Protection of cultured brain endothelial cells from cytokine-induced damage by α-melanocyte stimulating hormone

Materials

All reagents were purchased from Sigma-Aldrich Corporation (subsidiary of Merck KGaA, Darmstadt, Germany) unless otherwise indicated.

Cell culture

Primary rat brain endothelial cells were isolated from three week-old outbred Wistar rats (Harlan Laboratories, Indianapolis, IN, USA) as described in our previous paper (Veszelka et al., 2013). Forebrains were collected in ice-cold sterile phosphate buffered saline (PBS). After meninges were removed the tissue was cut into 1 mm³ pieces by scalpel and digested with enzymes (1 mg/ml collagenase type II, and 15 μg/ml deoxyribonuclease type I; Roche, Basel, Switzerland) in Dulbecco’s modified Eagle medium (DMEM/F12, Gibco; Life Technologies, Carlsbad, CA, USA) at 37 °C for 55 min. Microvessels were separated from myelin rich fraction by centrifugation in 20% BSA–DMEM gradient (1,000×g, 20 min, three times). The collected vessels were further digested with enzymes (1 mg/ml collagenase–dispase, and 15 μg/ml deoxyribonuclease type I; Roche, Basel, Switzerland) in DMEM/F12 at 37 °C for 35 min. Brain microvascular endothelial cell clusters were separated on a 33% continuous Percoll gradient (1,000×g, 10 min), collected, and washed twice in DMEM/F12. Cells were seeded onto Petri dishes (100 mm; Orange Scientific, Braine-l’Alleud, Belgium) coated with collagen type IV and fibronectin (100 μg/ml each). Cultures were maintained in DMEM/F12 supplemented with 15% plasma-derived bovine serum (First Link, Wolverhampton, UK), 5 μg/ml insulin-transferrin-sodium selenite (Pan Biotech, Aidenbach, Germany), 1 ng/ml basic fibroblast growth factor (Roche, Basel, Switzerland), 100 μg/ml heparin, and 5 μg/ml gentamycin. For the first three days of culture cells were grown in medium with 3 μg/ml puromycin to eliminate P-glycoprotein negative contaminating cell types (Perrière et al., 2005). When endothelial cells reached 90% confluency, they were subcultivated for different experiments.

Primary rat brain pericytes were isolated by the same method, except that the second digestion lasted only for 15 min. After isolation cells were plated onto uncoated dishes and they were cultured in low-glucose DMEM medium (Gibco, Life Technologies, Carlsbad, CA, USA) supplemented with 10% FBS (Pan Biotech, Aidenbach, Germany) and gentamycin. No puromycin was applied. Pericytes were used at third passage for experiments.

Primary rat glial cells were isolated from one-day-old Wistar rats. After meninges were removed brain tissue was mechanically dissociated by a syringe equipped with a long needle. The tissue homogenate was filtered through a nylon mesh (40 μm; Millipore, Billerica, MA, USA) to remove large tissue pieces and vessels. Cell clusters in the filtrate were plated onto uncoated 75 cm² flasks (TPP, Trasadingen, Switzerland), cultured until 90% confluency in DMEM containing 10% FBS (Lonza, Basel, Switzerland) and gentamycin. For the BBB co-culture model glial cells were passaged at a cell number of 5 × 10⁴ cells/well for 12-well plates (Corning Costar, Corning, NY, USA) and cultured for two weeks before use. Confluent glia cultures contained 90% of GFAP immunopositive astroglia, and 10 % CD11b immunopositive microglia.

To induce BBB characteristics brain endothelial cells were co-cultured with brain pericytes and glial cells (Nakagawa et al., 2009) using 12-well format tissue culture inserts (Transwell clear, polyester membrane, 0.4 μm pore size, 1.12 cm² surface; Corning Costar, Corning, NY, USA). Rat pericytes (1.5 × 10⁴ cells/cm²) were subcultivated to the bottom side of the insert membranes and brain endothelial cells (7.5 × 10⁴ cells/cm²) were passaged to the upper side of the collagen type IV and fibronectin coated inserts. Both compartments received endothelial culture medium. After two days of co-culture, brain endothelial cells reached confluency and 550 nM hydrocortisone was added to the culture medium to tighten junctions (Walter et al., 2015). Before experiments 250 μM 8-(4-chlorophenylthio)-cAMP and 17.5 μM Ro 20-1724 (Roche, Basel, Switzerland) were added to the endothelial cells for 24 h to tighten junctions and elevate resistance (Deli et al., 2005; Perrière et al., 2005).

Treatments

The neuropeptide α-MSH (Mw. 1664.8 Da) was tested at 1 pM to 1 μM concentrations. The use of cytokine cocktails is common in CNS culture studies (Sochocka, Diniz & Leszek, 2017). We have also used a combination of TNF-α (50 ng/ml) and IL-1β (25 ng/ml) on intestinal epithelial cells in our previous study (Váradi et al., 2017). Based on this study, we have tested a combination of the pro-inflammatory cytokines TNF-α (10–50 ng/ml) and IL-1β (10–25 ng/ml) for 24 h to induce damage in brain endothelial cells. The control group received culture medium only.

Total RNA isolation and RT-PCR

Brain microvessels were isolated from adult rat brains as published earlier (Veszelka et al., 2007). For receptor expression analysis total RNA was isolated from isolated brain microvessels and brain endothelial cells by TRI Reagent (Molecular Research Center, Cincinnati, OH, USA). One microgram of total RNA was transcribed to cDNA by Maxima First Strand cDNA Synthesis Kit (ThermoFisher, Waltham, MA, USA). Specific oligonucleotide primers were designed for the Mc1r gene (XM_006222790). Primer sets were MC1R_fwd 5′-TGCACCTCTTGCTCATCGTT-3′ and MC1R_rvs 5′-ACCTCCTTGAGTGTCATGCG-3′. The predicted length of PCR product was 160 bps. Primers for β-actin were used as internal controls (NM_031144). Primer sets were ACT_fwd 5′-TACTCTGTGTGGATTGGTGGC-3′ and ACT_rvs 5′-GGTGTAAAACGCAG CTCAGTAA-3′. The predicted length of PCR product was 150 bps. PCR was performed with FIREPol DNA polymerase (Solis BioDyne, Tartu, Estonia) in T100 thermal cycler (BioRad, Hercules, CA, USA). After initial heat inactivation (95 °C for 3 min) the following cycling conditions were applied: melting 94 °C for 15 s, annealing 50 °C for 15 s, polimerization 72 °C for 20 s (35 cycles). After a final 5 min extension at 72 °C PCR products were analyzed on 3% MetaPhor agarose gel (Cambrex BioScience, Rockland, ME, USA) and fragments were verified by capillary DNA sequencing.

Measurement of cell viability: real-time cell electronic sensing and MTT assay

Real-time cell electronic sensing is an impedance-based, label-free technique for dynamic monitoring of living adherent cells, including barrier forming epithelial and brain endothelial cells (Kiss et al., 2014; Lénárt et al., 2015). The RTCA-SP instrument (ACEA Biosciences, San Diego, CA, USA) registers the impedance of cells in every 10 min and for each time point cell index is calculated as (Rn − Rb)/15, where Rn is the cell-electrode impedance of the well when it contains cells and Rb is the background impedance of the well with the medium alone. E-plates, special 96-well plates with built in gold electrodes, were coated with collagen type IV and fibronectin for brain endothelial cells at room temperature and dried for 20 min under UV and air-flow. For background measurements culture medium (60 μl) was added to each well, then 50 μl of rat brain endothelial cell suspension was distributed at a cell density of 5 × 10³ cells/well. After cells reached a steady growth phase they were treated with α-MSH and cytokines.

In parallel we also used an endpoint colorimetric cell viability assay (Kiss et al., 2014). The yellow 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) dye is converted by viable cells to purple formazan crystals. Brain endothelial cells were grown in 96-well plates coated with collagen type IV and fibronectin. After treatment of confluent monolayers, cells were incubated with 0.5 mg/ml MTT solution in cell culture medium for 3 h in CO₂ incubator. Formazan crystals were dissolved in dimethyl sulfoxide, and dye concentration was determined by absorbance measurement at 570 nm with a multiwell plate reader (Fluostar Optima; BMG Labtechnologies, Offenburg, Germany).

Measurement of permeability for marker molecules

Penetration of fluorescein isothiocyanate-labeled dextran (FITC-dextran, MW: 4.4 kDa) and Evans blue-labeled albumin (EBA, MW: 67 kDa) across the BBB model was determined in permeability studies (Veszelka et al., 2007). Rat brain endothelial cells co-cultured with pericytes and glia on cell culture inserts were treated with cytokines and/or α-MSH for 1 h. After treatment inserts were transferred to 12-well plates containing 1.5 ml Ringer-HEPES buffer (150 mM NaCl, 2.2 mM CaCl₂, 0.2 mM MgCl₂, 5.2 mM KCl, 6 mM NaHCO₃, 5 mM glucose, and 10 mM HEPES, pH 7.4) in the lower compartments. In the upper compartments culture medium was replaced by 500 μl Ringer-HEPES buffer containing EBA complex (1 mg/ml bovine serum albumin and 167.5 μg/ml Evans blue) and FITC-dextran (100 μg/ml). The plates were kept for 30 min on a rocking platform (100 rpm) at 37 °C in an incubator with 5% CO₂. After the incubation the concentrations of the marker molecules in samples from the upper and lower compartments were determined by a fluorescence multiwell plate reader (Fluostar Optima; BMG Labtechnologies, Offenburg, Germany; excitation wavelength: 485 nm, emission wavelength: 520 nm for FITC-dextran; excitation wavelength: 584 nm, emission wavelength: 680 nm for EBA). The apparent permeability coefficients (P_app) were calculated by the following equation (Váradi et al., 2017): (1) $$P_{\text{app}}=\frac{{\left[C\right]}_B\times V_B}{A\times {\left[C\right]}_A\times t}$$where [C]_B is the concentration of the tracer in the lower (basal) compartment after 1 h, [C]_A is the concentration of the tracer in the upper (apical) compartment at 0 h, V_B is the volume of the basal compartment (1.5 ml) and A is the surface area available for permeability (1.12 cm²).

Immunohistochemistry

Measurement of ROS production

Reactive oxygen species production was detected by chloromethyl-dichloro-dihydro-fluorecein diacetate (DCFDA; Life Technologies, Carlsbad, CA, USA) as we described previously (Veszelka et al., 2013; Lénárt et al., 2015). The indicator penetrates into the cells and interacts with intracellular esterases. After its oxidation fluorescent molecules are produced, and the fluorescent signal is proportional to the produced ROS. Confluent brain endothelial cells were cultured in black 96-well plates with glass bottoms (Corning Costar, Corning, NY, USA). After 1 h treatment cells were incubated with Ringer-HEPES buffer containing 2 μM DCFDA for 1 h at 37 °C. Hydrogen peroxide (100 μM) was used as a reference compound (positive control). Fluorescence was measured at 485 nm excitation and 520 nm emission wavelengths by a Fluostar Optima multiwell plate reader (BMG Labtechnologies, Offenburg, Germany) every 5 min for 1 h. Fluorescence intensities are shown in arbitrary units.

Statistical analysis

Statistical analysis was done by GraphPad Prism 5.0 software (GraphPad Software Inc., La Jolla, CA, USA). Data are presented as mean ± SEM. Comparison of groups was performed using ANOVA and Dunnett or Bonferroni tests. Differences were considered significant at P < 0.05.

Expression of melanocortin-1 receptor on rat brain microvessels and cultured brain endothelial cells

MC1R gene expression was detected on both freshly isolated rat brain microvessels and confluent rat brain endothelial cell cultures with reverse transcription followed by PCR amplification using the specifically designed rat primers (Fig. 1A). This result was strengthened by immunohistochemistry for MC1R. The brain endothelial staining (Fig. 1B) was similar to the apical surface staining of MC1R on Caco-2 cells in our previous work (Váradi et al., 2017). Positive MC1R immunostaining was found on freshly isolated brain microvessels, too (Fig. 1C).

Effect of α-MSH on rat brain endothelial cell metabolism and viability

Treatment of rat brain endothelial cells with α-MSH peptide alone (1 pM to 1 μM) did not have any significant effect on cell kinetics measured by impedance, indicating good cell viability, attachment and barrier integrity. Followed by real-time cell electronic sensing the cells showed no change after α-MSH treatment, the kinetic curves ran similarly to the untreated control samples (Fig. 2A). In addition, the colorimetric MTT assay also did not detect any metabolic change in cells after 24 h treatment with α-MSH as compared to control (Fig. 2B). For further experiments 1 and 10 pM concentrations of α-MSH were selected, which correspond to the physiological range of the neurohormone in the blood (Kovács et al., 2001; Magnoni et al., 2003).

Effect of α-MSH on cytokine treated rat brain endothelial barrier integrity: real-time cell kinetics

To determine the treatment concentrations of TNF-α and IL-1β to induce brain endothelial damage, four different combinations of cytokine concentrations (10 + 10, 10 + 25, 25 + 25, 50 + 25 ng/ml) were tested (Fig. 3A). The cytokine combination in the tested concentration range decreased brain endothelial viability. There were no statistically significant differences between the four cytokine treatment groups. For further experiments the smallest, 10 + 10 ng/ml concentrations of the cytokines were selected. Treatment of rat brain endothelial cells with TNF-α and IL-1β decreased the cell index by more than 50% at 6 h time point which further dropped by 24 h (Fig. 3B). Low concentration of α-MSH (1 pM) significantly protected against the cytokine-induced cell viability decrease (Fig. 3C). There was no statistically significant difference between the two cytokine + α-MSH treatment groups.

Effect of α-MSH on cytokine treated rat brain endothelial barrier integrity: permeability for marker molecules

The integrity of brain endothelial monolayers was tested by its permeability for marker molecules. A low permeability for both FITC-dextran (1.91 ± 0.02 × 10⁻⁶ cm/s) and large biomolecule albumin (0.88 ± 0.13 × 10⁻⁶ cm/s) was measured on the triple co-culture model of the BBB. Cytokine treatment significantly increased the permeability of the co-cultures for both markers (Fig. 4). The lower, 1 pM concentration of α-MSH peptide significantly blocked the barrier opening effect of the cytokines for FITC-dextran and albumin, while 10 pM α-MSH significantly strengthened the cellular barrier for the bigger marker molecule.

Effect of α-MSH on cytokine treated rat brain endothelial barrier integrity: immunostaining for claudin-5 and β-catenin junctional proteins

Rat brain endothelial cells form tight paracellular barrier, which was visualized by the localization of integral membrane TJ protein claudin-5 and adherens junction protein β-catenin (Fig. 5). These junctional proteins appeared in the cell membranes in a continuous, belt-like manner. In the control, untreated group continuous stainings were visible without gaps at the cell border. In the cytokine treated group gaps and fragmented junctional staining (arrowheads, Fig. 5A) and cytoplasmic redistribution of junctional proteins (asterisks, Fig. 5A) were observed. In cells treated with both α-MSH and cytokine the staining of TJ proteins was more similar to the control group. The object number on the immunostained pictures significantly increased due to cytokine treatment as compared to the control and α-MSH treated group (Figs. 5B and 5C).

Effect of α-MSH on cytokine treated rat brain endothelial cells: ROS production

Treatment with TNF-α and IL-1β significantly increased ROS production in brain endothelial cells compared to the basal ROS production of the control group (Fig. 6). Low concentrations of α-MSH peptide alone did not change the basal ROS production, while it could statistically significantly decrease the cytokine-induced ROS production in cultured brain endothelial cells. Hydrogen peroxide treatment was used as a reference inducer of ROS in the assay, as in our previous study (Veszelka et al., 2013).

Effect of α-MSH on cytokine-induced NF-κB nuclear translocation in rat brain endothelial cells

Cytokine treatment induced inflammatory reaction can be measured reliably by the translocation of the p65 subunit of NF-κB transcriptional factor into cell nuclei. In the control group the localization of NF-κB p65 is mostly cytoplasmic (Fig. 7A). Treatment with inflammatory cytokines induced the nuclear translocation of the transcription factor in rat brain endothelial cells (Fig. 7). This NF-κB nuclear translocation, which was also quantified based on the fluorescent intensity of the immunostainings, was inhibited by α-MSH at 1 pM concentration (Fig. 7A), and which was the most effective in other experiments.