cell culture medium and added it to the neuronal cultures. The cells were incubated with the solution for 3 h at 37 °C, and the absorbance was assessed at a wavelength of 570 nm (using 600 nm as a reference) with an Infinite M200 PRO microplate reader (Tecan, Mannedorf, Switzerland) and i-control software. The presence of viable cells induces a reduction in the color of blue resazurin to pink resorufin, and the absorbance level is proportional to the number of living cells (Lescat et al. 2019). All the results were compared with the absorbance levels of vehicle-treated cells and are presented as a percentage of the control ± SEM.

Assessment of Neurodegeneration with Fluoro-Jade C

To assess the level of degenerating neurons in response to hypoxia/ischemia and amorfrutin B post-treatment, we decided to use Fluoro-Jade C (Wnuk et al. 2021a, Wnuk, Przepiórska et al. 2021; Pietrzak-Wawrzyńska et al. 2023). This fluorochrome dye was diluted in distilled water, resulting in a 0.005% working solution. After the experiment, 100 µl of culture medium was replaced with Fluoro-Jade C reagent per well, and the cells were incubated for 1 h. Subsequently, the fluorescence was measured at an excitation wavelength = 490 nm and an emission wavelength = 525 nm using an Infinite M200 PRO microplate reader, and i-control software. The fluorescence data were compared to those of the vehicle-treated cells and are presented as a percentage of the control ± SEM. Fluoro-Jade C is a green fluorescent dye, and the fluorescence intensity is proportional to the number of degenerating neurons.

Immunofluorescence Staining with IBA1

Immunofluorescence staining with IBA1 and confocal microscopy were used to detect and visualize microglia, which were cultured on glass coverslips. After 6 h of hypoxia/ischemia and 18 h of amorfrutin B treatment, the cultures were fixed with 4% paraformaldehyde. After 3 washes, the cells were incubated for 1 h with blocking buffer containing 5% normal donkey serum and 0.3% Triton X-100 in 0.01 M PBS. Then, the microglial cells were treated with an anti-IBA1 primary antibody (diluted 1:100) and incubated at 4 °C for 24 h. The last step of the assay involved incubating the cells with a secondary antibody, after which the slides were washed, mounted, and coverslipped. The negative control with blocking buffer depleted of primary antibody excluded the nonspecific binding of the secondary antibody to the sample components and confirmed the specific binding of the IBA1 to its target antigen. The microscopic preparations were analyzed with a Leica TCS SP8 WLL confocal laser scanning microscope (DMi8-CS, Leica Microsystems, Wetzlar, Germany), and the pixel intensity was assessed using ImageJ software (FIJI version; 1.54f). The frequency of the brightest pixels in the region of interest was quantified by determining the mean fluorescence intensity excluding the background. The immunofluorescence signal corresponded to the IBA1 expression level.

Estimation of Caspase-1 Activity

Caspase-1 activation was assessed in microglial cells exposed to hypoxia/ischemia and amorfrutin B post-treatment. Cultures were lysed with buffer containing DTT, and the cell lysates were incubated with caspase-1 substrate (Ac-Trp-Val-Ala-Asp-pNA; Sigma‒Aldrich, USA) at 37 °C. According to the manufacturer, this colorimetric substrate cleaves specifically at aspartic acid in the presence of caspase-1, and the level of released p-nitroaniline was determined by measuring the absorbance at 405 nm. After 1 h of incubation, the level of p-nitroaniline was measured with an Infinite M200 PRO microplate reader and i-control software. The results were compared to the absorbance of the vehicle-treated cells (normoxic conditions) and are presented as the percentage of the control ± SEM.

Assessment of the Mitochondrial Membrane Potential with JC-1

To determine the mitochondrial membrane potential in microglial cells subjected to hypoxia/ischemia and amorfrutin B post-treatment, the JC-1 was added to the plate and incubated for 1 h at 37 °C, as previously described (Rzemieniec et al. 2015; Wnuk et al. 2020). This lipophilic dye enters mitochondria, and in healthy cells, forms aggregates, which emit red fluorescence (an indicator of increased mitochondrial membrane potential). The JC-1 solution was replaced with PBS, and both red (540 nm/590 nm) and green (490 nm/525 nm) fluorescence intensities were measured with an Infinite M200 PRO microplate reader. The data are expressed as the red to green fluorescence ratio and are presented as a percentage of the control ± SEM.

Fluorescence Staining with Calcein AM and Hoechst 33342

Microglial cell cultures were subjected to hypoxia/ischemia and amorfrutin B post-treatment. After the experiment, the cells were incubated with Hoechst 33342 (2 µg/ml solution) at room temperature for 5 min. Then, the cultures were incubated with 2 µM calcein AM reagent at room temperature for 10 min, as previously described (Kajta et al. 2019). The cells were visualized with a Leica DM IL LED Inverted Microscope (Leica Microsystems, Wetzlar, Germany) using constant illumination settings. The living cells presented green cytoplasm (calcein AM), while blue fluorescence staining was specific for the nuclei of microglial cells (Hoechst 33342).

Microglial Morphometric Analysis

The total number of cells was estimated basing on the number of cell nuclei (Hoechst 33342), while morphometric analysis was performed basing on staining of living cells (calcein AM). The semi-automated analysis with ImageJ (Java-based FIJI version; 1.54f) was performed to assess: the ramification index, cell body area, minimum Feret diameter, and microglia cell number. The basic step of analysis involved converting the grayscale picture into a binary image with setting the threshold value. Segmenting the image and separating the foreground from the background constituted a crucial action in determining cell number and morphological status. Particle analysis provided valuable insights into the total number of particles in image (microglial cell counting), area occupied by each particle (cell body surface), and closest possible distance between the two parallel tangents of an object (minimum Feret diameter). The ramification index was determined by fractal analysis as presented in previous studies (Becker et al. 2018; Kogel et al. 2021; Wittekindt et al. 2022). It was calculated in three steps: i. estimation of the cell area (Ac), ii. using convex hull algorithm to measure projection area (Ap), iii. determining the ratio of Ac to Ap and interpretation of obtained results. Quiescent microglia exhibit small bodies and ramified processes (small Ac and large Ap), while activated microglia is characterized by hypertrophy of soma and retracted processes (similar Ac and Ap). Consequently, the ramification index of activated microglia approaches the value close to 1. An increased ramification index indicates morphological changes and heightened reactivity of microglia. The results of all analyses are presented as mean or as the percentage of the control ± SEM.

Measurement of MTT Reduction

A colorimetric MTT assay was used to evaluate the mitochondrial function and metabolic activity of microglial cells subjected to hypoxia/ischemia and amorfrutin B treatment. The cell cultures were treated with MTT solution, and after 1 h of incubation at 37 °C, the reagent was replaced with 100% DMSO, which was used to dissolve the formazan crystals. The MTT assay is based on reducing MTT by oxidoreductase enzymes to purple formazan, and the intensity of the purple color is proportional to the metabolic activity of the cells. The absorbance at 570 nm was measured with an Infinite M200 PRO microplate reader, and the data were analyzed with i-control software and are presented as a percentage of the control ± SEM.

Determination of LDH Release

A Cytotoxicity Detection Kit was used to assess the lactate dehydrogenase level in cell-free supernatants collected immediately after the experiment. The supernatants were incubated with the appropriate reaction mixture for 30–60 min, resulting in a colorimetric reaction, as previously described (Kajta et al. 2001). The intensity of the red color was measured using an Infinite M200 PRO microplate reader at a wavelength of 490 nm and was proportional to the extent of plasma membrane damage in response to hypoxia/ischemia. The results obtained from the experiments were analyzed with an i-control and are presented as a percentage of the control ± SEM.

Assessment of the Proliferation Potential with BrdU

To quantify the incorporation of BrdU into newly synthesized DNA from active proliferating microglial cells, a BrdU cell proliferation assay was applied after hypoxia/ischemia and amorfrutin B post-treatment, according to the manufacturer’s protocol. The cell medium was complemented with 20 µl of working stock of BrdU per well. After 2 h of incubation at 37°C, the reagent was removed, and the cells were reconstituted with fixative/denaturing solution. The next step involved incubating the cells with an anti-BrdU antibody (1:100 dilution), followed by exposure to a peroxidase-conjugated goat anti-mouse IgG HRP. The cells were washed 3 times with buffer and subjected to substrate solution, followed by stop solution. The absorbance of each well was measured at 450 nm with an Infinite M200 PRO microplate reader, and the data were analyzed with i-control software. The results are presented as a percentage of the control ± SEM.

qPCR Analysis of mRNAs Specific to the Genes Encoding IL1B, IL10, TNFA, PPARG, PGC1A and BCL2

Microglial cells were subjected to hypoxic/ischemic conditions and amorfrutin B post-treatment, after which total RNA was collected and isolated using an RNeasy Mini Kit (Qiagen, Hilden, Germany) as previously described (Wnuk et al. 2016, 2018). After spectrophotometric determination of the RNA content in the sample, reverse transcription, and quantitative polymerase chain reaction (qPCR) were performed using a CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). The RNA (1000 ng) was reverse-transcribed to the cDNA template in a final volume of 50 µl using a high-capacity cDNA reverse transcription kit. The obtained products were amplified using TaqMan Gene Expression Assays specific for genes encoding IL1B, IL10, TNFA, PPARG, PGC1A, and BCL2. The entire volume of the reaction mixture contained 10 μl of Fast Probe qPCR Master Mix, 8 μl of RNase-free water, 1 µl of cDNA, and 1 μl of TaqMan Gene Expression Assay. The qPCR process consisted of series of temperature changes: 2 min at 50 °C and 10 min at 95 °C, followed by 55 cycles of 15 s at 95 °C and 1 min at 60 °C. The mRNA expression level was detected in the 40th (last) cycle; however, due to the small amount of material, we decided to increase the number of cycles from 40 to 55. The data were analyzed using the Ct for each sample and the delta Ct method. The reference gene was chosen from among GAPDH, HPRT and ACTB, while geNorm, NormFinder, and BestKeeper pointed to ACTB as the most stable reference gene. The obtained results are presented as fold changes ± SEM.

ELISAs for IL-1β, IL-10, PPARγ, PGC1α, and BCL2

After 24 h of experiment, the microglial cells were gently washed with ice-cold PBS and lysed with ice-cold RIPA lysis buffer supplemented with a protease inhibitor cocktail. The collected samples were sonicated and centrifuged (15,000 × g for 20 min at 4 °C), and the obtained supernatants were collected as previously described (Kajta et al. 2017; Rzemieniec et al. 2019). The protein concentration was estimated using the Bradford method, and bovine serum albumin was used as a standard. Both standards and samples of known concentrations (~ 2 µg/µl) were added to plates precoated with IL-1β, IL-10, PPARγ, PGC1α, and BCL2. Next, biotin-labeled detection antibodies and streptavidin-HRP were added to each well. All the wells were washed with buffer, and the reaction was completed by adding substrate solutions. The color of the reaction developed according to the concentration of the protein of interest, and the reaction was terminated by the addition of acidic stop solution. The absorbance was measured at 450 nm with an Infinite M200 PRO microplate reader, and the data are presented as a percentage of the control value ± SEM or pg/mg of total protein.

Data Analysis

All the results were obtained as the absorbance or fluorescence units per well for alamarBlue™, caspase-1, MTT, LDH, BrdU or Fluoro-Jade C, JC-1, fluorescence units for qPCR; and pg per mg of total protein for the ELISAs. The statistical analysis of the data and the estimation of significance were conducted with analysis of variance (ANOVA), and the post hoc Newman‒Keuls test. A p value less than 0.05 was considered to indicate statistical significance and is presented as follows: *p < 0.05, **p < 0.01, and ***p < 0.001 (compared to the control groups); #p < 0.05, ##p < 0.01, and ###p < 0.001 (compared to the hypoxic group); ^p < 0.05, ^^p < 0.01, and ^^^p < 0.001 (compared to the ischemic group); and $$$p < 0.001 (compared to the cells subjected to both hypoxia/ischemia and amorfrutin B post-treatment. The effects of amorfrutin B or GW9662 on selected basic parameters related to cell survival under normoxic conditions are presented in the supplementary information (Figs. S1 and S2, Tables S1 and S2). Differences in the effects of the hypoxic and ischemic models and effects of amorfrutin B action between the hypoxic and ischemic conditions are also presented as the supplementary material (Figs. S3–S10).