Nitrative Stress in Cerebral Endothelium is Mediated by mGluR5 in Hyperhomocysteinemia

Abstract

Hyperhomocysteinemia (HHcy) disrupts nitric oxide (NO) signaling and increases nitrative stress in cerebral microvascular endothelial cells (CMVECs). This is mediated, in part, by protein nitrotyrosinylation (3-nitrotyrosine; 3-NT) though the mechanisms by which extracellular homocysteine (Hcy) generates intracellular 3-NT are unknown. Using a murine model of mild HHcy (cbs^+/– mouse), we show that 3-NT is significantly elevated in cerebral microvessels with concomitant reductions in serum NO bioavailability as compared with wild-type littermate controls (cbs^+/+). Directed pharmacology identified a receptor-dependent mechanism for 3-NT formation in CMVECs. Homocysteine increased expression of inducible NO synthase (iNOS) and formation of 3-NT, both of which were blocked by inhibition of metabotropic glutamate receptor-5 (mGluR5) with the specific antagonist 2-methyl-6-(phenylethynyl) pyridine hydrochloride. Activation of mGluR5 is both sufficient and necessary to drive the nitrative stress because direct activation using the mGluR5-specific agonist (RS)-2-chloro-5-hydroxyphenylglycine also increased iNOS expression and 3-NT formation while knockdown of mGluR5 receptor expression by short hairpin RNA (shRNA) blocked their increase in response to Hcy. Nitric oxide derived from iNOS was required for Hcy-mediated formation of 3-NT because the effect was blocked by 1400 W. These results provide the first evidence for a receptor-dependent process that explains how plasma Hcy levels control intracellular nitrative stress in cerebral microvascular endothelium.

Keywords

blood—brain barrier homocysteine microvascular oxidative stress peroxynitrite

Introduction

Homocysteine (Hcy) is an aminothiol formed during methionine metabolism. Its elevation in plasma (hyperhomocysteinemia; HHcy) is an independent risk factor for stroke, silent brain infarcts, white-matter lesions, Alzheimer's disease (AD), and dementia. Damage to cerebral microvascular endothelium in HHcy leads to opening of the blood—brain barrier in rodents (Beard et al, 2011; Kamath et al, 2006) and in humans (Lehmann et al, 2003) while increasing the risk of stroke (Eikelboom et al, 2000). Cerebral small vessel disease is directly related to the level of Hcy as a consequence of endothelial dysfunction (Hassan et al, 2004).

Homocysteine is metabolized by remethylation in the methionine cycle or by transsulfuration, a pathway comprising two enzymes, cystathionine β-synthase and cystathionine γ-lyase. Current therapies for elevated Hcy are limited to vitamin supplements, which serve as cofactors in the pathways of Hcy metabolism (remethylation and transsulfuration), but these therapies have shown low efficacy in altering disease risk or progression. Mild-to-moderate HHcy (mild: 15 to 30 μmol/L and moderate: 30 to 100 μmol/L) occurs in 5% to 10% of the general population and may exceed 90% in hemodialysis patients (Carmel and Jacobsen, 2001). Without a better understanding of the mechanisms by which Hcy alters cerebral endothelial function, therapeutic options will remain limited.

Production of superoxide, a potent reactive oxygen species, contributes significantly to cerebrovascular dysfunction in HHcy (Dayal et al, 2004). Similarly, disrupted nitric oxide (NO) signaling is a commonly reported outcome of HHcy and a significant player in neurovascular diseases (Faraci, 2011). Nitrotyrosinylation of proteins occurs when superoxide quenches NO to form peroxynitrite (ONOO⁻), which binds tyrosine residues in proteins to produce 3-nitrotyrosine (3-NT). Widespread regulation of microvascular function is manifest through 3-NT modification of proteins involved in cell survival (e.g., caspases) and matrix architecture (e.g., metalloproteinases) as well as in reduced protection from the superoxide itself (by inhibiting superoxide dismutase activity).

One of the fundamental gaps in knowledge regarding the actions of Hcy is the identity of a ‘sensor’ or receptor by which the extracellular molecule drives intracellular oxidative/nitrative stress. A survey of the literature identified glutamate receptors as one target of Hcy in neurons (Boldyrev and Johnson, 2007; Shi et al, 2003; Zieminska and Lazarewicz, 2006). Of note, metabotropic glutamate receptor-5 (mGluR5) is a Gα_q/11-coupled receptor that is particularly sensitive to several acidic derivatives of Hcy (though Hcy was not tested directly) (Shi et al, 2003). The downstream effector of mGluR5, protein kinase C, is a well-known activator of inducible NOS (iNOS; Fiebich et al, 1998), while the expression of iNOS is a major mechanism for 3-NT formation (Drew and Leeuwenburgh, 2002). Here, we first tested the hypothesis that cerebral microvessels are under nitrative stress in a murine model of HHcy. We then dissected the pathway by which Hcy induces nitrative stress in cerebral microvascular endothelial cells (CMVECs).

Materials and methods

All procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Idaho State University and performed in accordance with the NIH (National Institutes of Health) Guide for the Care and Use of Laboratory Animals.

In-Vivo Murine Model of Hyperhomocysteinemia

Cbs heterozygous knockout mice (cbs^+/–) on standard chow are a model of mild HHcy that matches the etiology in many humans (Beard and Bearden, 2011). Homozygous knockouts develop severe HHcy and most die by 5 weeks of age; this is the reason why they were not used in this study. To examine the effect of mild HHcy on cerebral microvessels, we used cbs^+/– mice on standard chow and compared them with wild-type littermates (cbs^+/+) on the same diet (n = 24; mean ± standard deviation: 27.5 ± 3.9 g, 5.7 ± 0.5 months of age). This age was chosen because it is ∼25% of the lifespan of these mice, making the subject animals young adults. We maintain an in-house breeding colony descended from founders obtained by our laboratory from Dr Steven Lentz (University of Iowa) in 2008. These mice are on a C57B1/6 strain background, backcrossed at least 12 generations, and maintained on a 12/12 hours dark/light cycle with food and water ad libitum.

Quantification of Homocysteine Concentration

Blood was collected into serum separator tubes (Becton Dickinson, Franklin Lakes, NJ, USA) at the time of euthanasia of each mouse. Total serum Hcy concentration was measured using a commercial enzyme immunoassay kit, Homocysteine-EIA Kit (Bio-Rad, Hercules, CA, USA), according to the manufacturer's instructions.

Brain Microvascular Isolation and Depletion from Hyperhomocysteinemic Mice

Brains were harvested from cbs^+/– and cbs^+/+ mice (n = 12/group) and the microvessels were isolated by a modification of the method of Gerhart et al (1988). Briefly, three brains were collected, pooled, and homogenized on ice in 1 mL of cold stock buffer (low glucose DMEM (Dulbecco's Modified Eagle's Medium; Sigma, St. Louis, MO, USA) plus 25 mmol/L HEPES and 1% dextran, pH 7.4) in a glass tissue grinder using a teflon pestle. The homogenate was filtered through a series of nylon mesh membranes (300 μm, then twice through 100 μm; Spectrum, Houston, TX, USA), mixed with an equal volume of 40% dextran in stock buffer, and centrifuged at 3,500 g for 30 minutes at 4°C. The supernatant with the lipid layer was removed and the pellet was resuspended in stock buffer. The suspension was passed through a 25-μm nylon mesh membrane (Bio-Design, Carmel, NY, USA). The microvessels on the surface of the membrane were washed with stock buffer four times, collected from the membrane, and then centrifuged at 3,000 g for 15 minutes at 4°C, the supernatant was removed, and the microvessel pellets resuspended in incubation buffer. These are identified throughout this manuscript as the ‘microvascular fraction‗. The purity and quantity of each preparation were routinely checked by light microscopy. Supernatants were pooled and labeled as the ‘vessel-depleted fraction’. Each sample n = 1' represents brains pooled from three mice of the respective groups.

3-Nitrotyrosine and 4-Hydroxynonenal Formation in Brain Fractions from Hyperhomocysteinemic Mice

For 3-NT and 4-hydroxynonenal (4-HNE) immunodetection, the proteins were electrophoresed onto 4% to 12% Bis-Tris reducing gels (NuPAGE Novex, Grand Island, NY, USA). After transfer, the nitrocellulose membranes were probed with an anti-HNE antibody (HNE11-S; Alpha Diagnostics, San Antonio, TX, USA) followed by a secondary antibody. The polypeptides were visualized by chemiluminescence with substrate (Pierce, Rockford, IL, USA). Optical densities of the immunopositive bands were quantified with Image J analysis software (NIH, USA). The 3-NT-modified proteins were detected with an anti-nitrotyrosine antibody (3-NT, AB5411; Chemicon-Millipore, Billerica, MA, USA) and quantified as described for HNE. The secondary horseradish peroxidase-linked goat anti-rabbit IgG was from Santa Cruz Biotechnology, (Santa Cruz, CA, USA). The membranes were reprobed with a β-actin antibody (Sigma) to validate loading uniformity and normalize values.

Glutathione Measurements in Isolated Cerebral Microvessels from Hyperhomocysteinemic Mice

Glutathione (GSH) concentrations in the brain were determined by reverse phase high-performance liquid chromatography (HPLC) by the method of Winters et al (1995). The HPLC system (Thermo Scientific Corporation, Rockford, IL, USA) consisted of a Finnigan Spectra SYSTEM vacuum membrane degasser (model SCM1000), a gradient pump (model P2000), an autosampler (model AS3000), and a fluorescence detector (model FL3000) with λ_ex = 330 nm and λ_em = 376 nm. The HPLC column used was a Reliasil ODS-1 C₁₈ column (5 μm packing material) with 250 mm × 4.6 mm i.d. (Column Engineering, Ontario, CA, USA). The mobile phase (70% acetonitrile and 30% water) was adjusted to a pH of 2.0 through the addition of 1 mL/L of both acetic and o-phosphoric acids (Fisher Scientific, Fair Lawn, NJ, USA). Brains were homogenized in a serine-borate buffer (100 mmol/L Tris-HCl, 10 mmol/L boric acid, 5 mmol/L l-serine, 1 mmol/L DETAPAC (diethylenetriaminepentaacetic acid), pH 7.4) on ice. Homogenates were derivatized with 1.0 mmol/L NPM [N-(1-pyrenyl)-maleimide] (Sigma) in acetonitrile. Briefly, HPLC grade water was added to each sample to make a volume of 250 μL, and 750 μL of NPM (1 mmol/L in acetonitrile) was added. This mixture was incubated for 5 minutes at room temperature and the reaction was stopped by adding 10 μL of 2N HCl. The samples were then filtered through a 0.45-μm acrodisc filter (Advantec MFS, Dublin, CA, USA) and injected into the HPLC system. The NPM derivatives were eluted from the column isocratically at a flow rate of 1 mL/min (Ridnour et al, 1999; Winters et al, 1995).

Plasma Nitric Oxide (NOx; Nitrate + Nitrite) Measurements in Hyperhomocysteinemic Mice

Blood was drawn by cardiac puncture of isoflurane-anesthetized mice (cbs^+/+ and cbs^+/–) and placed in EDTA blood collection tubes (BD). Blood was centrifuged at 5,000 g for 10 minutes at 20°C and plasma supernatant was collected. The plasma was filtered through a 30-kDa molecular weight cutoff microfuge ultrafiltration device (Amicon, Millipore, Billerica, MA, USA) and filtrate was used for analysis. Total plasma nitrate and nitrite (NO_x) was measured using a nitrate/nitrite colorimetric assay kit (780001; Cayman Chemical, Ann Arbor, MI, USA) according to the manufacturer's instructions. After reaction procedures, absorbance at 540 nm was measured in samples and standards using a microplate reader (Synergy HT, BioTek, Winooski, VT, USA), and results were expressed as NO_x in μmol.

Brain Microvascular Endothelial Cell Culture

An immortalized line of mouse CMVECs (bEnd.3; Montesano et al, 1990) was purchased from ATCC (Manassas, VA, USA). The CMVECs were cultured at 37°C in a humidified incubator with 5% CO₂ and balance room air in DMEM high glucose (DMEM-H; Invitrogen, Grand Island, NY, USA), 10% bovine calf serum (Hyclone FetalClone III; Fisher Scientific), and 1 × gentamicin or penicillin/streptomycin and amphotericin B (Invitrogen) in plastic culture flasks. All experiments were performed with cells from the third to sixth passages. For conditioned media experiments, RAW 264.7 cells (ATCC) were grown to ∼70% confluence in the same media and treated with 1 ng/mL lipopolysaccharide to induce secretion of cytokines that stimulate oxidative/nitrative stress in endothelial cells; after 12 hours, the conditioned media were harvested and used for treating CMVECs.

MTT Assay

Cerebral microvascular endothelial cell viability was determined by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay. The CMVECs were treated after 3 days at confluence with various concentrations of Hcy from 0 to 1,600 μmol/L (n = 8). After 48 hours of Hcy treatment, MTT was dissolved in media (0.5 mg/mL) and added to each well of cells, and the plate was incubated for 2 hours at 37°C. The medium containing MTT was removed, cells were rinsed with phosphate-buffered saline (PBS), and dimethyl sulfoxide (DMSO) was added. Spectrophotometric absorbance of each sample was measured at 560 nm using a microplate reader (Synergy HT, BioTek).

Brain Microvascular Endothelial Cell Treatments

Cerebral microvascular endothelial cells were treated after 3 days at confluence. d,l-Homocysteine (Sigma-Aldrich, St Louis, MO, USA) was dissolved in sterile PBS as a stock solution and diluted in culture media on treatment. A selective mGluR5 antagonist, 2-methyl-6-(phenylethynyl) pyridine hydrochloride (MPEP; Tocris, Ellisville, MO, USA), was dissolved in media as a stock concentration. The mGluR5 agonist, (RS)-2-chloro-5-hydroxyphenylglycine (CHPG; Tocris), was dissolved in DMSO as a stock solution. Both were diluted in culture media on treatment. The selective iNOS antagonist 1400 W (N-[[3-(aminomethyl)phenyl]methyl]-ethanimidamide dihydrochloride; Cayman Chemical) was dissolved in sterile PBS as a stock solution and dissolved in culture media on treatment. An equal volume of PBS and DMSO in media was used as vehicle controls for respective experiments. The above chemicals were incubated in culture media over bEnd.3 cells for the times and at the doses indicated for respective experiments (below).

Bradykinin Stimulation Test

Cerebral microvascular endothelial cells were treated for 48 hours with increasing doses of Hcy. The media was then replaced with fresh media and treated with either bradykinin (10⁻⁴ mol/L) or vehicle for 60 minutes. The media were harvested and used for NO_x analysis. Because Hcy can form nitrosothiols with NO, thereby quenching NO and confounding data interpretation, samples were pretreated with mercuric chloride (5 mmol/L for 10 minutes), which cleaves the Hcy-NO bond, allowing measurement of any NO bound with Hcy. The samples were assayed for NO_x with the same kit used for the in-vivo plasma measurements (above).

Cerebral Microvascular Endothelial Cell Western Blotting

The CMVECs were lysed and denatured in Laemmli sample buffer with 10% β-mercaptoethanol, separated by SDS-PAGE, blotted onto PVDF membranes, and probed using respective antibodies at 1:1,000 overnight at 4°C: rabbit anti-iNOS (sc-651; Santa Cruz Biotechnology), rabbit anti-nitrotyrosine (A21285; Invitrogen). Appropriate host species and alkaline phosphatase-conjugated secondary antibodies were applied for 1 hour and immunopositive bands were imaged (Versadoc; Bio-Rad) from the chemiluminescent signal (Lumiphos; Pierce) and quantified (Image J; NIH; USA). Data were normalized to total protein for each sample after reprobing of the blots with mouse anti-β-actin HRP conjugate (sc-47778; Santa Cruz Biotechnology).

Brain Microvascular Endothelial Cell Whole-Cell ELISA

After respective treatments, CMVECs were washed twice in ice-cold PBS. Phosphate-buffered saline was completely removed and the cells were incubated in 4% paraformaldehyde in PBS for 15 minutes, rinsed twice in Tris-buffered saline (TBS), permeabilized with 1% Triton X-100 in TBS for 15 minutes, and blocked in TBS with 0.01% Tween-20 (TBST) with 3% bovine serum albumin for 2 hours at room temperature. Samples were incubated with anti-3-NT (Invitrogen) diluted 1:1,000 or with anti-iNOS (Santa Cruz Biotechnology) diluted 1:250 in TBST with 3% bovine serum albumin at 4°C overnight. Samples were washed three times for 10 minutes each using TBST. Samples were incubated with alkaline phosphatase-conjugated goat anti-rabbit IgG (Pierce) for 30 minutes, washed as above and detected using a colorimetric ELISA substrate (BluePhos; KPL, Gaithersburg, MD, USA). Absorbance was measured at 620 nm using a plate reader (Synergy HT, BioTek).

Short Hairpin RNA

The mGluR5 knockdown was achieved using the lentiviral vector GIPZ containing the shRNAmir sequence V3LMM_455252 from Open Biosystems (Huntsville, AL, USA), and the nonsilencing shRNAmir (catalog # RHS4348) was used as a negative control. Optimal multiplicity of infection was determined following the manufacturer's instructions. A stable culture was generated by growing these cells in the presence of 10 μg/mL puromycin for 7 days. Percentage of mGluR5 knockdown was quantified with cell-surface ELISA by immunolabeling for the extracellular domain of mGluR5 (sc-47147; Santa Cruz Biotechnology).

AutoDock Prediction

AutoDock is a computer-based modeling system that predicts protein—protein interactions. The three-dimensional (3D) crystal structures of three ligands (glutamate, CHPG, and Hcy) and mGluR5 (protein data bank code is 3LMK) were obtained through a protein data bank. All ligands were given hydrogen atoms, defined ROOT and rotatable bonds, and assigned charges by the default settings. Then, the macromolecule mGluR5 was given hydrogen atoms, checked for missing atoms, and assigned charges and solvation parameters. Map Types and GRID dimension were chosen as each ligand and 100, respectively. Then, the best conformers and docked energy were calculated and displayed.

Data Analysis

Data were compared with one-way analysis of variance with Tukey or Dunnett post hoc analysis when overall significance was found; α was set at 0.05 for this project. All mean values are reported with their n and standard error of the mean. For comparison between only two groups, the unpaired Student's t-test was used. Statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software, San Diego, CA, USA).

Results

Homocysteine Concentration

Serum Hcy concentration was significantly higher in cbs^+/– mice (10.5 ± 0.9 μmol/L) as compared with cbs^+/+ mice (4.8 ± 0.6 μmol/L), consistent with our previous report in this model (Beard and Bearden, 2011).

Evidence of Nitrative and Oxidative Stress in Brains of Mice with Mild Hyperhomocysteinemia

Analysis of brain microvascular and vessel-depleted fractions from cbs^+/– versus cbs^+/+ mice showed that proteins from the microvascular fraction of cbs^+/– brains have significantly higher 3-NT immunolabeling. The mean 3-NT labeling of proteins from vessel-depleted fractions was also higher in cbs^+/– mice; however, the data had considerable variability and differences for vessel-depleted fractions were not statistically significant (Figures 1A and 1B). As a second marker of oxidative stress in the brains of these mice, we measured GSH levels in both fractions. Though not statistically significant, mean GSH levels were lower in the microvascular fractions of cbs^+/– mice and similarly or slightly higher in the vessel-depleted fractions (Figures 1C and 1D). As an indicator of lipid peroxidation, we measured 4-HNE in the brains of these mice. The 4-HNE levels were significantly higher in the vessel-depleted (parenchymal) fractions but not in the microvascular fractions (Figures 1E and 1F).

Figure 1

Glutathione (GSH) and total protein nitrotyrosinylation are decreased in brain microvessels of hyperhomocysteinemia (HHcy) mice compared with wild-type littermates, while 4-hydroxynonenal (4-HNE) was increased in vessel-depleted fraction. Using a murine model of HHcy (cbs^+/–) and wild-type (WT) littermates, vessel-depleted brain homogenates and brain microvascular fractions were collected. Protein nitrotyrosinylation (3-nitrotyrosine; 3-NT) was quantified (A) in the vessel-depleted fraction and (B) the microvascular fraction. Glutathione was quantified (C) in the vessel-depleted fraction and (D) the microvascular fraction. 4-hydroxynonenal was quantified (E) in the vessel-depleted fraction and (F) the microvascular fraction. For consistency across experiments, data are presented relative to controls (WT). *Indicates P < (n = 4).

Elevated Homocysteine Is Associated with Decreased Extracellular Nitric Oxide Bioavailability

Decreased plasma nitrate + nitrite (NO_x) is commonly used as a marker of endothelial dysfunction. We measured plasma NO_x levels in cbs^+/– mice. NO_x was significantly lower in cbs^+/– mice compared with their wild-type littermates, indicating lower NO bioavailability in HHcy (Figure 2A). To confirm a direct effect of Hcy on endothelium-derived NO and the potential involvement of mGluR5, we treated confluent CMVECs with Hcy at 20 μm, MPEP at 20 μm, and CHPG at 25 μm for 48 hours, followed by 1 hour of bradykinin stimulation (Figure 2B). Homocysteine decreased the amount of NO_x in culture media after stimulation. This effect was rescued by treatment with the mGluR5 antagonist MPEP. Direct activation of mGluR5 with CHPG mimicked the effects of Hcy treatment.

Figure 2

Homocysteine (Hcy) is associated with decreased extracellular nitric oxide (NO) bioavailability. (A) As a marker of NO bioavailability, serum NO_x measurements were made in hyperhomocysteinemia (HHcy) mice and their wild-type (WT) littermates (n = 4). *Indicates P < compared with control. (B) Brain microvascular endothelial (bEnd.3) cells were treated with Hcy, 2-methyl-6-(phenylethynyl) pyridine hydrochloride (MPEP), (RS)-2-chloro-5-hydroxyphenylglycine (CHPG), and dimethyl sulfoxide (DMSO) (vehicle control for CHPG) or water (vehicle control for Hcy and MPEP) for 48 hours. Media NO_x measurements were made 60 minutes after stimulation with bradykinin (10⁻⁴ mol/L) (n = 6). *Indicates P < compared with control and Hcy + MPEP; data for the three control conditions (water, MPEP alone, DMSO; n = 6 each) were not different from one another and were pooled.

Homocysteine Increases Inducible Nitric Oxide Synthase Expression in Cerebral Microvascular Endothelial Cells in a Dose- and Time-Dependent Manner

To identify a potential source of increased oxidative/nitrative stress in the cerebral endothelium, we examined the effect of Hcy on iNOS expression. Confluent CMVECs were treated with 0, 2, or 20 μmol/L Hcy for 24 hours. In all, 20 μmol/L Hcy increased iNOS expression> 2-fold as determined by western blotting (Figure 3A). To determine if the effect of Hcy persisted beyond 24 hours, we treated CMVECs with 20 μmol/L Hcy for 0, 2, and 9 days. At both 2 and 9 days, iNOS expression remained> 2-fold greater than baseline levels indicating a sustained effect of chronic HHcy (Figure 3B), which is consistent with the chronic disease state in patients. Cells were treated with 20 μmol/L Hcy, 20 μmol/L MPEP, and 25 μmol/L CHPG to determine the influence of mGluR5 in the elevated expression of iNOS. CHPG had the equivalent effect of Hcy on iNOS expression. Treating with MPEP provided a partial rescue compared with control and reduced expression significantly from the Hcy treatment group (Figure 3C). Homocysteine did not affect the viability of cells in culture (Figure 3D).

Figure 3

Homocysteine (Hcy) dose and time dependently increases brain microvascular endothelial cell expression of inducible nitric oxide synthase (iNOS) via metabotropic glutamate receptor-5 (mGluR5). (A) Brain microvascular endothelial (bEnd.3) cells were treated with increasing doses of Hcy for 24 hours. (B) Using 20 μmol/L Hcy, bEnd.3 cells were treated daily for 9 days and iNOS expression was quantified by western blot after 2 and 9 days treatment. (C) Cells were treated with 20 μmol/L Hcy, 20 μmol/L 2-methyl-6-(phenylethynyl) pyridine hydrochloride (MPEP), and 25 μmol/L (RS)-2-chloro-5-hydroxyphenylglycine (CHPG) (n = 10/group) for 48 hours. Dimethyl sulfoxide (DMSO) is the vehicle control for CHPG. (D) Cell viability was determined by MTT assay 48 hours after treatment with the indicated doses of Hcy For consistency across experiments, data are presented relative to controls (WT or 0). *Indicates P < compared with vehicle controls, ^# indicates P < compared with Hcy.

Homocysteine Increases Protein Nitrotyrosinylation in Cerebral Microvascular Endothelial Cells in a Dose- and Time-Dependent Manner

Finding significantly greater 3-NT labeling in brain microvascular isolates of cbs^+/– mice, we determined the direct effect of Hcy on 3-NT modifications of proteins from CMVECs in culture. Treating CMVECs for 24 hours with 0, 2, and 20 μmol/L Hcy led to a significant increase in total protein 3-NT as evidenced by increased 3-NT immunolabeling of multiple bands on western blotting of cell lysates (Figure 4A). For quantification of 3-NT labeling, one band was arbitrarily chosen for densitometric analysis of Hcy-mediated dose- and time-dependent increase of 3-NT (Figures 4B and 4C). Given that all protein bands showed increased 3-NT immunolabeling in response to Hcy (Figure 4A), we developed a whole-cell ELISA protocol which allowed us to repeat the Hcy dose-dependent 3-NT modifications with a second technique and for higher throughput analysis of receptor-dependent mechanisms. To confirm the efficacy of the ELISA, we treated CMVECs with media preconditioned by lipopolysaccharide-stimulated macrophages (diluted 1:4 in fresh media) as a positive control. Using this technique, treatment of CMVECs for 48 hours with both positive controls and Hcy treatments (20, 60, and 80 μmol/L) led to a significant increase in 3-NT expression (Figure 4D). For subsequent experiments, the whole-cell ELISA technique was used.

Figure 4

Homocysteine (Hcy) induces protein nitrotyrosinylation (3-nitrotyrosine; 3-NT) by triggering metabotropic glutamate receptor-5 (mGluR5) and inducible nitric oxide synthase (iNOS) activity. (A) Brain microvascular endothelial (bEnd.3) cells were treated with increasing doses of Hcy for 24 hours. Cells were harvested and total protein nitrotyrosinylation was determined by western blot. (B) Given the increase in 3-NT immunolabeling for all protein bands, the band at ∼55 kDa was used for quantification. (C) Using 20 μmol/L Hcy, bEnd.3 cells were treated daily for 9 days and 3-NT immunolabeling was quantified by western blot after 2 and 9 days treatment. (D) Homocysteine dose-response of 3-NT was determined using whole-cell ELISA. Further experiments were performed using ELISA. (E) To determine if Hcy treatment leads to protein nitrotyrosinylation via activation of mGluR5, the specific mGluR5 agonist ((RS)-2-chloro-5-hydroxyphenylglycine; CHPG) and antagonist (2-methyl-6-(phenylethynyl) pyridine hydrochloride; MPEP) were used. (F) To determine the role of iNOS in Hcy-mediated protein nitrotyrosinylation, a specific iNOS inhibitor was used (1400 W). For consistency across experiments, data are presented relative to controls (WT). * indicates P < when compared with appropriate vehicle controls; and ^# indicates P < compared with Hcy group.

Homocysteine-Induced 3-Nitrotyrosine Modifications in Cerebral Microvascular Endothelial Cells Depend on Metabotropic Glutamate Receptor-5 Activation and Inducible Nitric Oxide Synthase Activity

With increased 3-NT modifications in vivo and in vitro, and the possibility that Hcy could be modulating its effects through mGluR5 activation on CMVECs, we tested the hypothesis that Hcy-induced protein nitrotyrosinylation in CMVECs is dependent on mGluR5 activation. Directed pharmacology using an mGluR5-specific agonist, CHPG, significantly increased 3-NT formation. The selective mGluR5 antagonist, MPEP, blocked Hcy-mediated 3-NT formation while MPEP alone had no effect on baseline 3-NT levels (Figure 4E). Given the increased expression of iNOS in CMVECs treated with Hcy, we then tested if iNOS activity was necessary for Hcy-induced nitrotyrosinylation. An iNOS-specific inhibitor, 1400 W at 10 μmol/L, ameliorated Hcy-induced protein nitrotyrosinylation in CMVECs (Figure 4F).

Short Hairpin RNA

After shRNA and puromycin selection, cell-surface expression of mGluR5 in knockout cells (KO) was 47 ± 12% of that in nonsilencing control cells (53% knockdown; P <). Compared with untreated cells, Hcy (20 μmol/L) increased expression of iNOS and 3-NT in nonsilencing cells (P <) but did not increase expression in KO cells (Figure 5).

Figure 5

Knockdown of metabotropic glutamate receptor-5 (mGluR5) expression blocks the increase in inducible nitric oxide synthase (iNOS) and 3-nitrotyrosine (3-NT) expression caused by homocysteine (Hcy). Lentiviral knockdown of mGluR5 (Short hairpin; shRNA) and nonsilencing control cells were treated with Hcy (20 μmol/L) for 48 hours. Inducible nitric oxide synthase and 3-NT were quantified by ELISA. * indicates P <.

Homocysteine Is Predicted to Bind to the Orthosteric Site of Metabotropic Glutamate Receptor-5

Pharmacology shows a critical role of mGluR5 in Hcy-mediated iNOS expression and subsequent 3-NT of cellular proteins. We questioned whether Hcy is likely to bind directly with mGluR5 and specifically, whether it may bind the orthosteric region of the receptor. As a first step, and to complement the experiments presented here, we used a computational model that predicts protein interactions (AutoDock Tools). Binding regions of mGluR5 are known for glutamate (agonist) and CHPG (agonist); i.e., the orthosteric binding site. Figure 6 shows that sites bound by glutamate and CHPG are those already reported, demonstrating the accuracy of the model prediction. Homocysteine was predicted to bind the same region, which is consistent with activation of the receptor. Further, Hcy was predicted to bind the same amino acids as glutamate, S152 and T175, and has a docked energy similar to that for CHPG.

Discussion

The major new contribution of these studies is the identification of a receptor that is both necessary and sufficient for Hcy-mediated oxidative and specifically nitrative stress in cerebral microvascular endothelium. In vivo and in vitro, Hcy increased nitrative stress as quantified by 3-NT formation. Hcy-mediated nitrative stress required mGluR5 activation and iNOS expression. HHcy is an independent risk factor for stroke (Eikelboom et al, 2000), and several clinical trials report a significant relation between Hcy levels and silent brain infarcts or white-matter lesions (Seshadri et al, 2008). Moreover, cognitive decline after stroke is worsened by HHcy (Khedr et al, 2009). Effective therapies for ameliorating cerebrovascular damage in HHcy are needed. These must be grounded in a deeper understanding of the mechanisms by which Hcy initiates downstream effects. Our study begins to dissect novel mechanisms that may be useful therapeutic targets for future pharmacologic intervention.

Nitrative stress contributes to microvascular degeneration and loss of brain mass (, 2011; Sirinyan et al, 2006), which may be a factor in Hcy-associated white-matter lesions. Indeed, chronic nitrative stress manifests in some diseases of the central nervous system. For example, a general theory of the pathogenesis of AD and other neurodegenerative diseases involves this type of nitrative stress (Malinski, 2007). This damage extends to the vasculature where nitrative stress is thought to underlie impaired reactivity of peripheral resistance vasculature in HHcy (Ungvari et al, 2002). Cerebrovascular nitrative stress is implicated in the etiology of neurovascular damage associated with blood—brain barrier disruption (Sandoval and Witt, 2008) and stroke (Liang et al, 2004). We previously showed that Hcy increases iNOS expression and 3-NT formation in human umbilical vein endothelial cells (Bearden et al, 2010). The present studies show complementary results in CMVECs and extend the findings by identifying a receptor-dependent mechanism. In the present study, formation of 3-NT in response to Hcy required both the activation of mGluR5 and the increased expression and activity of iNOS. An association between Hcy and glutamate receptors in controlling intracellular processes has been reported. Kuszczyk et al (2009) reported that mGluR5 antogonists reduced the Hcy-mediated hyperphosphorylation of τ proteins in neurons, a process central to AD. In human cardiac microvascular endothelial cells, Moshal et al (2006) showed a pertussis toxin-sensitive mechanism (G-protein coupled receptor, sensitive) for Hcy-mediated secretion of matrix metalloproteinase-9, which is consistent with our findings because mGluR5 is a G-protein coupled receptor.

Figure 6

AutoDock predicts orthosteric binding of homocysteine (Hcy) to metabotropic glutamate receptor-5 (mGluR5). AutoDock, a computer-based modeling software program, was used to predict mGluR5 binding sites and calculate docked energies for (A) glutamate, (B) (RS)-2-chloro-5-hydroxyphenylglycine (CHPG) (a selective agonist), and (C) Hcy. Glutamate and CHPG are known to bind mGluR5 at the regions predicted by the program, demonstrating accuracy in our modeling. Homocysteine is predicted to bind the same amino acids as glutamate with an energy similar to that for CHPG, consistent with experimental results.

Homocysteine-induced production of NO in vascular smooth muscle cells depends on an increase in iNOS expression and activity (Welch et al, 1998). Our data extend these findings to vascular endothelium, demonstrating a consistent mechanism by which Hcy increases nitrative burden in blood vessels. Contrary to other forms of NOS, iNOS is calcium insensitive. When expressed, iNOS continuously generates large quantities of NO which leads to peroxynitrite formation when superoxide is available. Both NO and superoxide are made by iNOS, with heightened production of superoxide if l-arginine levels are low; hence, iNOS is a significant producer of peroxynitrite (Xia and Zweier, 1997). Peroxynitrite may contribute to endogenous regulation and signaling at low levels but is also toxic to cells at high concentrations. This is the primary function of iNOS during the ‘oxidative burst’ of macrophages during phagocytosis of cells and microbes. In the brain, after middle cerebral artery occlusion to produce stroke-like ischemia, iNOS expression peaks within 14 to 24 hours. Remarkably, stimulation of the cerebellum significantly impaired the increase in iNOS with a concomitant reduction in infarct volume (Galea et al, 1998), suggesting a direct link between iNOS activity and neural damage during ischemic injury. The oxidative stress imposed during an ischemia-reperfusion injury leads to opening of the blood—brain barrier, which is significantly rescued if iNOS activity is inhibited after the injury (Yao et al, 2005). Thus, iNOS has a crucial role in disrupting the normal functional properties of cerebral microvasculature and protection of neurons in the acute setting of stroke and ischemia-reperfusion injuries. Our results suggest that these processes may also have a role in the more insidious and progressive damage of chronic low grade HHcy.

Previous experiments reported increased levels of lipid peroxidation including 4-HNE in vulnerable regions of the AD brain (Butterfield et al, 2002). Furthermore, 4-HNE has been identified in early onset AD with mild-cognitive decline, suggesting its role in early pathogenesis. 4-Hydroxynonenal is a commonly used marker of neuronal stress secondary to ischemic conditions, such as stroke, that manifest from impaired cerebral blood flow. 4-Hydroxynonenal in cerebral endothelium is associated with an increase in blood—brain barrier permeability (Mertsch et al, 2001); however, we did not find an increase in 4-HNE in the microvascular fraction. Therefore, cbs^+/– mice present with differential oxidative stress in vasculature (i.e., peroxynitrite and 3-NT formation) as compared with brain parenchyma (i.e., lipid oxidation and 4-HNE formation). These data show a difference in the substrate for oxidative stress in respective fractions: protein in microvessels (protein nitration) but lipid in brain parenchyma (4-HNE). Whether there is significant disruption of the white matter (e.g., white-matter lesions) in cbs^+/– mice remains to be determined. It is possible that the greater lipid content of brain parenchyma, owing primarily to oligodendrocytes of white matter, presents a larger substrate potential for free radicals to attack lipid while induction of iNOS in vascular endothelium, and concomitant increases in local NO production, pushes the balance toward peroxynitrite. Teasing out these mechanisms will require focused experiments in future studies. Overexpression of cellular GSH rescues Hcy-mediated reduction in endothelial cell production of NO (Weiss et al, 2001), demonstrating a significant role of redox imbalance in reducing bioavailability of NO, which is consistent with the present study. These results provide evidence that elevated Hcy reduces NO bioavailability by an mGluR5-dependent mechanism.

A strength of our study is the time course of treatments. Many studies have focused on the acute (hours) or subchronic (24 hours) effects of Hcy on cells or tissues. Our experiments, with treatment periods as long as 9 days, show that iNOS expression and 3-NT remain elevated under chronic treatment. In the clinical setting, HHcy is a chronic condition. Therefore, we propose that 3-NT formation in HHcy is a continual and ongoing cellular burden that may be modifiable through pharmacologic intervention at mGluR5. An additional strength of our study is the use of a physiologically relevant level of Hcy (20 μmol/L d/l-Hcy in culture and mild HHcy in vivo, 10.5 ± 0.9 μmol/L). Collectively, these studies provide evidence for the idea that even mild elevations in Hcy levels may accumulate significant cerebral microvascular nitrative stress over a lifetime. This study does have limitations. Specifically, we have not yet identified the source of superoxide for quenching NO to form peroxynitrite. While iNOS itself is a likely candidate as discussed above, another potential source is NAD(P)H oxidase. For example, Ungvari et al (2003) reported a critical role for NAD(P)H oxidase in Hcy-mediated superoxide production in coronary arteries from a diet-induced rat model of HHcy. Alternatively, formation of peroxynitrite can drive a feed-forward cascade for further peroxynitrite production because the molecule can uncouple tetrahydrobiopterin from endothelial NO synthase; uncoupled endothelial NO synthase shuttles electrons to oxygen to form superoxide (Laursen et al, 2001). We previously reported that diet-induced HHcy in mice impairs endothelial NO synthase expression and phosphorylation at S1177 coincident with reduced NO-dependent dilation in mouse mesenteric resistance vessels (Looft-Wilson et al, 2008). However, we have yet to quantify levels of tetrahydrobiopterin or endothelial NO synthase uncoupling in the models used in the present set of studies.

These results define mGluR5 as a necessary and sufficient player in Hcy-mediated 3-NT formation in cerebral microvascular endothelium. Computer modeling simulations predict that Hcy is a ligand for the orthosteric site of mGluR5, consistent with the experimental data. The strong association among HHcy, nitrative stress, and neurovascular disease (e.g., stroke and cognitive impairment) suggests a causative role. We propose that mGluR5 may be a potential therapeutic target in the cerebral endothelial dysfunction caused by HHcy, which could be useful in treating Hcy-related neurovascular disease.

Footnotes

The authors declare no conflict of interest.

References

Beard

Jr Bearden

(2011) Vascular complications of cystathionine beta-synthase deficiency: future directions for homocysteine-to-hydrogen sulfide research. Am J Physiol Heart Circ Physiol 300:H13–26

Beard

Jr Reynolds

Bearden

(2011) Hyperhomocysteinemia increases permeability of the blood-brain barrier by NMDA receptor-dependent regulation of adherens and tight junctions. Blood 118:2007–14

Bearden

Beard

Jr Pfau

(2010) Extracellular transsulfuration generates hydrogen sulfide from homocysteine and protects endothelium from redox stress. Am J Physiol Heart Circ Physiol 299:H1568–76

Boldyrev

Johnson

(2007) Homocysteine and its derivatives as possible modulators of neuronal and nonneuronal cell glutamate receptors in Alzheimer's disease. J Alzheimers Dis 11:219–28

Butterfield

Castegna

Lauderback

Drake

(2002) Evidence that amyloid beta-peptide-induced lipid peroxidation and its sequelae in Alzheimer's disease brain contribute to neuronal death. Neurobiol Aging 23:655–64

Carmel

Jacobsen

(2001) Homocysteine in Health and Disease. Cambridge, UK, New York: Cambridge University Press

Dayal

Arning

Bottiglieri

Boger

Sigmund

Faraci

Lentz

(2004) Cerebral vascular dysfunction mediated by superoxide in hyperhomocysteinemic mice. Stroke 35:1957–62

Drew

Leeuwenburgh

(2002) Aging and the role of reactive nitrogen species. Ann NY Acad Sci 959:66–81

Eikelboom

Hankey

Anand

Lofthouse

Staples

Baker

(2000) Association between high homocyst(e)ine and ischemic stroke due to large- and small-artery disease but not other etiologic subtypes of ischemic stroke. Stroke 31:1069–75

10.

Faraci

(2011) Protecting against vascular disease in brain. Am J Physiol Heart Circ Physiol 300:H1566–82

11.

Fiebich

Butcher

Gebicke-Haerter

(1998) Protein kinase C-mediated regulation of inducible nitric oxide synthase expression in cultured microglial cells. J Neuroimmunol 92:170–8

12.

Galea

Golanov

Feinstein

Kobylarz

Glickstein

Reis

(1998) Cerebellar stimulation reduces inducible nitric oxide synthase expression and protects brain from ischemia. Am J Physiol 274:H2035–45

13.

Gerhart

Broderius

Drewes

(1988) Cultured human and canine endothelial cells from brain microvessels. Brain Res Bull 21:785–93

14.

Grimm

Hoehn

Davies

Grune

(2011) Protein oxidative modifications in the ageing brain: consequence for the onset of neurodegenerative disease. Free Radic Res 45:73–88

15.

Hassan

Hunt

O'Sullivan

Bell

D'Souza

Jeffery

Bamford

Markus

(2004) Homocysteine is a risk factor for cerebral small vessel disease, acting via endothelial dysfunction. Brain 127:212–9

16.

Kamath

Chauhan

Kisucka

Dole

Loscalzo

Handy

Wagner

(2006) Elevated levels of homocysteine compromise blood-brain barrier integrity in mice. Blood 107:591–3

17.

Khedr

Hamed

El-Shereef

Shawky

Mohamed

Awad

Ahmed

Shehata

Eltahtawy

(2009) Cognitive impairment after cerebrovascular stroke: Relationship to vascular risk factors. Neuropsychiatr Dis Treat 5:103–16

18.

Kuszczyk

Gordon-Krajcer

Lazarewicz

(2009) Homocysteine-induced acute excitotoxicity in cerebellar granule cells in vitro is accompanied by PP2A-mediated dephosphorylation of tau. Neurochem Int 55:174–80

19.

Laursen

Somers

Kurz

McCann

Warnholtz

Freeman

Tarpey

Fukai

Harrison

(2001) Endothelial regulation of vasomotion in apoE-deficient mice: implications for interactions between peroxynitrite and tetrahydrobiopterin. Circulation 103:1282–8

20.

Lehmann

Regland

Blennow

Gottfries

(2003) Vitamin B12-B6-folate treatment improves blood-brain barrier function in patients with hyperhomocysteinaemia and mild cognitive impairment. Dement Geriatr Cogn Disord 16:145–50

21.

Liang

Dawson

(2004) What have genetically engineered mice taught us about ischemic injury? Curr Mol Med 4:207–25

22.

Looft-Wilson

Ashley

Billig

Wolfert

Ambrecht

Bearden

(2008) Chronic diet-induced hyperhomocysteinemia impairs eNOS regulation in mouse mesenteric arteries. Am J Physiol Regul Integr Comp Physiol 295:R59–66

23.

Malinski

(2007) Nitric oxide and nitroxidative stress in Alzheimer's disease. J Alzheimers Dis 11:207–18

24.

Mertsch

Blasig

Grune

(2001) 4-Hydroxynonenal impairs the permeability of an in vitro rat blood-brain barrier. Neurosci Lett 314:135–8

25.

Montesano

Pepper

Mohle-Steinlein

Risau

Wagner

Orci

(1990) Increased proteolytic activity is responsible for the aberrant morphogenetic behavior of endothelial cells expressing the middle T oncogene. Cell 62:435–45

26.

Moshal

Sen

Tyagi

Henderson

Steed

Ovechkin

Tyagi

(2006) Regulation of homocysteine-induced MMP-9 by ERK1/2 pathway. Am J Physiol Cell Physiol 290:C883–91

27.

Ridnour

Winters

Ercal

Spitz

(1999) Measurement of glutathione, glutathione disulfide, and other thiols in mammalian cell and tissue homogenates using high-performance liquid chromatography separation of N-(1-pyrenyl)maleimide derivatives. Methods Enzymol 299:258–67

28.

Sandoval

Witt

(2008) Blood-brain barrier tight junction permeability and ischemic stroke. Neurobiol Dis 32:200–19

29.

Seshadri

Wolf

Beiser

Selhub

Jacques

Yoshita

Rosenberg

D'Agostino

DeCarli

(2008) Association of plasma total homocysteine levels with subclinical brain injury: cerebral volumes, white matter hyperintensity, and silent brain infarcts at volumetric magnetic resonance imaging in the Framingham Offspring Study. Arch Neurol 65:642–9

30.

Shi

Savage

Hufeisen

Rauser

Grajkowska

Ernsberger

Wroblewski

Nadeau

Roth

(2003) L-homocysteine sulfinic acid and other acidic homocysteine derivatives are potent and selective metabotropic glutamate receptor agonists. J Pharmacol Exp Ther 305:131–42

31.

Sirinyan

Sennlaub

Dorfman

Sapieha

Gobeil

Jr Hardy

Lachapelle

Chemtob

(2006) Hyperoxic exposure leads to nitrative stress and ensuing microvascular degeneration and diminished brain mass and function in the immature subject. Stroke 37:2807–15

32.

Ungvari

Csiszar

Bagi

Koller

(2002) Impaired nitric oxide-mediated flow-induced coronary dilation in hyperhomocysteinemia: morphological and functional evidence for increased peroxynitrite formation. Am J Pathol 161:145–53

33.

Ungvari

Csiszar

Edwards

Kaminski

Wolin

Kaley

Koller

(2003) Increased superoxide production in coronary arteries in hyperhomocysteinemia: role of tumor necrosis factor-alpha, NAD(P)H oxidase, and inducible nitric oxide synthase. Arterioscler Thromb Vasc Biol 23:418–24

34.

Weiss

Zhang

Heydrick

Bierl

Loscalzo

(2001) Overexpression of cellular glutathione peroxidase rescues homocyst(e)ine-induced endothelial dysfunction. Proc Natl Acad Sci USA 98:12503–8

35.

Welch

Upchurch

Jr Farivar

Pigazzi

Brecher

Keaney

Jr Loscalzo

(1998) Homocysteine-induced nitric oxide production in vascular smooth-muscle cells by NF-kappa B-dependent transcriptional activation of Nos2. Proc Assoc Am Physicians 110:22–31

36.

Winters

Zukowski

Ercal

Matthews

Spitz

(1995) Analysis of glutathione, glutathione disulfide, cysteine, homocysteine, and other biological thiols by high-performance liquid chromatography following derivatization by n-(1-pyrenyl)maleimide. Anal Biochem 227:14–21

37.

Xia

Zweier

(1997) Superoxide and peroxynitrite generation from inducible nitric oxide synthase in macrophages. Proc Natl Acad Sci USA 94:6954–8

38.

Yao

Zhou

Zhu

(2005) The effect of inducible nitric oxide synthase inhibitor on microvascular permeability after cerebral ischemia/reperfusion. Methods Find Exp Clin Pharmacol 27:701–4

39.

Zieminska

Lazarewicz

(2006) Excitotoxic neuronal injury in chronic homocysteine neurotoxicity studied in vitro: the role of NMDA and group I metabotropic glutamate receptors. Acta Neurobiol Exp (Wars) 66:301–9