To test the hypothesis that endothelial dysfunction in hyperhomocysteinemia was due to increased levels of nitrotyrosine and matrix metalloproteinase (MMP) activity in response to antagonism of peroxisome proliferator-activated receptor-α (PPAR-α), cystathionine β-synthase (CBS) −/+ mice were bred, tail tissue was analyzed for genotype by PCR, and tail vein blood was analyzed for homocysteine (Hcy) by spectrofluorometry. To induce PPAR-α, mice were administered 8 μg/ml of ciprofibrate (CF) and grouped: 1) wild type (WT), 2) WT + CF, 3) CBS, 4) CBS + CF (n = 6 in each group). In these four groups of mice, plasma Hcy was 3.0 ± 0.2, 2.5 ± 1.2, 15.2 ± 2.6 (P < 0.05 compared with WT), 11.0 ± 2.9 μmol/l. Mouse urinary protein was 110 ± 11, 86 ± 6, 179 ± 13, 127 ± 9 μg · day−1 · kg−1by Bio-Rad dye binding assay. Aortic nitrotyrosine was 0.099 ± 0.012, 0.024 ± 0.004, 0.132 ± 0.024 (P < 0.01 compared with WT), 0.05 ± 0.01 (scan unit) by Western analysis. MMP-2 activity was 0.053 ± 0.010, 0.024 ± 0.002, 0.039 ± 0.009, 0.017 ± 0.006 (scan unit) by zymography. MMP-9 was specifically induced in CBS −/+ mice and inhibited by CF treatment. Systolic blood pressure (SPB) was 90 ± 2, 88 ± 16, 104 ± 8 (P < 0.05 compared with WT), 96 ± 3 mmHg. Aortic wall stress [(SPB · radius2/wall thickness)/2(radius + wall thickness)] was 10.2 ± 1.9, 9.7 ± 0.2, 16.6 ± 0.8 (P < 0.05 compared with WT), 13.1 ± 2.1 dyn/cm2. The results suggest that Hcy increased aortic wall stress by increasing nitrotyrosine and MMP-9 activity.
- extracellular matrix
- matrix metalloproteinase
- tissue inhibitor of metalloproteinase
- cystathionine β-synthase
- nitric oxide
homocysteine (Hcy) is involved in nucleic acid methylation (50). There are four ways by which hyperhomocysteinemia (HHcy) is developed: 1) by methionine-rich protein diet, 2) vitamin B12/folate deficiency, 3) heterozygous/homozygous in cystathione β-synthase (CBS) activity and B6 deficiency, and 4) renovascular stenosis and volume retention. Although the treatment of vitamin B12/folate reduced the level of plasma Hcy and ameliorated vascular dysfunction, in part, by conversion to methionine (37), the mechanism of other causes of HHcy is unclear. In human, heterozygosity (−/+) in CBS activity was associated with HHcy and increased oxidative stress (8). In CBS −/+ mice, Hcy induced endothelial dysfunction (11). Hcy caused vascular disease (20), including arteriosclerosis (43), endothelial cell desquamation (40), thromboresistance (23), smooth muscle cell proliferation (44), collagen synthesis (44), oxidation of low-density lipoprotein (17), increased monocyte adhesion to the vessel wall (21), platelet aggregation (7), coagulation (33), blood rheology (25), and activation of plasminogen and matrix metalloproteinase (MMP) (16). Hcy decreased endothelial nitric oxide (NO) concentration and promoted the formation of nitrotyrosine in the vessel wall (29). Hcy induced oxidative stress (2) and activated NF-κB (5). A negative correlation between high Hcy level and peroxisome proliferator-activated receptor (PPAR) expression has been suggested (6). The agonists of PPAR attenuated oxidative stress-mediated vascular dysfunction (12) and hypertension (34). In addition, PPAR promoted the synthesis of superoxide dismutase (SOD) and catalase (18) and decreased NADPH oxidase (18). Although Hcy induced constrictive vascular remodeling and PPAR agonists ameliorated remodeling in vitro (30), the increased oxidative stress was associated with increased MMP activity (45), and the agonists of PPAR decreased oxidative stress and MMP activity in macrophages (24). It was unclear whether PPAR ameliorated Hcy-mediated MMP activation in vivo. We hypothesized that Hcy downregulated PPAR-α, thereby causing oxidation of NO to nitrotyrosine with consequent increased activation of MMP. MMP-induced changes in vascular wall structure, in turn, may alter wall stress and hemodynamics.
MATERIALS AND METHODS
A breeding pair of mice heterozygous for the CBS gene (C57BL/6J-Cbs tm1Unc) was obtained from Jackson Laboratories and bred at the mice breeding facility of the University of Mississippi Medical Center. CBS −/+ was created by inactivating CBS gene using homologous recombination in C57BL/6J mouse embryonic stem cells by disrupting the coding sequence of the CBS gene with the Neo gene as described (49). Homozygous −/− CBS mice do not survive past 2–3 wk. However, the heterozygous (−/+) CBS-deficient mice and their wild-type (+/+) littermate controls survive normally. At this age, CBS −/+ mice develop homocysteinemia. To minimize sex differences, we performed all experiments in male mice. Animal room temperature is maintained between 22 and 24°C. A 12-h light-dark cycle was maintained by artificial illumination. In accordance with the National Institutes of Health Guidelines for animal research, all animal procedures are reviewed and approved by the Institutional Animal Care and Use Committee of the University of Mississippi Medical Center, Jackson. To induce PPAR-α, we administered ciprofibrate (CF; Sigma) to CBS −/+ and wild-type +/+ mice at 40 μg/day in drinking water. The mice were grouped as follows: 1) wild type,2) wild type plus CF, 3) CBS, and 4) CBS plus CF. To determine selectivity of CF in the absence of Hcy, we administered CF to wild-type mice. In humans, 100 mg/day of CF has a potent effect (12). On the basis of the fact that the binding constant between CF and PPAR is in the micromolar range (31), 8 μg/ml of CF were administered to mice in drinking water. Because mice continuously drink and excrete ∼5 ml of water/day, each ingests 2 mg · kg−1 · day−1CF. This produced a blood concentration of ∼32 μmol/l, enough to saturate most binding sites on PPAR. The animals were fed standard chow and water ad libitum. To determine whether treatment with CF caused any change in food and water intake, we measured food and water every 2 days during the treatment period. There was no difference in food and water intake in either group. Because previous studies have demonstrated significant vascular dysfunction at 12 wk of homocysteinemia (26), CF was administered for 12 wk.
Genotype and phenotype determination.
The tail vein blood and tissue from offspring at the age of ∼8 wk weighing ∼20 ± 3 g were collected and analyzed for1) genomic DNA by PCR using specific CBS primers (49) and 2) the levels of plasma Hcy. The DNA was extracted and amplified by PCR for sequences in intron 3 and Neo insert in CBS −/+ mice (49). The PCR primers used were: 5′-GCCTCTGTCTGCTAACCTA-3′ and 5′-GAGGTCGACGGTATCGATA-3′ (49). On the basis of the genotype and the levels of plasma Hcy, the mice were categorized as wild type (+/+) and CBS (−/+) (49). This produced wild-type littermates from the same breed of mice.
Aortic morphology and in situ MMP activity.
The aorta was stained with trichrome and van Gieson for collagen and elastin as described (46). The intima-media thickness was measured by a digital micrometer. To determine total MMP activity in the aortas of CBS −/+ and +/+ mice, we performed in situ zymography as described (46). Briefly, freshly isolated aortic segments were laid onto gelatin-gel preequilibrated with Triton X-100 and incubation buffer. The gels after 18 h were stained for lytic activity with Coomassie blue.
Plasma Hcy and urinary protein.
Hcy was measured by a modification of a procedure by Frantzen et al. (13). Briefly, the plasma was reduced by trace amounts of reduced glutathione. The Hcy was converted toS-adenosyl-l-homocysteine (SAH) by incubating the plasma with SAH hydrolase (Sigma) and adenosine. The fluorescence of the incubate SAH was measured 422 nm when excited at 320 nm. The free nucleotides do not fluoresce. The adenosine and hydrolase were used in reference buffer. The standards of SAH (Sigma) were prepared. Because mouse urine had large quantities of proteins [primary mouse major urinary protein (MUP)] and males had much higher urinary MUP than females, to collect the urine, we caged male mice in 24-h metabolic cages for several days of acclimatization to reduce separation effects before hemodynamic measurements. MUP was measured by Bio-Rad dye binding assay.
Mice were anesthetized with tribromoethanol (100 mg/kg ip). This drug had minimal effect on cardiovascular function in mice (32). The aortic blood pressure (BP), heart rate, and systolic (SBP) and diastolic blood pressure were measured by a PE-10 catheter in the aorta through right common carotid artery (27). The catheter was connected to a pressure transducer (Micro-Med) positioned at the level of heart. Pulsatile arterial pressure signal was analyzed by a computer using customized software (Micro-Med). To determine plasma Hcy and to ensure mouse-to-mouse variation, we collected 0.5 ml of blood from each mouse by the same catheter.
Under deep anesthesia, the mice were euthanized by arresting the heart in diastole with injections of 0.2 ml/100 g body wt (iv) of a 20% KCl solution. The thoracic aorta was dissected. Aortic tissue homogenates were prepared (48). For PPAR measurements, aortic nuclear extracts were prepared (47). Bio-Rad dye binding assay was applied to estimate total protein concentration in the tissue extracts. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed with and without the reduction.
PPAR, nitrotyrosine, and β-actin.
The levels of nitrotyrosine and PPAR were measured by Western blot analysis using mouse monoclonal anti-nitrotyrosine antibody (Upstate Biotechnology) and mouse polyclonal anti-PPAR antibody (Calbiochem). The 10% SDS-PAGE at 20-mA/gel was carried out. The samples were prepared in reducing buffer. PPAR antibody recognized both α- and γ-isoforms. We established the specificity of antibodies by immunoprecipitating the antigen before loading them onto the gel, by antibody conjugated-agarose beads (Upstate Biotechnology). To determine whether total protein loaded onto the gel was identical, we performed β-actin Western blots using anti-β-actin antibody (Sigma). Alkaline phosphatase-conjugated secondary antibody was used as the detection system. Bands on blots were scanned by a Bio-Rad GS-700 densitometer.
Zymographic analysis of MMP activity.
To determine MMP-2 and -9 activity, we performed gelatin substrate gel zymography containing 1% gelatin in 8% SDS-PAGE (46). The aortic tissue homogenates were loaded onto the gel under identical condition of total protein. The scanned band intensity was normalized by β-actin.
Preparation of Hcy, acetylcholine, nitroprusside, and CF solutions.
The concentration of Hcy was determined by spectrophometric titration with dithio-bis-nitrobenzoate (absorption measured at 412 nm) using ε412 nm of 13,600 M−1cm−1(48). Concentrations of acetylcholine, nitroprusside, and CF were based on weight measurements. All dilutions from stock solutions were made before the experiment. Buffer was used as vehicle control.
Aortic ring preparation and vascular contractility.
The aortic function was measured as described (29). The lumen diameter was measured by a micrometer. The wall thickness was measured by putting micrometer edges on both sides of the aortic wall. The wall stress was measured as follows: (SBP · radius2/wall thickness)/2(radius + wall thickness). The ring was aerated with 95% O2 and 5% CO2 (pH 7.4) and equilibrated at 37°C continuously. Aortic rings were perfused with physiological salt solution containing (in mM) 131.5 NaCl, 0.2 KCl, 1.2 NaH2PO4, 1.2 MgCl2, 0.5 CaCl2, 23.5 NaHCO3, and 11.2 glucose. The rings were mounted in between two stainless steel wires, one connected to a force transducer and the other connected to a micrometer, in an isometric myograph (World Precision Instrument). The signal from the ring under experimentation was digitalized by on-line analysis using CVMS software (World Precision Instrument). To evaluate the viability of aortic ring, we contracted the ring three times by inducing active muscle tone using 20 mM CaCl2, rinsed, and reequilibrated before experiment. To accommodate for different sizes and lengths of different rings, we normalized the generated tension in grams with the weight of the tissue in grams (38). Precontracted rings with CaCl2 were relaxed by endothelium-dependent acetylcholine and endothelium-independent nitroprusside. The optimum dose, an effective concentration (EC50), was measured by a dose-dependence curve between amplitude of contraction and added concentrations of vasoactive agents. We measured the levels of PPAR, nitrotyrosine, and MMP activity in aortic segments after function measurements and observed similar results before and after.
Values are given as means ± SE from n = 6 in each group. Differences between groups were evaluated by using ANOVA, followed by the Bonferroni post hoc test (42), focusing on the effects of −/+ CBS (+/+ mice to −/+ mice) and treatment (−/+ mice treated with CF compared with −/+ mice). P < 0.05 was considered significant.
Genotype/phenotype of CBS mice.
PCR analysis of genomic DNA revealed one PCR product of 1.5 kb of a normal allele and two products of normal and homologous disrupted alleles with different electrophoretic mobility in −/+ alleles. There were two groups of mice: one with Hcy levels between 3 and 6 μM and the other with Hcy levels between 8 and 15 μM. On the basis of genotype and the levels of Hcy, mice were divided into wild-type (+/+) littermates and heterozygous (−/+) CBS mice, respectively.
Morphology and in situ MMP activity.
Histological analysis revealed increased collagen and decreased elastin in aortic media of CBS −/+ mice compared with wild-type mice (Fig. 1). There was significant MMP activity in the aortas of −/+ compared with +/+ mice. In +/+ mice, only adventitia regions of the aortas showed high MMP activity, whereas in −/+ mice, the adventitia, media, and lumen showed high MMP activity (Fig. 2).
BP in CBS −/+ mice.
CBS −/+ mice demonstrated an increase in protein excretion in urine compared with wild-type +/+ mice. Treatment with CF decreased the levels of MUP in CBS −/+ mice to control levels (Table 1).
The levels of plasma Hcy in CBS −/+ mice did not decrease to the control levels after CF treatment (Table 1).
Levels of PPAR.
To determine whether CF had classical biological effect, such as liver proliferation, we measured liver wt/body wt ratios. The results suggested reduced liver size in CBS −/+ mice. Treatment with CF increased liver wt-to-body wt ratios (Table 1). CF tended to induce both PPAR-α and -γ (Fig. 3).
CF inhibits nitrotyrosine generation.
The levels of nitrotyrosine were increased in CBS −/+ mice compared with wild-type +/+ littermates (P < 0.01). Treatment with CF inhibited the generation of nitrotyrosine in both the −/+ as well as wild-type +/+ controls (Fig.4).
Activation of MMP-2 and -9 in−/+ CBS mice.
Zymographic analysis revealed increased MMP-9 activity in CBS mice. Treatment with CF completely inhibited the MMP-9 and decreased MMP-2 activity by 50% in −/+ as well as wild-type +/+ control mice (Fig.5).
Improvement of aortic function by CF.
Aortic response to acetylcholine was attenuated in CBS −/+ mice compared with wild-type +/+ mice. The dose-response curves were shifted to the left in CBS −/+ mice after CF treatment compared with CBS −/+ mice (Fig. 6). These results suggest that treatment with CF improved vascular function in CBS −/+ mice.
A linear relationship between SBP and plasma levels of Hcy has been suggested in elderly patients (41). Some studies have also linked PPAR to hypertension (1, 10) and suggested that peroxisome proliferator fibrates may improve hemodynamic (34) and renal parameters (14, 15). We suggest that in the absence of PPAR response, reactive oxygen species generated by Hcy, in part, induced hypertension by increasing vessel wall stress in CBS −/+ mice. The administration of Hcy in rats increased BP (26). The changes in BP by Hcy were independent of amount of Hcy but dependent on the duration, suggesting structural changes by Hcy (26). Histological analysis revealed that aortic intima-media was thickened in the CBS −/+ mice compared with controls (Table 1 and Fig. 1). Similar results were obtained by others (9). The thickening of aortas was associated with increased MMP activity (Fig. 2). Others have shown that CF inhibits the synthesis of the ECM component fibrinogen (19) and that the agonist of PPAR causes regression of intima-media thickness in humans (28). The ECM, particularly elastin (50% of the normal vessel wall), surrounding the vascular smooth muscle cell enhances muscle cell compliance with load. Elastin is a good substrate for MMP-2 and -9 (38). To reduce vessel wall stress, especially in the absence of endothelial NO and HHcy, the MMP is activated to dilate the vessel by remodeling and degrading elastin. Because elastin turnover is remarkably lower than collagen (36), the degraded elastin is replaced by stiffer collagen, and the Hcy induces collagen synthesis (44). Therefore, vessel wall stress is increased (Table 1). Previously we quantified the decreased elastin and increased collagen in homocysteinemic aortas (29). Here we demonstrated increased collagen and decreased elastin in aortas of HHcy mice.
Treatment with peroxisome proliferator decreased lipid profile (35). A number of studies have demonstrated that the levels of Hcy in fasting and loading conditions are not changed after fibrate therapy. In fact, some studies reported a slight increase (3). Here is the paradox: drugs that repair vascular endothelial function may increase Hcy accumulation. Our results suggest that plasma Hcy levels were not altered by fibrate treatment in CBS −/+ mice (Table 1). In addition, the levels of PPAR were increased after fibrate treatment (Fig. 3), suggesting that CF induced PPAR expression.
Hcy induced multiorgan damage (26) and increased the MUP in CBS −/+ mice (Table 1). CF treatment reduced the injury response by Hcy (Table 1). Others have shown that treatment with fibrate decreases renovascular resistance and hypertension (15). The mouse model of abrogation of endothelial NO generation has suggested an increase in BP (39). Hcy decreased bioavailability of endothelial NO. The increase in Hcy in CBS −/+ mice was associated with the increase in SBP (Table 1). The treatment with Hcy induced both nitrotyrosine and MMP-2 and -9 activity (29). Treatment with nicotinamide, an inhibitor of poly(ADP-ribose)synthetase, an enzyme that can be activated by oxidants and peroxynitrite (4), inhibited the Hcy-mediated nitrotyrosine generation and MMP activation (29). Nicotinamide also ameliorated Hcy-mediated vascular dysfunction (29). Here we report that the activation of MMP and treatment with CF decreased nitrotyrosine levels and MMP activity, respectively (Figs. 4 and 5).
Numerous studies have demonstrated impairment of vasodilatory response of aortas after acute treatment with Hcy. In CBS −/+ mice, the response to acetylcholine was attenuated (9). Folic acid regressed the Hcy-mediated vascular dysfunction in CBS −/+ mice (22). CF has been shown to improve the oxidative stress-mediated vascular dysfunction (12). Our results suggest that CF reversed Hcy-mediated vascular dysfunction in CBS −/+ mice (Fig. 6), in part by decreasing nitrotyrosine and MMP activity. Under normal conditions, NO binds to the metal ion and keeps the MMP in latent form. However, in HHcy, NO favors peroxynitrite formation and nitrates the neighboring tyrosine residues. Consequently, the levels of peroxynitrite were increased and generated nitrotyrosine. Also it is known that Hcy induced endothelial nitric oxide synthase as well as superoxide (2). It is a paradox that Hcy may increase both NO and nitrotyrosine. However, this can be explained by the scenario that, during HHcy, Hcy may promote NO generation. The increased oxidative stress may favor nitration and promote nitrotyrosine. The decrease in nitrotyrosine in CF-treated mice may be due, in part, to the fact that PPAR increased antioxidant enzymes such as SOD/catalase and decreases NADPH oxidase (18). Collectively, these studies suggest a role of both the superoxide and NO in generation of nitrotyrosine. Although we have not established a cause and effect relationship between PPAR expression and MMP activity, previous studies by us and others have suggested decreased MMP activity after CF treatment (24, 30). The results of this study suggest that Hcy downregulated PPAR-α, thereby causing oxidation of NO to nitrotyrosine with consequent increased activation of MMP. MMP-induced changes in vascular wall structure, in turn, may alter wall stress and hemodynamics.
The goal of this study was to test whether the high MMP and nitrotyrosine arterial levels in HHcy animals are secondary to Hcy-induced downregulation of PPAR-α. We have shown that CF, a PPAR-α agonist, increased PPAR expression in HHcy mouse aorta while reducing aortic nitrotyrosine and MMP levels. However, MMP and nitrotyrosine levels were also downregulated by CF in wild-type mice, a phenotype in which CF did not upregulate PPAR expression. We speculated that this phenotype may be due to a threshold level of PPAR in normal mice. It is possible that CF had other effects on vascular muscle, perhaps inducing transcription of other enzymes. Indeed, other effects of CF are likely another possible explanation for the observation that CF downregulated MMP activity and nitrotyrosine levels but not PPAR expression, especially in wild-type mice. There was a basal level of nitrotyrosine in the aorta of wild-type +/+ mice. However, it was lower than in −/+ mice (Fig. 4). Similar levels have been reported by others (9). Treatment with CF decreased nitrotyrosine levels in CBS −/+ mice. Collectively, these studies suggest that Hcy interferes with NO signaling via the cGMP pathway, which may impact remodeling, structure, and function in the vessel wall.
This work was supported in part by National Institutes of Health Grants GM-48595 and HL-71010 and by the Kidney Care Foundation.
Address for reprint requests and other correspondence: S. C. Tyagi, Dept. of Physiology and Biophysics, Univ. of Mississippi Medical Center, 2500 No. State St., Jackson, MS 39216-4505 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published October 25, 2002;10.1152/ajplung.00183.2002
- Copyright © 2003 the American Physiological Society