Tetrahydrobiopterin, L-Arginine and Vitamin C Act Synergistically To

Home , Tetrahydrobiopterin

Tetrahydrobiopterin, L-Arginine and Vitamin C Act Synergistically to Decrease Oxidative Stress, Increase Nitric Oxide and Improve Blood Flow after Induction of Hindlimb Ischemia in the Rat

Jinglian Yan, Guodong Tie, and Louis M Messina

Department of Surgery, University of Massachusetts Medical School, Worcester, Massachusetts, United States of America

Nitric oxide (NO) derived from endothelial nitric oxide synthase (eNOS) is a potent vasodilator and signaling molecule that plays an essential role in vascular remodeling of collateral arteries and perfusion recovery in response to hindlimb ischemia. In ischemic conditions, decreased NO bioavailability was observed because of increased oxidative stress, decreased L-arginine and tetrahydrobiopterin. This study tested the hypothesis that dietary cosupplementation with tetrahydrobiopterin (BH4), L-arginine, and vitamin C acts synergistically to decrease oxidative stress, increase nitric oxide and improve blood flow in response to acute hindlimb ischemia. Rats were fed normal chow, chow supplemented with BH4 or L-arginine (alone or in combination) or chow supplemented with BH4 + L-arginine + vitamin C for 1 wk before induction of unilateral hindlimb ischemia. Cosupplementation with BH4 + 2+ L-arginine resulted in greater eNOS expression, Ca -dependent NOS activity and NO concentration in gastrocnemius from the ischemic hindlimb, as well as greater recovery of foot perfusion and more collateral artery enlargement than did rats receiving either agent separately. The addition of vitamin C to the BH4 + L-arginine regimen did further increase these dependent variables, although only the increase in eNOS expression reached statistical significances. In addition, rats given all three supplements demonstrated significantly less Ca2+-independent activity, less nitrotyrosine accumulation, greater glutathione:glutathione disulfide (GSH:GSSG) ratio and less gastrocnemius muscle necrosis, on both macroscopic and microscopic levels. In conclusion, cosupplementation with BH4 + L-arginine + vitamin C significantly increased vascular perfusion after hindlimb ischemia by increasing eNOS activity and reducing oxidative stress and tissue necrosis. Oral cosupplementation of L-arginine, BH4 and vitamin C holds promise as a biological therapy to induce collateral artery enlargement. Online address: http://www.molmed.org doi: 10.2119/molmed.2011.00103

INTRODUCTION loss of eNOS-derived NO production At least two characteristics of eNOS Endothelial dysfunction is a hallmark (that is, functional inactivation of eNOS) render it susceptible to oxidative stress. of peripheral artery disease (PAD) (1). and inactivation of NO by a superoxide First, eNOS transcription, posttransla- – Central to the development of endothelial anion (O2 ) to form peroxynitrite tional modification and trafficking to the dysfunction, regardless of its cause, is a (OONO–) (2,3). It is likely that all three caveolae are attenuated by the accumula- reduction in the bioavailability of nitric mechanisms contribute to the endothelial tion of reactive oxygen species (ROS) oxide (NO) derived from endothelial ni- dysfunction characteristic of PAD because within the endothelial cell (4). Second, the tric oxide synthase (eNOS). Three funda- increased oxidative stress is a common eNOS cofactor tetrahydrobiopterin (BH4) mental mechanisms can compromise NO antecedent in the pathogenesis of this dis- is highly susceptible to oxidation (5). BH4 bioavailability: loss of eNOS expression, ease and can reduce NO bioavailability. maintains eNOS in its functional dimeric form; in the absence of BH4, eNOS be- comes uncoupled so that the electron flux is diverted away from the L-arginine bind- Address correspondence to Louis M Messina, Department of Surgery, University of Massa- ing site and instead reduces molecular chusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655. Phone: 508-856- – oxygen, generating O2 (6). This circum- 5599; Fax: 508-856-8329; E-mail: [email protected]. stance initiates a vicious cycle, wherein Submitted March 22, 2011; Accepted for publication February 21, 2012; Epub – eNOS catalytic activity produces O2 , not www.molmed.org) ahead of print February 24, 2012. NO, worsening existent oxidative stress. These molecular characteristics of eNOS predict that several therapeutic options might prove effective for the

676 | YAN ET AL. | MOL MED 18:676-684, 2012 RESEARCH ARTICLE

treatment of PAD, namely dietary sup- respectively. This action would enhance Preparation plementation with an antioxidant, or eNOS-derived NO production and, by Severe ischemia was induced in the – with the eNOS substrate L-arginine, or quenching existent O2 , reduce NO inac- left hindlimb. The femoral artery was lig- – with the eNOS cofactor BH4. Vitamin C, tivation by its reaction with O2 . ated between the inguinal ligament and or L-ascorbic acid, is a potent antioxidant In our previous work (19), we observed popliteal fossa, and the ligated section and has been shown to preserve BH4 decreased eNOS expression, decreased and its branches were excised. This pro- levels and enhance endothelial NO pro- bioavailable NO and increased oxidative cedure was carried out under anesthesia duction in vitro (7). ONOO– reacts with stress in hindlimb ischemic rats, and oral with 2% isofluorane. The untreated right BH4 6–10 times faster when in the pres- supplementation of BH4 increased the hindlimb served as an internal control ence of ascorbate. The intermediate prod- beneficial effect of eNOS gene transfer. for each rat. An additional group of uct of the reaction between ONOO– and The goal of this study was to test the hy- sham-operated rats were also used for BH4 is the trihydrobiopterin radical pothesis that combined dietary supple- selected assays. These rats underwent • (BH3 ), which is also reduced back to mentation with BH4, L-arginine and vita- isolation of the femoral artery in the left BH4 by ascorbate. Thus, ascorbate does min C will act synergistically to improve hindlimb under 2% isofluorane anesthe- not protect BH4 from oxidation but hindlimb perfusion recovery and preser- sia, but the artery was left intact. rather recycles the BH3 radical back to vation of tissue integrity in response to BH4 (8). Vitamin C levels are low in PAD severe hindlimb ischemia. To this end, we Study Design patients (9), and acute (10) or short-term generated severe hindlimb ischemia in All rats were fed standard chow in (11) vitamin C supplementation reduces the rat by means of femoral artery exci- powder form and water ad libitum PAD symptoms; however, cross-sectional sion. Measured dependent variables in- (Deans Feeds, Redwood City, CA, USA). epidemiological surveys have failed to cluded gastrocnemius muscle eNOS ex- Animals were randomly selected to re- find a clear link between long-term vita- pression and activity, hindlimb laser ceive normal chow (control), chow with min C intake and PAD symptoms or dis- Doppler perfusion and collateral artery a single added supplement (BH4 or ease progression (12,13). L-arginine sup- enlargement, and gastrocnemius nitroty- L-arginine), chow supplemented with plementation showed exciting promise in rosine accumulation and tissue necrosis. BH4 + L-arginine or chow supplemented short-term studies of PAD (14,15), but with BH4 + L-arginine + L-ascorbic acid. this effect was not observed in a subse- MATERIALS AND METHODS Dietary supplementation was com- quent long-term study by the same group menced 7 d before the induction of (16). BH4 improves eNOS-dependent va- Materials hindlimb ischemia and was continued sodilation in long-term smokers and pa- Dietary supplements included the fol- until sacrifice of the animal. Chow was tients with type 2 diabetes, conditions as- lowing: BH4 (10 mg/kg/d; Schircks Lab- replaced every 2 d. The time of sacrifice sociated with increased oxidative stress oratories, Jonas, Switzerland); L-arginine, varied with the measured endpoint (17,18). To our knowledge, BH4 has not provided as L-arginine α-ketoglutarate under consideration, as described below. been specifically evaluated as a therapeu- (hereafter L-arginine; 88.5 mg/kg/d; tic modality in PAD. Body Tech, North Bergen, NJ, USA); and Western Blotting An important gap in our understand- L-ascorbic acid (that is, vitamin C; 88.5 Rats were sacrificed 14 d after induc- ing of BH4, L-arginine and L-ascorbic mg/kg/d, Sigma-Aldrich, St. Louis, MO, tion of ischemia for measurement of acid in the prevention and treatment of USA). The dose of tetrahydrobiopterin gastrocnemius muscle eNOS, phosphory- PAD is the potential synergistic effect of was selected on the basis of a published lated eNOS (p-eNOS) and nitrotyrosine combined therapy, and there is convinc- report of its use in rats (35). And the dose expression. The timing of death was se- ing evidence to suggest that such an ap- of L-arginine and vitamin C was on the lected on the basis of previous work that proach would prove successful. For ex- basis of their clinical dose in patients. demonstrated maximal postischemic ample, supplementation with L-arginine change in these variables at this time (19). alone might prove deleterious in the face Animals Samples were homogenized in liquid ni- of endothelial oxidative stress inasmuch All protocols were approved by the trogen and transferred to NP-40 lysis as the resultant increase in eNOS cat- Institutional Animal Care and Use Com- buffer, comprised of 50 mmol/L HEPES – alytic activity might generate O2 , not mittee at the University of California, San (pH 7.5), 150 mmol/L NaCl, 10% glycerol,

NO, if BH4 levels were reduced by oxi- Francisco, and University of Massachu- 1.5 mmol/L MgCl2, 1 mmol/L ethylenedi- dation. Thus, cosupplementation of L- setts Medical School. Adult male Sprague- aminetetraacetic acid (EDTA), 100 mmol/L arginine with L-ascorbic acid and BH4 Dawley rats weighing 265–285 g (Charles NaF, 1% NP 40, 1 mmol/L phenylmethyl- might enhance the therapeutic outcome River Laboratories, Wilmington, MA, sulfonyl fluoride (PMSF) and 1 g/mL by reducing oxidant stress and preserv- USA) were maintained in a clean housing aprotinin. The lysates were centrifuged, ing eNOS in its functional dimeric form, facility on a 12-h light/dark cycle. the supernatant recovered and the protein

MOL MED 18:676-684, 2012 | YAN ET AL. | 677 BH4, L -ARGININE AND VITAMIN C IN LIMB ISCHEMIA

concentration determined (Pierce Biotech- nology, Rockford, IL, USA). A total of 100 µg protein per sample was separated on a 7.5% or 12% sodium dodecyl sulfate–polyacrylamide gel for determination of eNOS or nitrotyrosine, respectively, and then electroblotted on nitrocellulose membranes (Bio-Rad, Hercules, CA, USA). Membranes were incubated overnight at 4°C with mouse monoclonal anti-nitrotyrosine (Cayman Chemical, Ann Arbor, MI, USA; 1:1,000), mouse monoclonal anti-eNOS (BD Transduction Laboratories, San Diego, CA, USA; 1:1,000) or rabbit anti-p-eNOS (cell signaling) and then incubated for 2 h with horseradish peroxidase–conjugated anti- mouse or rabbit IgG antibody (Pierce Biotechnology; 1:5,000). Immunoreactive bands were visualized using the en- hanced chemiluminescence system (Amersham, Arlington Heights, IL, USA). Band density was quantitated by standard densitometry, and the intensity of Figure 1. Effects of BH4, L-arginine (L-arg) and vitamin C (VitC) on eNOS and phosphorylated the band of interest was expressed as a eNOS expression in the ischemic gastrocnemius muscle. Muscle was harvested 14 d after the induction of hindlimb ischemia, and supernatants from muscle homogenates were used in function of the α-tubulin band. these assays. (A) Expression of eNOS was increased by all dietary additives, although the greatest increase was noted in rats that received BH4 + L-arginine + vitamin C. (B) Expression eNOS Activity of p-eNOS was increased in rats fed BH4 + L-arginine or BH4 + L-arginine + vitamin C. Control Rats were sacrificed 14 d after induc- rats were fed a standard diet, whereas the four treatment groups consumed diets supple- tion of ischemia for determination of gas- mented with additives noted in the bar graph. Concentrations of these additives are noted trocnemius muscle NOS activity. The tim- in the text. Data are mean ± standard deviation (sd); n = 5–6. *p < 0.05 versus control; †p < ing of sacrifice was selected on the basis 0.05 versus BH4 or L-arginine groups; ‡p < 0.05 versus BH4 + L-arginine group. of previous work that demonstrated maximal postischemic change in this variable at this time (19). Muscles were homoge- and expressed as carboxypeptidase M centration measurement according to the nized in ice-cold buffer (250 mmol/L (CPM)/µg protein. The Ca2+-dependent protocol of the Nitrate/Nitrite fluoro- Tris-HCl, pH 7.4, 10 mmol/L EDTA, NOS activity, which corresponds to the metric assay kit (Cayman Chemical). 10 mmol/L EGTA [ethylene glycol-bis{β- sum of the endothelial and neural NOS aminoethyl ether}-N,N,N´N´-tetraacetic isoforms, was calculated by subtracting Assay of GSH:GSSG Ratio acid]) and centrifuged, and the protein in the NOS activity measured in the absence The left gastrocnemius muscles were ho- the supernatant was adjusted to 5 µg/mL. of CaCl2 from the NOS activity measured mogenized and deproteinated. And then 2+ Samples were incubated in 10 mmol/L in the presence of CaCl2. The Ca -inde- the supernatant was used for the measure- NADPH, 1 µCi/µL 14C-Arg, 6 mmol/L pendent NOS activity corresponds to in- ment of glutathione disulfide (GSSG) and

CaCl2, 50 mmol/L Tris-HCl (pH 7.4), flammatory isoform of NOS, or iNOS. glutathione (GSH) according to the Glu- 6 µmol/L BH4, 2 µmol/L flavin adenine tathione assay kit (Cayman Chemical). dinucleotide (FAD) and 2 µmol/L flavin NOx (Nitrite + Nitrate) Assay mononucleotide (FMN) for 30 min at The left gastrocnemius muscles after Hindlimb Perfusion 37°C. The reaction was stopped with ischemic d 14 were collected and homog- Hindlimb blood flow was determined 400 µL of 50 mmol/L HEPES, pH 5.5, and enized in phosphate-buffered saline by means of laser Doppler imaging (Moor 5mmol/L EDTA. Identical samples were (PBS). After centrifugation, ultrafilter tis- Instruments, Devon, UK). Flow was mea- prepared without CaCl2, and all reactions sue homogenates were passed through a sured preoperatively, immediately after were performed in duplicate. The radioac- 10-kDa molecular weight cutoff filter. Fil- arterial excision, and then 3, 7, 14, 21, 28, tivity of the sample eluate was measured trate was used for nitrite and nitrate con- 35 and 42 d after induction of ischemia.

678 | YAN ET AL. | MOL MED 18:676-684, 2012 RESEARCH ARTICLE

The timing of sacrifice was selected on the basis of previous work that demonstrated maximal postischemic muscle necrosis at this time (19). The muscle was cut transversely into three 2-mm sections. Two sections were used for nitro blue tetrazolium (NBT) staining, while the third was frozen (–80°C) in optimal cut- ting temperature (OCT) embedding com- pound. Sections for NBT staining were incubated in PBS containing 0.033% NBT (Fisher Biotech, Austin, TX, USA) and 0.133% NADH (Roche Diagnostics, Indi- anapolis, IN, USA) at 21°C for 10 min. The samples were then fixed in 4% paraformaldehyde for 24 h. The areas of viable tissue, indicated by dark blue color, and nonviable tissue, indicated by white color, were measured by quantita- tive image analysis (ImagePro; Media Cy- bernetics, Bethesda, MD, USA). Data Figure 2. Effects of BH4, L-arginine (L-arg) and vitamin C (VitC) on NOS activity in the ischemic gastrocnemius muscle. Muscle was harvested 14 d after the induction of hindlimb were expressed as the ratio of nonviable ischemia, and supernatants from muscle homogenates were used in these assays. tissue to total tissue area. Measurements (A) Ca2+-dependent NOS activity was increased by all dietary additives, although the were made on the four exposed cut sur- greatest increase was noted in mice receiving BH4 + L-arginine or BH4 + L-arginine + vita- faces, and the average was taken and min C. Data are mean ± sd; n = 5–6. *p < 0.05 versus control; †p < 0.05 versus BH4 or used as a single data point for each ani- L-arginine groups; ‡p < 0.05 versus BH4 + L-arginine group. (B) NOx (nitrite + nitrate) was in- mal. The frozen section was used to pre- creased in mice receiving BH4 + L-arginine or BH4 + L-arginine + vitamin C. Data are pare cryosections (10 µm) for hema- † mean ± sd; n = 5. *p < 0.05 versus control; p < 0.05 versus BH4 or L-arginine groups. 2+ toxylin and eosin (H&E) staining to (C) Ca -independent NOS activity was less in mice receiving BH4 + L-arginine or BH4 + evaluate histological integrity of the mus- L-arginine + vitamin C than in control mice. Data are mean ± sd; n = 5–6. *p < 0.05 versus cle, as well as to detect the presence of an control. In all graphs, control rats were fed normal chow and underwent hindlimb ischemia, whereas the four treatment groups consumed diets supplemented with additives inflammatory infiltrate. noted in the bar graph. Concentrations of these additives are noted in the text. STATISTICAL ANALYSIS Analyses were carried out by means of Scans were obtained during inhalation of patella. The number of intersections be- analysis of variance (ANOVA). Post hoc 1% isofluorane while core body tempera- tween contrast-filled vessels and gridlines Student-Newman-Keuls tests were car- ture was maintained between 36.8 and was determined independently by three ried out if the ANOVA F statistic was sig- 37.2°C. Scans were repeated three times, blinded observers. The angioscore was cal- nificant to determine sites of difference and the average for each rat was deter- culated as the average ratio of intersections within the ANOVA format. Probability mined. Data were expressed as the ratio to the total number of gridlines. Within values less <0.05 were accepted as signif- of ischemic to non-ischemic hindlimb. the experimental setting of this study, the icant for all statistical calculations. angioscore is a marker of collateral artery Angiograms enlargement; thus, as collateral arteries di- RESULTS Angiograms were performed 42 d after late and remodel in response to femoral induction of ischemia. Barium sulfate artery excision, their diameters increase, Effects of Oral BH4, L-Arginine, and (2.5 mL; EZPaqe, Merry X-Ray, South San enhancing their visibility on the X-ray Vitamin C on Gastrocnemius eNOS Francisco, CA, USA) was infused into the film, thus increasing the angioscore (19). and p-eNOS Expression infrarenal aorta after ligation of the proxi- Dietary supplementation of BH4, mal aorta and inferior vena cava during Nitroblue Tetrazolium Staining to L-arginine and vitamin C significantly af- inhalation of 2% isofluorane. A grid was Detect Muscle Necrosis fected eNOS and p-eNOS expression in superimposed over the film between the The left gastrocnemius muscle was re- the ischemic gastrocnemius muscle (Fig- greater trochanter of the femur to the moved 7 d after induction of ischemia. ure 1). Rats given single supplementation

MOL MED 18:676-684, 2012 | YAN ET AL. | 679 BH4, L -ARGININE AND VITAMIN C IN LIMB ISCHEMIA

NOx concentration in ischemic gastrocnemius was significantly increased from rats fed BH4 and L-arginine or BH4, L-arginine and vitamin C (Figure 2B). There were no changes in rats fed with either BH4 and L-arginine alone.

Effects of Oral BH4, L-Arginine and Vitamin C on Gastrocnemius Ca2+- Independent NOS Activity Ca2+-independent NOS activity was greater in the ischemic gastrocnemius from rats fed normal chow than in the ischemic gastrocnemius of dietary fed rats (Figure 2C). Rats fed a single supplement (BH4 or L-arginine), or the combination of these two agents, demonstrated Ca2+-independent NOS activity statistically similar to the normal chow group (that is, there was no beneficial effect noted in these dietary intervention Figure 3. Effects of BH4, L-arginine (L-arg) and vitamin C (VitC) on oxidative stress in the is- groups). However, rats provided with chemic gastrocnemius muscle. Muscle was harvested 14 d after the induction of hindlimb three dietary supplements (BH4 + L-argi- ischemia, and supernatants from muscle homogenates were used in these assays. (A) Ni- nine + vitamin C) demonstrated Ca2+-in- trotyrosine expression was decreased by all dietary additives, although this decrease was dependent NOS activity levels in the is- greatest in rats receiving BH4 + L-arginine + vitamin C. (B) The GSH:GSSG ratio was increased in rats receiving BH4 + L-arginine; the addition of vitamin C to this regimen chemic gastrocnemius muscle that were caused an additional increase in this ratio. Data are mean ± sd; n = 5–6. *p < 0.05 versus lower than the other dietary intervention † control, BH4 or L-arginine groups; p < 0.05 versus BH4 + L-arginine groups. groups.

Effects of Oral BH4, L-Arginine and with BH4 or L-arginine demonstrated given a single dietary supplement (BH4 Vitamin C on Gastrocnemius 2+ similar levels of eNOS expression, and or L-arginine) demonstrated Ca - Oxidative Stress these levels were significantly greater dependent NOS activities that were simi- Dietary supplementation affected ni- than that of rats fed normal chow. Rats lar to each other and greater than that of trotyrosine accumulation and the ratio of given two (BH4 + L-arginine) or three rats fed normal chow. Rats given two GSH versus GSSG in the ischemic gas- (BH4 + L-arginine + vitamin C) supple- (BH4 + L-arginine) or three (BH4 + trocnemius muscle (Figure 3). Rats given ments also had greater levels of eNOS L-arginine + vitamin C) supplements dis- a single dietary supplement (BH4 or 2+ expression than rats given single supple- played significantly greater Ca -depen- L-arginine), or the combination of these ments. Hence, the combination of BH4 dent NOS activity than rats given a dietary agents, displayed similar nitroty- and L-arginine had an additive effect on single dietary supplement. The combina- rosine levels. These levels were signifi- eNOS expression, and the addition of vi- tion of BH4 and L-arginine generated an cantly less than the level noted in rats tamin C provided an additional benefi- additive effect. The addition of vitamin C fed normal chow. Rats provided with cial effect. However, increased p-eNOS further increased Ca2+-dependent NOS all three dietary supplements (BH4 + expression was only observed in BH4 + activity, although it did not reach statisti- L-arginine + vitamin C) exhibited a ni- L-arginine– and BH4 + L-arginine + vita- cal significance. trotyrosine level significantly less than min C–fed rats. rats given a single supplement (BH4 or Effects of Oral BH4, L-Arginine and L-arginine) or the combination of these Effects of Oral BH4, L-Arginine, and Vitamin C on Gastrocnemius NOx two agents. The ratio of reduced versus 2+ Vitamin C on Gastrocnemius Ca - Levels oxidized glutathione (GSH:GSSG) was Dependent NOS Activity The final products of NO in vivo are ni- also measured as another index of ox- 2+ – – Dietary supplements affected Ca - trite (NO2 ) and nitrate (NO3 ). NOx is idative stress in the ischemic gastrocne- – – dependent NOS activity in the ischemic the sum of NO2 and NO3 and it is the mius muscle. GSH:GSSG ratio was in- gastrocnemius muscle (Figure 2A). Rats best index of total NO production. The creased in the ischemic gastrocnemius of

680 | YAN ET AL. | MOL MED 18:676-684, 2012 RESEARCH ARTICLE

chow. Rats provided with two (BH4 + L-arginine) or three (BH4 + L-arginine + vitamin C) supplements showed significantly greater recovery of foot perfusion than rats fed normal chow or rats given a single dietary supplement. This difference was evident at the later phase of recovery, on d 21, 28 or 42 after induction of ischemia, which was also the time of maximal collateral artery wall remodeling (20). A similar pattern was noted for collateral artery angioscores determined on d 42 after induction of ischemia. Hence, the angioscore was significantly greater in rats given two (BH4 + L-arginine) or three (BH4 + L-arginine + vitamin C) supplements than in rats fed normal chow or in rats given a single supplement (Figures 4B, C).

Effects of Oral BH4, L-Arginine and Vitamin C on Gastrocnemius Muscle Necrosis The extent of gastrocnemius necrosis was affected by the provision of dietary supplements. Rats given a single supplement (BH4 or L-arginine) or the combination of these agents manifest a similar degree of gastrocnemius muscle necrosis. Moreover, the extent of necrosis noted in these dietary intervention groups was Figure 4. Effects of BH4, L-arginine and vitamin C on perfusion recovery after induction of similar to that noted in rats fed normal hindlimb ischemia. (A) Laser Doppler perfusion imaging (LDPI) data, expressed as the ratio chow (that is, these dietary regimens did of blood flow from the ischemic to nonischemic hindlimbs, was determined before, imme- not improve postischemic muscle in- diately after and then serially over the ensuing 6 wks. Group identity is shown in the color tegrity). However, rats provided with key. Data are mean ± sd; n = 6. *p < 0.05 for BH4 + L-arginine or BH4 + L-arginine + vitamin C all three dietary supplements (BH4 + groups versus all other groups. (B) The angioscore, determined 42 d after the induction of L-arginine + vitamin C) had significantly hindlimb ischemia and calculated as described in the text, was greater in rats receiving less gastrocnemius necrosis than rats fed BH4 + L-arginine or BH4 + L-arginine + vitamin C than in all other groups. Data are mean ± sd; normal chow, rats provided with a single n = 6. *p < 0.05 for BH4 + L-arginine or BH4 + L-arginine + vitamin C groups versus all other groups. (C) Representative angiogram from each study group. dietary supplement or rats given the combination of BH4 + L-arginine. This difference was evident on macroscopic rats fed the two dietary combination Effects of BH4, L-Arginine and Vitamin and microscopic levels. The percentage (BH4 + L-arginine) or the three dietary C on Hindlimb Perfusion of the cut surface of the ischemic gastroc- combination (BH4 + L-arginine + vita- Dietary supplementation significantly nemius muscle that was necrotic, deter- min C). Addition of vitamin C had a increased the recovery of hindlimb per- mined by NBT staining, was significantly beneficial effect in increasing GSH:GSSG fusion after induction of severe ischemia, less in the BH4 + L-arginine + vitamin C ratio. Taken together, three dietary sup- and this effect was regimen and time de- group than in all other groups (Fig- plementation (BH4 + L-arginine + vita- pendent (Figure 4A). Rats given a single ures 5A, B). Groups fed normal chow, or min C) would be much more effective in supplement (BH4 or L-arginine) demon- supplemented with BH4, L-arginine, or decreasing ischemic muscles oxidative strated similar degree of perfusion recov- both agents, demonstrated similar histo- stress after induction of hindlimb ische- ery in the foot, and this level was also logical evidence of severe necrosis: mus- mia in rats. similar to that noted in rats fed normal cle nuclei were nearly absent, intra-

MOL MED 18:676-684, 2012 | YAN ET AL. | 681 BH4, L -ARGININE AND VITAMIN C IN LIMB ISCHEMIA

myofiber vacuolization was substantial and the distance between myofibers was large (Figure 5C). A pronounced inflammatory infiltrate was also present in these groups. In contrast, the BH4 + L-arginine + vitamin C group demonstrated good preservation of muscle his- tology and only a limited inflammatory cell infiltrate.

DISCUSSION The following study hypothesis was supported by our findings: dietary cosupplementation with tetrahydrobiopterin (BH4), L-arginine and vitamin C acts synergistically to decrease oxidative stress, increase nitric oxide and thereby improve perfusion and tissue recovery in response to acute hindlimb ischemia more than supplementation with a single supplement. Interestingly, two patterns of effect emerged. Cosupplementation with BH4 + L-arginine increased the dependent variables eNOS and p-eNOS expression, Ca2+-dependent NOS activity, foot perfusion and the collateral artery angioscore more than the addition of either component separately, whereas the addition of vitamin C provided a further beneficial effect on these variables, although only eNOS expression reached statistical significance. In addition, co- administration of all three dietary supplements had a significantly greater ef- Figure 5. The effects of BH4, L-arginine (L-arg) and vitamin C (VitC) on necrosis in the ische- fect than BH4 or L-arginine, given mic gastrocnemius muscle. (A) Gastrocnemius muscle was removed 7 d after induction of individually or in combination, when the hindlimb ischemia. Transverse sections of this muscle were stained with nitroblue tetra- dependent variables of Ca2+-independent zolium to determine the ratio of necrotic versus viable surface area. This ratio was less in NOS activity, oxidative stress or muscle rats receiving BH4 + L-arginine + vitamin C than in all other groups. Data are mean ± sd; necrosis were measured. n = 6. *p < 0.05 for BH4 + L-arginine + vitamin C groups versus all other groups. (B) Represen- eNOS and p-eNOS expression, Ca2+- tative images from each study group. (C) Histological sections of gastrocnemius muscle dependent NOS activity, foot perfusion taken from the ischemic hindlimb 7 d after the induction of hindlimb ischemia were stained with H&E. Representative sections are shown. Note the loss of myofibers and the in- and the collateral artery angioscore are tense cellular infiltrate in the control, BH4 and L-arginine groups. Best preservation of muscle linked by established cause-and-effect re- architecture was consistently observed in rats receiving BH4 + L-arginine + vitamin C. lationships. eNOS-derived NO is a potent vasodilator (2); hence, the increased eNOS expression and activity present in of the vasodilator effect of NO (24), but postischemic arteriogenesis, the process the BH4 + L-arginine group should result also indirect, insofar as eNOS-derived wherein existing collateral arteries un- in an NO-dependent increase in foot per- NO is critical to collateral artery remod- dergo remodeling designed to restore fusion, and this expectation was realized eling. These effects include mobilization vascular conductance (26). This process by our findings. Moreover, eNOS- of endothelial progenitor cells from bone was evidenced by the increased an- derived NO is a critical determinant in marrow and their subsequent homing to gioscore, a marker of collateral artery the response to hindlimb ischemia the ischemic hindlimb (25). Once there, enlargement, in rats provided with L- (21–23). This effect is direct, on the basis endothelial progenitor cells participate in arginine + BH4 dietary supplements.

682 | YAN ET AL. | MOL MED 18:676-684, 2012 RESEARCH ARTICLE

may also restore BH4 or NO levels, since tivity of vitamin C, or indirect, due to the we observed an increase in eNOS activ- beneficial effect of vitamin C on BH4 lev- ity and p-eNOS expression, although it els (8,27), insofar as BH4 also exhibits po- did not reach statistical difference be- tent antioxidant activity (7). cause of the number of samples ana- Although it is well established that en- lyzed, dosage of vitamin C or time of dothelial dysfunction related to vascular data collection. Cosupplementation with oxidative stress is a critical factor in PAD BH4 + L-arginine may have sufficiently pathogenesis, dietary supplementation restored endothelial BH4 levels and with L-arginine or antioxidants, such as thereby intracellular redox balance. vitamin C, have had equivocal effects on Hence, the addition of vitamin C did not long-term outcome (12,13,16). Dietary further p-eNOS expression or activity, supplementation with L-arginine alone and hence NO bioavailability. has a beneficial effect when given acutely Ca2+-independent NOS activity and tis- (that is, via intravenous infusion [14] or sue nitrotyrosine accumulation were sig- for short duration [2 months] [15]), and nificantly lower in rats receiving all three these clinical results are consistent with dietary supplements than in rats receiving the positive effects observed in rats pro- Figure 6. Schematic illustration of possible BH4 or L-arginine, or the combination of vided with dietary L-arginine. However, mechanism of three dietary combined the two. When measured by methods long-term administration of L-arginine regimen can increase NO bioavailability used herein, Ca2+-independent NOS activ- (6 months) not only failed to demonstrate and decrease oxidative stress, accord- ity is an authentic reflection of iNOS ac- a beneficial effect, but resulted in a degree ingly increase blood flow recovery and re- tivity, inasmuch as the assay was con- of eNOS-dependent vascular reactivity duce tissue necrosis. A•–, ascorbate; AH –, dehydroascorbate. ducted in vitro, in the absence of shear significantly less than that of the placebo stress, which can activate eNOS in the ab- group (16). The present findings demon- sence of Ca2+ via phosphorylation (32). strated increased iNOS activity after in- Vitamin C likely exerted its beneficial The marked elevation of iNOS activity in duction of ischemia. If a similar circum- effects in this study through a variety of rats fed normal chow indicates the pres- stance is present in PAD, then the singular molecular mechanisms. In its capacity as ence of postischemia inflammation, which dietary supplementation with L-arginine, an antioxidant, it enhances NO bioavail- is also evidenced by the cellular inflam- the substrate for all NOS isoforms, might – ability by quenching O2 , thus limiting matory infiltrate in this group. Nitrotyro- serve to worsen vascular inflammation, a the inactivation of NO that occurs when sine accumulation is indicative of critical participant in the pathogenesis of – – – O2 and NO combine to produce OONO OONO -induced damage (33), and it is in- PAD (1). Vitamin C reduces vascular in- (3). Vitamin C also stabilizes existing teresting that the group that exhibited the flammation (34), improves redox balance BH4 (8) and increases endothelial BH4 least amount of tissue damage (rats pro- (11) and eNOS-dependent vascular reac- synthesis (27), thus minimizing eNOS vided with all three supplements) also tivity (10), but these effects have only “uncoupling,” which in turn lessens gen- had the least nitrotyrosine accumulation. been evaluated on a short-term basis, – eration of O2 by eNOS and reduces vas- Moreover, rats in the triple therapy group whereas retrospective cross-sectional cular oxidative stress (7). However, BH4 demonstrated a virtual absence of nitroty- studies have failed to confirm that dietary is itself a potent antioxidant (7), and ad- rosine. We interpret the present findings supplementation with antioxidants im- ministration of exogenous BH4 has been to indicate that muscle injury after ische- proves PAD outcome (12,13). We interpret established to increase endothelial BH4 mia is not entirely contingent upon loss of the present findings to indicate that provi- levels (28). Moreover, L-arginine directly perfusion, inasmuch as cosupplementa- sion of BH4 + L-arginine + vitamin C act- stimulates eNOS expression (29), en- tion with the antioxidant vitamin C ing synergystically might prove a useful hances eNOS activity by a receptor- clearly reduced tissue injury and elimi- therapeutic alternative in PAD treatment. dependent G protein–linked process (30) nated nitrotyrosine accumulation, but did To this end, the use of sapropterin dihy- and limits the inhibitory effect of asym- not affect perfusion or collateral artery en- drochloride, a synthetic form of (6R)-L- metric dimethylarginine on eNOS- largement. Instead, we speculate that vita- erythro-5,6,7,8-tetrahydrobiopterin re- derived NO production (31). We propose min C provided an antioxidant effect that cently approved for the treatment of that under the experimental conditions limited tissue injury generated by inflam- phenylalanine hydroxylase deficiency imposed by hindlimb ischemia, addition matory cells for which action depends in (35), might provide a practical means for of vitamin C to BH4 + L-arginine signifi- part on oxidant production (for example, the provision of BH4. cantly decreased oxidative stress and ac- neutrophils and macrophages). This effect In conclusion, cosupplementation with cordingly decreased tissue necrosis. It could be direct, due to the antioxidant ac- BH4 + L-arginine + vitamin C resulted in

MOL MED 18:676-684, 2012 | YAN ET AL. | 683 BH4, L -ARGININE AND VITAMIN C IN LIMB ISCHEMIA

increased eNOS activity and NO concen- 9. Langlois M, Duprez D, Delanghe J, De Buyzere blood flow recovery but not for arteriogenesis. tration as well as greater foot blood flow M, Clement D. (2001) Serum vitamin C concen- Arterioscler. Thromb. Vasc. Biol. 27:1–8. recovery than rats receiving normal tration is low in peripheral arterial disease and is 25. Aicher A, et al. (2003) Essential role of endothelial associated with inflammation and severity of nitric oxide synthase for mobilization of stem chow or either agent separately. The ad- atherosclerosis. Circulation. 103:1863–8. and progenitor cells. Nat. Med. 9:1370–6. dition of vitamin C to the BH4 + 10. Silvestro A, et al. (2002) Vitamin C prevents en- 26. Helisch A, Schaper W. (2003) Arteriogenesis: the L-arginine regimen further reduced ox- dothelial dysfunction induced by acute exercise development and growth of collateral arteries. idative stress and tissue injury in ische- in patients with intermittent claudication. Athero- Microcirculation. 10:83–97. mic muscles (Figure 6). The clearly supe- sclerosis. 165:277–83. 27. Huang A, Vita J, Venema R, Keaney J Jr. (2000) rior outcome of rats provided with BH4 + 11. Wijnen M, et al. (2001) Antioxidants reduce oxida- Ascorbic acid enhances endothelial nitric-oxide tive stress in claudicants. J. Surg. Res. 96:183–7. synthase activity by increasing intracellular L -arginine + vitamin C warrants investi- 12. Donnan P, Thomson M, Fowkes F, Prescott R, tetrahydrobiopterin. J. Biol. Chem. 275:17399–406. gation of a cosupplementation strategy Housley E. (1993) Diet as a risk factor for periph- 28. Sawabe K, Wakasugi K, Hasegawa H. (2004) as a therapeutic alternative in PAD. eral arterial disease in the general population: Tetrahydrobiopterin uptake in supplemental ad- the Edinburgh Artery Study. Am. J. Clin. Nutr. ministration: elevation of tissue tetrahydro- ACKNOWLEDGMENTS 57:917–21. biopterin in mice following uptake of the exoge- et al. This work was supported by National 13. Klipstein-Grobusch K, (2001) Dietary an- nously oxidized product 7,8-dihydrobiopterin tioxidants and peripheral arterial disease: the and subsequent reduction by an anti-folate- Institutes of Health, National Heart, Rotterdam Study. Am. J. Epidemiol. 154:145–9. s e n s i t i v e p r o c e s s . J. Pharmacol. Sci. 96:124–33. Lung, and Blood Institute grant RO-1 14. Böger R, et al. (1998) Restoring vascular nitric 29. Kohil R. (2004) Dietary L-arginine supplementa- HL-75353 (to LM Messina), as well as oxide formation by L-arginine improves the tion enhances endothelial nitric oxide synthesis grants from the Pacific Vascular Research symptoms of intermittent claudication in pa- in streptozotocin-induced diabetic rats. J. Nutr. Foundation and the Wayne and Gladys tients with peripheral arterial occlusive disease. 134:600–8. J. Am. Coll. Card. 32:1336–44. 30. Joshi M, et al. (2007) Receptor-mediated activa- Valley Foundation (to LM Messina). 15. Maxwell A, Anderson B, Cooke J. (2000) Nutri- tion of nitric oxide synthesis by arginine in en- tional therapy for peripheral arterial disease: a dothelial cells. Proc. Natl. Acad. Sci. U. S. A. DISCLOSURE double-blind, placebo-controlled, randomized 104:9982–7. The authors declare that they have no trial of HeartBar. Vasc. Med. 5:11–9. 31. Böger R, Ron E. (2005) L-arginine improves vas- competing interests as defined by Molecu- 16. Wilson A, Harada R, Nair N, Balasubramanian cular function by overcoming the deleterious ef- lar Medicine, or other interests that might N, Cooke J. (2007) L-arginine supplementation in fects of ADMA, a novel cardiovascular risk fac- peripheral arterial disease: no benefit and possi- tor. Alt. Med. Rev. 10:14–8. be perceived to influence the results and ble harm. Circulation. 116:188–95. 32. Ayajiki K, Kindermann M, Hecker M, Fleming I, discussion reported in this paper. 17. Ueda S, et al. (1999) Tetrahydrobiopterin restores Busse R. (1996) Intracellular pH and tyrosine endothelial function in long-term smokers. J. Am. phosphorylation but not calcium induce shear REFERENCES Coll. Card. 35:71–5. stress-induced nitric oxide production in native 1. Silvestro A, Oliva G, Brevetti G. (2002) Intermit- 18. Heitzer T, Krohn, K, Albers S, Meinertz T. (2000) endothelial cells. Circ. Res. 78:750–8. tent claudication and endothelial dysfunction. Tetrahydrobiopterin improves endothelium-de- 33. Beckman J, Koppenol W. (1996) Nitric oxide, su- Eur. Heart J. 4:B35–40. pendent vasodilation by increasing nitric oxide peroxide, and peroxynitrite: the good, the bad, 2. Marletta M. (1993) Nitric oxide synthase structure activity in patients with type II diabetes mellitus. and ugly. Am. J. Physiol. 271:C1424–37. and mechanism. J Biol. Chem.; 268:12231–12234. Diabetologia. 43:1435–8. 34. Aguirre R, May J. (2008) Inflammation in the vas- 3. Pacher P, Beckman J, Liaudet L. (2007) Nitric 19. Yan J, et al. (2010) Oral tetrahydrobiopterin im- cular bed: importance of vitamin C. Pharmacol. oxide and peroxynitrite in health and disease. proves the beneficial effect of adenoviral- Ther. 119:96–103. Physiol. Rev. 87:315–424. mediated eNOS gene transfer after induction of 35. Yamamizu K, Shinozaki K, Ayajiki K, Gemba M, 4. Peterson T, et al. (1999) Opposing effects of reac- hindlimb ischemia. Mol. Ther. 18:1482–9. Okamura T. (2007) Oral administration of both tive oxygen species and cholesterol on endothe- 20. Park B, et al. (2010) Endothelial nitric oxide syn- tetrahydrobiopterin and L-arginine prevents en- lial nitric oxide synthase and endothelial cell thase affects both early and late collateral arterial dothelial dysfunction in rats with chronic renal caveolae. Circ. Res. 85:29–37. adaptation and blood flow recovery after induc- failure. J. Cardiovasc. Pharmacol. 49:131–9. 5. Kuzkaya N, Weissman N, Harrison D, Dikalov S. tion of hindlimb ischemia in mice. J. Vasc. Surg. (2003) Interactions of peroxynitrite, tetrahydro- 51:165–73. biopterin, ascorbic acid, and thiols. J. Biol. Chem. 21. Yan J, Tang G, Wang R, Messina L. (2005) Opti- 278:22546–54. mization of adenovirus-mediated endothelial ni- 6. Bevers L, et al. (2006) Tetrahydrobiopterin, but tric oxide synthase delivery in rat hindlimb is- not L-arginine, decreases NO synthase uncou- chemia. Gene Ther. 12:1640–50. pling in cells expressing high levels of endothe- 22. Yu J, et al. (2005) Endothelial nitric oxide syn- lial NO synthase. Hypertension. 47:87–94. thase is critical for ischemic remodeling, mural 7. Schmidt T, Alp N. (2007) Mechanisms for the role cell recruitment, and blood flow reserve. Proc. of tetrahydrobiopterin in endothelial function Natl. Acad. Sci. U. S. A. 102:10999–1004. and disease. Clin. Sci. 113:47–63. 23. Lloyd P, Yang H, Terjung R. (2001) Arteriogenesis 8. Heller R, et al. (2001). L-ascorbic acid potentiates and angiogenesis in rat ischemic hindlimb: role endothelial nitric oxide synthesis via a chemical of nitric oxide. Am. J. Physiol. 281:H2528–38. stabilization of tetrahydrobiopterin. J. Biol. Chem. 24. Mees B, et al. (2007) Endothelial nitric oxide syn- 276:40–7. thase activity is essential for vasodilation during

684 | YAN ET AL. | MOL MED 18:676-684, 2012