|Home | About | Journals | Submit | Contact Us | Français|
Folate supplementation improves endothelial function in patients with hyperhomocysteinemia. Mechanistic insights into potential benefits of folate on vascular function in general population however, remain mysterious.
Expression of dihydrofolate reductase (DHFR) was markedly increased by folic acid (FA, 50 µmol/L, 24 hr) treatment in endothelial cells. Tetrahydrofolate (THF) is formed after incubation of purified DHFR or cellular extracts with 50 µmol/L of substrate dihydrofolic acid. THF could then be detected and quantified by high performance liquid chromatography (HPLC) with a fluorescent detector (295/365 nm). Using this novel and sensitive assay, we found that DHFR activity was significantly increased by FA. Furthermore, FA improved redox status of Ang II treated cells by increasing H4B and NO• bioavailability while decreasing superoxide (O2•−) production. It however failed to restore NO• levels in DHFR siRNA-transfected or methotrexate pre-treated cells, implicating a specific and intermediate role of DHFR. In mice orally administrated with FA (15 mg/kg/day, 16 days), endothelium upregulation of DHFR expression and activity occurred in correspondence to improved NO• and H4B bioavailability, and this was highly effective in reducing Ang II infusion (0.7 mg/kg/day, 14 days)-stimulated aortic O2•− production. 5’-methyltetrahydrofolate (5’-MTHF levels, GTPCH1 expression and activity remained unchanged in response to FA or Ang II treatment in vitro and in vivo.
FA supplementation improves endothelial NO• bioavailability via upregulation of DHFR expression and activity, and protects endothelial cells from Ang II-provoked oxidant stress both in vitro and in vivo. These observations likely represent a novel mechanism (intermediate role of DHFR) whereby FA induces vascular protection.
The most widely known therapeutic function of folic acid (FA) is to prevent birth defect (primarily neural tube defects), based on its role in neonatal development [1, 2]. Its other major therapeutic function is to treat patients with hyperhomocysteinemia . FA is required for the remethylation of homocysteine to form methionine, thus reducing the homocysteine level in the plasma. Oral supplementation of FA effectively decreases plasma homocysteine levels . Both hyperhomocysteinemia and hyperhomocysteinuria have been shown to be independent risk factors for atherosclerotic vascular diseases [4–10]. One of the molecular mechanisms responsible for hyperhomocysteinemia-induced disease is increased production of reactive oxygen species and consequent endothelial dysfunction [11, 12]. It is however unclear whether FA, besides reducing homocysteine levels metabolically, modulates endothelial function directly in subjects with or without hyperhomocysteinemia. It is also unclear whether FA treatment of hyperhomocysteinemic patients is beneficial in disease progression and prognosis.
As a FDA-approved agent, FA or combined vitamin therapy has been investigated for its role in cardiovascular therapeutics in various clinical conditions, and the outcomes (myocardial infarction, stroke, thromboembolic events and mortality) have remained controversial [5, 13–15]. However, it is worth noting that the larger trials (i.e. HOST , VISP , HOPE-2 ) were conducted in subjects with advanced atherosclerotic vascular diseases where disease regression might be more difficult. Even though, a significant reduction was observed in stroke morbidity with FA treatment after removal of the HOST trial [14, 18] (subjects with renal failure) based on meta-analysis [5, 19–21]. Therefore, FA supplementation can potentially benefit patients with vascular diseases, although the role of homocysteine during the course of interventions is unclear . More interestingly, FA supplementation was demonstrated to improve endothelial function in subjects without elevated homocysteine levels, indicating an independent effect . Whether this is attributed to direct nitric oxide (NO•) production and/or reduced oxidant stress, as well as the potential underlying mechanisms, remains to be fully elucidated.
Accumulating evidence has established that a deficiency in endothelial nitric oxide synthase (eNOS) cofactor tetrahydrobiopterin (H4B) causes eNOS to produce superoxide (O2•−) rather than NO•, resulting in eNOS uncoupling and increased oxidant stress [23, 24]. This change in eNOS enzymatic function is independent of gene regulation. Under pathological conditions such as diabetes and ischemic renal dysfunction where angiotensin II (Ang II) levels are elevated , we and others have shown that a deficiency in H4B salvage enzyme dihydrofolate reductase (DHFR) is responsible for reduced H4B and NO• bioavailability [23, 25–27]. Therefore in the present study we examined FA regulation of DHFR protein expression and activity, and DHFR-mediated changes in NO• and O2•− productions in cultured endothelial cells and Ang II-infused mice.
A sensitive HPLC-based activity assay for DHFR was first established. Endothelial cells were exposed to Ang II in the presence or absence of FA, followed by detection of NO• and O2•− production specifically and quantitatively using electron spin resonance (ESR). Basal FA regulation of DHFR expression and activity were also analyzed. In Ang II-infused mice, oral administration of FA was examined for its effects on aortic productions of NO• and O2•− Overall FA potently improved NO• bioavailability, intracellular H4B content while reducing oxidant stress both in vitro and in vivo. These responses were found dependent on an upregulation in DHFR. These data demonstrate an innovative mechanism whereby FA may protect against cardiovascular disorders in a general population.
Monoclonal antibodies for DHFR and eNOS were purchased from Research Diagnostics (Flanders, NJ) and BD Transduction Laboratories (San Jose, CA) respectively. Tetrahydrofolate, FA, dihydrofolic acid, recombinant DHFR, NADPH and all other reagents were purchased from Sigma-Aldrich in highest purity (St. Louis, USA).
Bovine aortic endothelial cells (BAECs, Cell Systems, Kirkland, WA) were grown in media 199 containing 10% fetal bovine serum (FBS) until confluence, and quiescent with 5% FBS media. Confluent cells were exposed to FA (50 µmol/L) or Ang II (100 nmol/L) or combination of these two for 24 hr, then rinsed with modified Kreb’s/HEPEs buffer, incubated with freshly prepared NO•-specific spin trap Fe2+(DETC)2 colloid (0.5 mmol/L) for 60 min. Gently collected suspensions of cells were snap-frozen in liquid nitrogen and loaded into a finger Dewar for analysis with an eScan ESR spectrophotometer (Bruker) at the following settings: center field, 3410; field sweep, 100 G; microwave frequency, 9.76 GHz; microwave power, 13.26 mW; modulation amplitude, 9.82 G; 10240 points of resolution; and receiver gain, 356. The aortic NO• production was measured as previously described .
Cells were lysed with lysis buffer (0.2 mol/L Tris, 1.5 mol/L NaCl, 10 mmol/L EDTA, 10 mmol/L EGTA, 25 mmol/L sodium pyrophosphate, 10 mmol/L β-glycerophosphate, 10 mmol/L Na3VO4, 1 mmol/L PMSF, 2 µmol/L leupeptin and 10% Triton, pH 7.4) for 1hr. The samples were then prepared for Western analysis. Forty µg of protein was separated by 10% SDS PAGE, transferred to nitrocellulose membranes, and probed with DHFR (1:1000) or eNOS (1:1000) antibody following standard immunoblotting procedure.
Endothelial cells were lysed using trichloroacetic acid containing 10 mmol/L DTT. Lysates were subjected to differential oxidation in acidic (0.2 mol/L trichloroacetic acid containing 2.5% I2/10% KI) or alkalytic (0.1 mol/L NaOH containing 0.9% I2/1.8% KI) solutions as briefly decribed previously . After centrifugation, 20 µl of supernatant was injected into a HPLC system. Excitation and emission wavelengths of 350 nm and 450 nm were used to detect fluorescent H4B and its oxidized species.
Cellular GTPCH1 mRNA was extracted using RNeasy kit from Qiagen (Huntsville, AL) according to manufacturer instruction. The cDNA was synthesized with script III from Invitrogen (Carlsbad CA) and Real-time PCR was performed on a Bio-Rad iCYCLER iQ real-time PCR machine as previously described .
Aortic segments were powdered under liquid nitrogen using mortar and pestle. To access GTPCH1 activity, tissue powders or cell pellets were lysed and incubated with GTP (0.75 mmol/L final concentration) at 37°C for 90 min in the dark. The reaction product was oxidized to neopterin triphosphate by acidic iodine (1% I2 and 2% KI in 2 mol/L TCA). After reduction of excessive iodine by ascorbic acid (2%) and adjustment of pH to 7.0, reaction products were monitored by HPLC with fluorescent detection (365/446 nm). The detailed chromatographic procedure has been described elsewhere .
Tissue powders or cell pellets were dissolved/re-suspended in ascorbic acid (57 mmol/L) solution. Samples were repeatedly frozen and thawed or sonicated to break cell membranes and then vortexed and incubated at 37°C for 60 min. Samples were diluted with buffer solution (0.2 mol/L potassium phosphate dibasic and 30 mmol/L mercaptoethanol, pH 8.5) and heated at 100°C for 10 min. After being vortexed for 30 sec and centrifuged at 10,000 g for 15 min, the supernatants were collected and analyzed using HPLC as previously described . Briefly, the 5’-MTHF signals were monitored by fluorescent detection (295/360 nm) under the mobile phase of methanol:0.6% acetic acid in water (14:86, v/v) at a flow-rate of 1 ml /min. The quantitative levels of 5’-MTHF were calculated based on a calibration curve established using known concentrations of 5’-MTHF going through the same preparation and detection procedures.
The C57BL/6 mice from Jackson Laboratories (6–9 weeks old, total 64) were separated into four groups. Two were put on standard mice chow (Harlan) and two on FA rich (15 mg/kg/day) chow. On average, mice weighed 25 g and consumed 3.5 g of diet daily. To prepare FA rich chow, standard mouse chow was grounded, mixed well with FA, and food tablets produced by a customerized press machine. All mice were fed ad libitum and tablets with FA were replaced every two days. Half of mice on each diet were then anesthetized and osmotic pumps (Durect corporation) containing angiotensin II (0.7 mg/kg/day, Sigma) were implanted subcutaneously after anesthetization. Two weeks later mice aortas were isolated and sectioned. Sections from well-defined regions, 1 mm below subclavian artery, were used for ESR analysis of O2•− production. The rest of the aorta was used for ESR measurement of NO• and western analysis of eNOS and DHFR expression. The procedures were carried out based on protocols approved by the UCLA Institutional Animal Care and Use Committee (IACUC).
As previously decribed , aortic segments (~3 mm) were mixed with freshly prepared spin trap (CMH, 0.5 mmol/L) in the presence or absence of 100 units/ml polyethylene glycol-conjugated superoxide dismutase (PEG-SOD), and loaded to glass capillarries for immediate kinetic analysis of O2•− production. The ESR settings were as the followings: biofield, 3494.50; field sweep, 9 G; microwave frequency, 9.75 GHz; microwave power, 21.02 mW; modulation amplitude, 2.47 G; 4,096 points of resolution; receiver gain, 1000; and kinetic time, 10 min. The PEG-SOD inhibitable O2•− signals at 10 min time point, normalized by protein concentrations, were compared among different experimental groups.
To determine changes in DHFR expression and activity in the mouse endothleium, an additional group of 32 mice were studied as described above. Upon harvest, whole aorta was cut open longitudinally and digested with collagenase (0.6 mg/ml) for 20 min at 37°C. The supernatant was centrifuged and the endothelial cells were collected from the pellet. Cells were lysed and forty µg of protein separated by 10% SDS/PAGE, transferred to nitrocellulose membranes, and probed with DHFR (1:250) or eNOS (1:1,000) antibody respectively. The endothelium-denuded vessels were also homogenated and subjected to Western analysis of eNOS and DHFR following identical procedure.
Data are presented as Means±SEM from independent experiments. Differences between control and treatment groups were compared by ANOVA and considered significant when p<0.05.
Post confluent bovine aortic endothelial cells were treated with folic acid (FA, 50 µmol/L, 24 hr) prior to Western analysis of DHFR and eNOS protein expressions. As shown by represent blot and grouped densitometric data, FA supplementation resulted in a significant upregulation of DHFR while not affecting eNOS expression (Figs. 1A and 1B). Of note, endothelial cell production of NO• was also increased (Figs. 1C and 1D).
In order to determine specifically DHFR activity from cells, a sensitive HPLC-based assay was established. Purified recombinant DHFR or cell lysates were incubated with dihydrofolate (50 µmol/L) and NADPH (200 µmol/L) for 20 min at 37°C, in a 0.1 mol/L potassium phosphate assay buffer (pH 7.4) containing 1 mmol/L DTT, 0.5 mmol/L KCl, 1 mmol/L EDTA and 20 mmol/L sodium ascorbate. At the end of the incubation, the reaction was terminated by addition of 0.2 mol/L trichloroacetic acid, and then product was stabilized by addition of stabilization solution (200 mg of sodium ascorbate and 30 mg of DTT in 1 ml of water, 1:10 dilution for the final working solution).
The product, tetrahydrofolate/tetrahydrofolic acid (THF), was measured using a Shimadzu HPLC system consisted of C-18 column (250- 4.6-mm C18 column Alltech, Deerfield) and fluorescent detector, at wavelengths of 295 nm for excitation and 365 nm for emission. The mobile phase was isocratic flow 1 ml/min of mixture of 7% acetonitrile with 5 mmol/L KH2PO4; pH 2.3. The retention time of tetrahydrofolic acid was at about 7.5 minute. The specific THF peaks, under different assay conditions, were highlighted in dotted frame in Fig. 2 tracing F. In Fig. 2, tracing A shows a standard peak of THF (50 nmol/L) in assay buffer. Tracing B represents the reaction containing only substrate dihydrofolic acid, indicating that THF peak was unique, without overlapping with the dihydrofolic acid peak. Tracing C & E demonstrate that when NADPH or substrate was missing from the reaction, no THF was produced. Tracing D indicates that an inactive enzyme (boiled to inactivate) also failed to generate a specific THF peak. Only tracing F shows a characteristic THF peak when recombinant enzyme was mixed with proper concentrations of dihydrofolic acid and NADPH in the right assay buffer. The constituents of each tracing are shown on the right panel of Fig. 2. The THF content was calculated against a standard curve prepared by using THF solutions in assay buffer. Data are presented as nmol production of THF.
As shown in Fig. 3A, THF formation was linearly correlated with time within 20 min incubation. A dose-dependent effect of electron donor NADPH (maximized at >200 µmol/L) on THF formation was observed (Fig. 3B). In addition, formation of THF was linearly correlated with protein content either recombinant protein or BAEC lysates (Figs. 3C&D). Using purified DHFR and dihydrofolic acid as the substrate, formation of THF at different pH was compared and pH 7.4–7.5 was found to be the optimal for the reaction (Fig. 3E). The Km of the enzyme was calculated to be 4.5±0.8 µmol/L and the Vmax for the enzyme to be 999.7±74.2 nmol/min/µg protein (Fig. 3F). Folic acid, when serving as a much weaker substrate, showed a Km of 28.8±5.3 µmol/L and the Vmax for the enzyme is 2.5±0.2 nmol/min/µg protein.
The validation parameters were specificity, linearity, extraction recovery, precision and accuracy. Blank cell lysates and pure tetrahydrofolate were screened to determine the specificity (see Fig. 2). Linearity was tested on three different sets of calibration curves. The lowest concentration of the linear regression defined the lower limit of qualification (LLOQ). The LLOQ for tetrahydrobiopterin was 25 nmol/L at a signal-to-noise (S/N) ratio of 5 (highest sensitivity, ×16 gain for HPLC setting).
The extraction recovery of tetrahydfolate was assessed at different concentrations of tetrahydrofolate, and calculated by comparing the peak area of tetrahydrofolate in assay buffer containing cell lysates to the peak area obtained with the same concentration of tetrahydrofolate in assay buffer. The mean recovery rate of tetrahydrobiopterin at low, middle and high concentrations (30, 50, 75 nmol/L) was greater than 95% (Table 1). The mean regression equation weighed by concentration (nM) was y = (331400 ± 15010)x − (1126000 ± 853400) (r2=0.9959).
Intra-day assay accuracy and precision of quality controls were measured using three determinations per concentration at the same day. Inter-day assay accuracy and precision of quality control were measured using three determinations per concentration on three consecutive days. The data of accuracy and precision are represented by relative standard deviation (R.S.D.) and relative error (RE) as shown in Table 2.
As shown in Figure 4A, Ang II decreased DHFR expression in endothelial cells which is consistent with our previous findings . FA supplementation not only increased DHFR expression but also preserved DHFR expression from being attenuated by Ang II (Fig. 4A). Similarly, DHFR activity was downregualted by Ang II but upregulated/restored by FA (Fig. 4B). The upregulation in DHFR expression and activity may increase H4B level via augmented salvation of H2B. Indeed, the ratio of oxidized H4B (H2B and other intermediates) over total pterins was decreased in FA treated cells. Moreover, FA treatment prevented H4B from being oxidized by Ang II (Figs. 4C and 4D). Meanwhile, bioavailable NO• was decreased by Ang II treatment, but markedly improved by FA pre-treatment (Fig. 4E).
In additional experiments, endothelial cells were transfected with control and DHFR siRNA prior to FA stimulation. As shown in Fig. 5A, DHFR siRNA efficiently attenuated DHFR protein expression. Although FA increased NO• production in control siRNA transfected cells, it failed to provoke any changes in NO• level in DHFR siRNA transfected cells, indicating a specific and intermediate role of DHFR in FA induced augmentation of NO• production (Fig. 5B). Moreover, pre-inhibition of DHFR with methotrexate (1 µmol/L) also attenuated the beneficial effects of FA on endothelial NO• bioavailability following Ang II stimulation (Fig. 5C). GTPCH1, the rate limiting enzyme for H4B de novo synthetic pathway were also examined for its expression and activity to further determine whether DHFR has an independent role in regulating H4B availability. As shown in Fig. 5D and 5E, both GTPCH1 mRNA level and activity were not changed by FA treatment. 5’-methyltetrahydrofolate (5’-MTHF), a bioactive form of FA which may have similar benefits on eNOS function , also remained unchanged with FA treatment.
To examine whether FA would exert beneficial effects on vascular endothelial function in vivo, control or Ang II infused mice were on FA-rich diet for 2 weeks. Aortas were then harvested and a 2 mm section of descending aorta was subjected to O2•− detection immediately using ESR. As shown by representative kinetic curves, and quantitative data for O2•− production at the end of 10 min, Ang II-dependent O2•− was completely attenuated by FA treatment (Figs. 6A&B). After O2•− measurement, the rest of the aortas were subjected to NO• detection by ESR. FA alone increased NO• production by 40% in control mice (Figs. 6C&D). Whereas aortic NO• production was markedly reduced by Ang II infusion, it was restored to near control level by oral administration with FA (Figs. 6C&D).
In additional experiments, DHFR expression was examined using intact or endothelium-denuded aortas. Of note, eNOS was completely absent in the denuded vessels (Fig. 7A) whereas its expression was abundant in the endothelial cells but downregulated by Ang II (Fig. 7A). Oral administration of FA significantly improved endothelial DHFR expression in untreated control mice, as shown by both representative blots and quantitative data (Figs. 7B&C), which is similar to what is observed in vitro (Fig. 1). In Ang II infused animals, endothelial DHFR expression was diminished, which was however significantly restored by FA administration (Figure 7B&C). Again, this beneficial effect of FA treatment was independent of GTPCH1 activity and intracellular 5’-MTHF level (Figs. 7D–F). These data seem to suggest that the beneficial effects of FA on vascular NO• bioavailability in Ang II infused mice are mediated by an upregulation of DHFR function.
The most significant finding of the study is that folic acid (FA) directly improves H4B and NO• bioavailability while reducing oxidant production both in vitro and vivo, and that these effects are mediated by an upregulation of dihydrofolate reductase (DHFR) expression and activity. FA treatment of control or Ang II stimulated endothelial cells increased NO• production, which was attenuated by DHFR siRNA or methotrexate. In Ang II-infused mice, aortic O2•− production was attenuated by FA administration, whereas NO• levels were elevated. Endothelial expression of DHFR was markedly increased by FA, which is corresponding to a similar increase in activity.
As shown in Fig. 1, FA upregulates DHFR expression and NO• production in healthy endothelial cells at baseline. This seems consistent to what we found earlier that endogenous DHFR function is important for basal NO• production . It is thus interesting to speculate that a deficiency in FA, i. e. caused by low dietary supplementation, may induce a DHFR deficiency that predisposes to vascular diseases. In consequent efforts to establish a sensitive and reliable HPLC-based activity assay for DHFR, FA was compared to dihydrofolic acid for substrate capability and found a second titer substrate with much less efficiency (Km 7-fold higher, Vmax 50-fold less). Nonetheless, it seems to implicate that when FA dietary supplementation is sufficient, DHFR activity would be important for its metabolism. Indeed, data from the Framingham Offspring Study demonstrated that a 19-bp polymorphism at intron 1 of DHFR gene is associated with increased plasma unmetabolized FA and decreased red blood cells folates .
Importantly, we developed a novel and sensitive HPLC-fluorescent based assay to measure DHFR activity. Using cell or tissue lysates, an optimal condition has been selected for maximal detection sensitivity: substrate dihydrofolate at 50 µmol/L, electron donor NADPH at 200 µmol/L, incubation for 20 min at 37°C, in a 0.1 mol/L potassium phosphate assay buffer (pH 7.4). The stabilization buffer maintains product (tetrahydrofolate, THF) stability at least overnight at 4°C to allow autosampling performance. The HPLC-coupled electronic detection (ECD) method developed by Aiso et al  has similar sensitivity limit to our assay; but is associated with disadvantages such as handling of radioactivity, cost consumption, and complexity of the assay procedure. DHFR activity has been assayed at various pH and/or ambient temperature ranging from 21 to 28°C [34, 35]. In this study, however, we assayed it at pH 7.4 and 37°C, which is more physiologically relevant. When dihydrofolic acid was used as a substrate, the reaction was not affected by pH (Fig. 3E), but significantly more sensitive at 37°C than at ambient temperature. Moreover, the product THF was stable in our assay system even after twenty min incubation at 37°C, and with the proper stabilizing buffer, can be analyzed by autosampler at 4°C overnight.
Elevated angiotensin II (Ang II) levels have been observed in many cardiovascular disorders where endothelial dysfunction occurs . It is an independent risk factor for atherosclerosis and its antagonists have been used clinically to treat patients with hypertension. Thus it would be a good model system to test the effects of FA on endothelial function, which has already been implicated in coronary patients [22, 37, 38] and patients with hyperhomocysteinemia [39, 40]. As expected, Ang II stimulation led to increased O2•− production and decreased NO• and H4B bioavailability. These responses were however all reversed significantly by FA pre-treatment. In addition, DHFR expression and activity were increased by FA. Pre-inhibition of DHFR by RNA interference or Methotrexate effectively abolished the beneficial effects of FA on endothelial cell functions. Besides these clear-cutting in-vitro observations, data collected from Ang II-infused mice also supported a strong impact of FA and an intermediate role of DHFR.
Mice infused with Ang II had markedly increased O2•− production and reduced NO• bioavailability. This was accompanied by an endothelial downregulation of DHFR. Oral administration of FA significantly improved NO• production, while reducing O2•− levels. It also reserved DHFR protein abundance from being diminished by Ang II. Changes in DHFR activity mirrored expressional regulations by Ang II and FA. Taken together, these data demonstrate that FA exerts it beneficial effects in vivo via regulation of DHFR expression and activity. It is important to note that in the endothelium-denuded aortas DHFR expression was upregulated by Ang II, indicating differential regulation of DHFR in different cell types. This seems logic as Ang II is known to stimulate vascular smooth muscle hypertrophy, and that DHFR in general is important for DNA synthesis and cell growth.
In summary, new data presented in the present study for the first time established that FA, via modulation of DHFR expression and activity, improves NO• bioavailability while reducing O2•− production under pathophysiological conditions. These findings may represent a common mechanism whereby FA supplementation exerts beneficial effects on cardiovascular system in populations with and without hyperhomocysteinemia.
The authors’ work has been supported by National Heart, Lung and Blood Institute (NHLBI) Grant HL077440 (HC), HL057244 (PLL and HC), HL080111 (PPP and HC), an American Diabetes Association Award 7-08-RA-23 (HC), and a Start-up Fund from the University of California Los Angeles (HC).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.