Vitamin C revisited

1Critically ill patients suffer from multiple organ dysfunction mostly occurring in the course of ischemia/reperfusion or septic shock. In these conditions, overwhelming amounts of reactive oxygen species (ROS) and reactive nitrogen species are generated. ROS are oxidizing agents produced during mitochondrial respiration and phagocytosis. In addition, ROS cause post-translational modifications of proteins, modifying their action and affecting cellular signaling, gene expression, oxygen sensing and other physiological processes. In low concentrations, ROS also enhance the antioxidant response via nuclear factor-erythroid 2-related factor 2 activation, and thereby promote cell survival. While ROS are crucial for body homeostasis and defense, they cause harm if abundant production overwhelms the antioxidant defense. In that case, ROS can induce reversible or irreversible injury to proteins, lipids and nucleic acids, thereby contributing to endothelial dysfunction, cellular injury and multiple organ dysfunction.

2Endothelial dysfunction is a uniform, ROS-mediated manifestation of ischemia/reperfusion and sepsis. Furthermore, ROS-induced damage to the glycocalyx, cellular membranes and junctions leads to increased permeability, adhesion of leukocytes and platelets with local activation of inflammation and coagulation, leads to loss of endothelial vasodilatation and attenuates the vascular response to vasoconstrictors. Subsequent hypotension, vascular leakage and microcirculatory flow impairment at reperfusion augment tissue hypoxia due to increased diffusion distance for oxygen and may thereby enhance cellular damage and organ failure. Ascorbate, the redox form of vitamin C, is a physiological antioxidant. We hypothesize that the early administration of a high pharmacological dose of vitamin C to patients with sepsis or after ischemia/reperfusion can reduce oxidative damage to endothelial and other cells, and thereby improve tissue perfusion and oxygenation, and mitigate subsequent organ dysfunction.

3Vitamin C also has other effects. Vitamin C improves immune function, and facilitates enteral uptake of nonheme iron, reduction of folic acid intermediates and synthesis of collagen (wound healing), cortisol, catecholamines and carnitine. These effects are beyond the scope of this review.

4The aim of this narrative review is to summarize the role of vitamin C in mitigating ROS-induced damage to endothelial and myocardial cells in ischemia/reperfusion or sepsis. By limiting endothelial dysfunction, vitamin C might improve tissue perfusion and reduce tissue hypoxia and subsequent organ dysfunction. Experimental and clinical studies on the use of vitamin C are reported with a focus on cardiovascular effects.

5Ischemia/reperfusion-induced and sepsis-induced endothelial dysfunction is initiated by increased amounts of ROS produced by the induction of enzymes such as nicotinamide adenine dinucleotide phosphate-oxidase (NOX) and uncoupling of mitochondrial oxidative phosphorylation and endothelial nitric oxide synthase (eNOS). ROS are additionally produced by xanthine oxidase, lipoxygenase and cyclooxygenase, and during oxidation of catecholamines (Figure 1).

6Unopposed ROS oxidize tetrahydrobiopterin (BH4), the cofactor of eNOS, and thereby reduce eNOS activity, the enzyme producing endothelial nitric oxide (eNO). eNO initiates vasodilatation by stimulating soluble guanylcyclase and increasing cyclic guanosine monophosphate in smooth muscle cells. It also inhibits platelet aggregation and adhesion of activated platelets and leukocytes. eNO is therefore crucial for patency of the microcirculation and its depletion hampers organ perfusion and oxygenation. In the absence of BH4, eNOS becomes uncoupled, producing superoxide (O2–) rather than nitric oxide (NO). O2– and NO yield peroxynitrite, the most damaging ROS.

7The underlying mechanisms for the effect of ascorbate on these pathways have been demonstrated in in vitro studies with cultured endothelial cells. Endothelial cells can accumulate ascorbate to millimolar levels and represent an appropriate model to study the effects of high-dose vitamin C. In vitro studies are presented focusing on protective effects of vitamin C (ascorbate) in ischemia/reperfusion and sepsis.

8Ascorbate decreases oxidative stress in endothelial cells by reducing the production of O2–, hydrogen peroxide and peroxynitrite. Mechanisms include prevention of NOX activation, decreased inducible nitric oxide synthase (iNOS) expression and increased NO bioavailability (Figure 1). NOX is the major source of ROS in endothelial and myocardial cells. Activation of NOX leads to the formation of intracellular O2–. Addition of ascorbate to endothelial cells exposed to oxidative stress prevented activation of NOX by inhibition of subunit p47phox expression (mediated by the Jak2/Stat1/IRF1 signaling pathway). NOX-derived ROS additionally increase the expression of iNOS, producing excessive NO. Abundant NO in the presence of O2– generates peroxynitrite. Ascorbate prevented iNOS expression. Ascorbate also scavenges O2–, but only at high levels ≥10 mmol/l.

9Furthermore, ascorbate can increase NO bioavailability by preventing BH4 oxidation and recovering oxidized BH4. Recovery of BH4 by ascorbate prevents uncoupling of eNOS and associated O2– production and restores eNOS activity and subsequent generation of eNO, which has a pivotal role in endothelial-dependent vasodilatation.

10ROS additionally increase endothelial permeability, causing edema and contributing to organ dysfunction. Ascorbate can tighten the endothelial barrier through several pathways.

11Constitutive eNO is required to control endothelial permeability and prevent loss of tight junctions between cells. Loading endothelial cells with ascorbate preserved eNO generation through eNOS and decreased endothelial permeability. This effect depended on eNOS and guanylatecyclase, suggesting that tightening of the endothelium involved NO generation by eNOS and subsequent NO-dependent activation of guanylatecyclase.

12Exposure of endothelial cells to lipopolysaccharide (LPS) increases endothelial permeability by inducing NOX-dependent protein phosphatase 2A activity and subsequent occludin dephosphorylation. Phosphorylated occludin is crucial to maintain tight junctions. Ascorbate protects against vascular leakage by inhibiting protein phosphatase 2A activation.

13Furthermore, oxidants and LPS increase apoptosis, impairing the endothelial barrier. LPS decreases Bcl-2 (which inhibits apoptosis) and increases Bax (which suppresses the ability of Bcl2 to block apoptosis). Ascorbate inhibited apoptosis and protected endothelial progenitor cells. Simultaneous administration with vitamin E had a synergistic effect on the prevention of apoptosis.

14Oxidative stress stimulates the expression of tissue factors and cellular adhesion molecules at the surface of platelets and endothelial cells, promoting adhesion of leucocytes to the endothelium and formation of microthrombi and thus impairing microcirculatory flow. In cultured endothelial cells, ascorbate inhibited the tumor necrosis factor alpha-induced expression of intracellular adhesion molecule-1 in a dose-dependent manner, probably by modulating the production of ROS and reactive nitrogen species. By preventing intracellular adhesion molecule expression, ascorbate reduces leukocyte plugging in microvessels and microcirculatory flow impairment.

15Ischemia/reperfusion injures not only the endothelium but also the myocardium, leading to stunning and arrhythmias. Loading isolated cardiomyocytes subjected to hypoxia/reoxygenation with ascorbate improved their resistance to cell death by decreasing ROS generation and inhibiting (proapoptotic) Bax expression, caspase-3 activation, and cytochrome-c translocation into the cytoplasm. Pretreatment with vitamin C or vitamin E of isolated cardiomyocytes exposed to singlet oxygen reduced the number of hypercontracted cardiomyocytes in a concentration-dependent manner. Simultaneous administration of both vitamins acted synergistically.

16In sepsis, ascorbate also influences macrophage activity and bacterial growth. Macrophages play an important role in sepsis, enhancing cytokine production as well as production of several types of ROS. ROS are necessary to overcome infections, but are only beneficial if their production is controlled. Incubation of macrophages with ascorbate regulated the phagocytic process by reducing adherence, chemotaxis, ingestion and O2– anion production. Furthermore, ascorbate has profound bacteriostatic activity. Ascorbate (in concentrations from 100 to 1,000 μM) significantly inhibited bacterial replication in dilute fecal samples in vitro.

17Beneficial effects of ascorbate pretreatment on organ function were observed in ischemia reperfusion injury models of rat heart and rabbit kidney, and of rat skeletal muscle, lung and liver. Studies are summarized in Table 1.

18In a rat model of cardiac arrest (ventricular fibrillation and electrical shock), intravenous ascorbate at the start of resuscitation alleviated postcardiac arrest myocardial necrosis and mitochondrial damage, reduced lipid peroxidation and improved the resuscitation success rate and 72-hour survival. In a pig model of coronary artery ischemia/reperfusion and preconditioning, intravenous ascorbate started 30 minutes before ischemic preconditioning or ischemia had no effect on infarct size, but abolished the reduction in infarct size by ischemic preconditioning. In a rabbit renal ischemia model, ascorbate ameliorated renal structure and function. Ascorbate also had a positive effect on muscle injury when administered intravenously (i.v.) during ischemia/reperfusion of the leg in rats. Experiments with different intravenous ascorbate doses to rats prior to hepatic ischemia reperfusion found antioxidant effects at low and moderate doses and predominant pro-oxidant effects at extremely high doses (1,000 mg/kg). In addition, vitamin C administered after ischemic preconditioning but before clamping of the liver blood flow reduced hepatic mitochondrial damage and dysfunction. Thus, in these animal models, ascorbate reduced ischemic organ injury and improved function, but abolished the protective effects of ischemic preconditioning on organ injury.

19The most frequently used sepsis animal models are feces injection into the peritoneum (FIP), cecal ligation and puncture (CLP) and intraperitoneal injection of LPS.

20Ascorbate plasma concentrations rapidly declined in lymphocytes and macrophages, muscle and plasma of septic rodents. Endotoxin also depleted myocardial ascorbate in guinea pig hearts after as early as 4 hours, even after dietary supplementation for 5 weeks.

21Ascorbate (76 mg/kg i.v.) restored blood pressure and prevented microvascular dysfunction in skeletal muscle of septic rats. In a mouse model of FIP, ascorbate (10 mg/kg i.v.) inhibited impairment of microvascular perfusion when administered early (0 hours) and reversed septic platelet adhesion and flow impairment with late administration (after 6 hours). Similarly, ascorbate improved capillary blood flow by an eNOS-dependent mechanism when administered 6 hours after FIP. In addition, ascorbate (200 mg/kg i.v.) administration before CLP protected against impaired arteriolar constriction and loss of catecholamine responsiveness and improved 24-hour survival in mice. Ascorbate prevented arteriolar vasoconstriction by inhibiting eNOS uncoupling and iNOS-derived and neuronal nitric oxide synthase-derived NO production, when given before but also 3 hours after CLP in mice. Moreover, ascorbate (200 mg/kg i.v.) prevented vascular leakage in a mouse CLP model by inhibiting excessive production of NO by iNOS and neuronal nitric oxide synthase, and of O2– by NOX, and by inhibiting the activation of protein phosphatase 2A, thereby preserving occludin phosphorylation. Ascorbate thus restored several seemingly contradictory disorders contributing the heterogeneity of the septic circulation. Ascorbate improved microcirculatory perfusion by NOX inhibition and arteriolar vasodilator responsiveness (neuronal nitric oxide synthase related), restored vasoconstrictor responsiveness by inhibiting iNOS expression and prevented vascular leakage.

22Vitamin C prevented the increase in endotoxin-induced myocardial uric acid accumulation, a marker of ischemia-induced oxidative stress, and protected against endotoxin-induced oxidative damage to proteins in the guinea pig liver. Ascorbate (100 mg/kg i.v.) reduced hepatic microvascular dysfunction during polymicrobial sepsis when administered immediately after CLP in rats by reducing oxidative stress and lipid peroxidation, and regulating hepatic vasoregulatory gene expression. In addition, ascorbate prevented the sepsis-induced decrease in several cytochrome P450 enzyme activities, thereby improving drug-metabolizing function. Ascorbate (200 mg/kg i.v.) also attenuated sepsis-induced acute lung injury in a mouse model of FIP or LPS and improved 72-hour survival. Finally, oral prefeeding with ascorbate decreased bacterial concentrations and improved survival after intraperitoneal injection of Klebsiella pneumonia in mice.

23In human volunteers, both ischemia/reperfusion injury (20 minutes of forearm ischemia) and low-dose LPS reduced plasma vitamin C concentrations and diminished acetylcholine-induced, endothelial-dependent vasodilatation. High-dose (24 mg/minute) intra-arterial vitamin C increased BH4 concentrations, reduced neutrophil oxidative burst and completely restored the response to acetylcholine, but not to glyceryl-trinitrate (endothelium-independent dilatation), supporting its endothelial protective effect. Vitamin C also corrected the LPS-induced decreased responsiveness to norepinephrine and angiotensin II. Both vasopressors act independent of the endothelium, but their effect is blunted by oxidative stress and inflammation. These volunteer studies originate from the same group.

24Several preclinical in vitro, animal and volunteer studies thus show that vitamin C in moderate to high doses can reduce ROS-induced microcirculatory flow impairment, microvascular leakage, decreased responsiveness to vasoconstrictors, and myocardial and other organ injury. Contradictory results in ischemia/reperfusion can partially be explained by the timing of administration, because vitamin C abrogates ischemic preconditioning. Furthermore, vitamin C reduces the overwhelming neutrophil response and inhibits bacterial replication.

25Vitamin C plasma concentrations depend on absorption, the distribution volume, cellular uptake, consumption and renal reabsorption and excretion. Patients with sepsis, hemorrhage, multiple organ failure, stroke, traumatic brain injury or after cardiac surgery have low vitamin C concentrations in plasma and leukocytes, probably due to increased consumption in the cell and high leukocyte turnover. Since intracellular ascorbate concentrations in mononuclear leucocytes and in granulocytes are respectively 80 and 25 times higher than in plasma, a high production and turnover of these cells contributes to depletion. Low plasma concentrations correlate with inflammation (C-reactive protein) and multiple organ failure, suggesting consumption during oxidative stress.

26Recommended doses of vitamins are generally based on preventing deficiency in healthy humans. In healthy volunteers, manifestations of vitamin C deficiency (fatigue and/or irritability) occurred at plasma concentrations below 20 μmol/l. Clear scurvy can develop below 11 μmol/l. With sufficient vitamin C intake (100 to 300 mg/day), plasma concentrations plateau at 70 to 85 μmol/l and do not exceed 220 μmol/l at maximal oral intake (3 g six times daily). Oral dose is limited by intestinal absorption and high oral intake causes diarrhea. Urinary excretion depends on the plasma concentration and is minimal at low plasma concentrations due to active tubular reabsorption. Threshold plasma concentration for excretion may be 55 μmol/l.

27Ascorbate is transported into the cell by ascorbate-specific membrane transporters and less so as dehydroascorbic acid via glucose transporters. Within the cell, dehydroascorbic acid can be rapidly reduced to ascorbate, thereby recycling ascorbate. The ascorbate transporter SVCT1 is expressed predominantly in epithelial tissues such as the intestine and kidney, maintaining optimal vitamin C concentrations in the body. The ascorbate transporter SVCT2 delivers ascorbic acid to tissues. Immune and inflammatory cells have intracellular concentrations 10 to 80 times higher than plasma, protecting them against ROS generated by respiratory burst or phagocytosis. Neurons have concentrations as high as 10 mmol/l, sufficient for scavenging O2–. Of note, ascorbate does not pass the blood–brain barrier. However, dehydroascorbic acid does so via the glucose transporter GLUT1 and is reduced to ascorbate after uptake in neurons. Intravenous administration of dehydroascorbic acid confers supraphysiologic concentrations of ascorbate in the brain.

28Notably, concentrations attained with high oral dosing are sufficient to modulate enzymes such as nicotinamide adenine dinucleotide phosphate but not for scavenging O2–, which reacts with NO at a rate 105-fold greater than that with ascorbate. A plasma concentration of 10 mmol/l would be required to compete with NO and for complete restoration of NO bioavailability. High plasma concentrations can be obtained with intravenous administration. However, mild dietary supplementation of vitamin C reduced peroxynitrite formation and atrial electrophysiological remodeling induced by rapid pacing in dogs, probably due to higher intracellular vitamin C concentration.

29High intravenous vitamin C doses, up to 3 to 6 g daily, are needed to restore normal plasma concentrations in critically ill patients. To attain plasma concentrations over 10 mmol/l for 3 hours, a short-term infusion of 30 to 100 g would be required. High pharmacological doses of vitamin C seem to be well tolerated. Prolonged oral intake of high-dose vitamin C increases the risk of oxalate kidney stones. However, this complication has not been reported with short-term high intravenous dosing. Of note, low dose ascorbate can also act as a pro-oxidant. However, after a 1 g vitamin C intravenous infusion, ascorbyl radical concentrations increased much more in healthy controls than in septic patients, who had a lower baseline concentration.

30Whereas most preclinical studies investigate the role of vitamin C alone, clinical studies often used a combination of antioxidants. Some studies are performed in conditions similar to animal models, reporting the use of vitamin C before the ischemic incident (coronary bypass surgery) or directly at reperfusion (percutaneous coronary interventions) or shock resuscitation (burns). In the critically ill studies, combinations of antioxidants were generally given after the hyperacute phase.

31In patients undergoing elective percutaneous coronary intervention, vitamin C in a dose of 1 g over 1 hour improved microcirculatory reperfusion, and left ventricular and renal function. This improvement was associated with reduced markers of oxidative injury.

32Atrial fibrillation is the most common arrhythmia after cardiac surgery, developing in 15 to 50 % of patients depending on several risk factors. Atrial fibrillation increases short-term and long-term morbidity and length of hospital stay, and necessitates anticoagulation to prevent stroke. Its pathogenesis is multimodal but accumulating evidence indicates a role of oxidative stress. Ischemia/reperfusion, atrial stress and angiotensin increase atrial NOX activity, which is associated with postoperative atrial fibrillation. Oxidative damage initiates breakdown of cell membranes, mitochondrial dysfunction, calcium overload, apoptosis and also inflammation by signaling activation of nuclear factor-kappaB and activator protein-1 transcription factors, thereby initiating electrophysiological remodeling.

33After cardiac surgery, a massive depletion of vitamin C has been observed. Several studies suggest a beneficial effect of vitamin C on the occurrence of new atrial fibrillation and some on enhanced recovery, although not all studies are positive (Table 2). In a matched control study, the perioperative use of vitamin C reduced the incidence of new postoperative atrial fibrillation in patients undergoing coronary artery bypass grafting. A subsequent randomized controlled trial found a reduction in postoperative atrial fibrillation when adding vitamin C to a β-blocker. In the largest trial, vitamin C did not reduce atrial fibrillation, but it reduced time on mechanical ventilation. Another clinical trial found a reduction in atrial fibrillation, but its incidence in the control group was extremely high. The most recent randomized controlled trial found a reduction in postoperative atrial fibrillation comparing preoperative ω-3 poly-unsaturated fatty acids with vitamin C and vitamin E supplementation with placebo (see Table 2). An older Chinese study using a very high dose of intravenous vitamin C (250 mg/kg) found less cardiac injury, better cardiac performance and shorter intensive care and hospital stay.

34Studies differ in the timing, route and dose of vitamin C, and in the combination of other antioxidants. Timing may be crucial in cardiac surgery, because preoperative episodes of ischemia and reperfusion protect the myocardium against perioperative ischemic damage and vitamin C may hamper the beneficial effects of ischemic preconditioning on reducing infarct size.

35Several clinical trials in critically ill patients have reported favorable results of high-dose vitamin C alone, or in combination with vitamin E or with selenium, zinc and vitamin B (Table 3). The main beneficial outcomes include reduction in pulmonary morbidity and new organ failure, less mechanical ventilation days and shorter length of ICU and/or hospital stay. Some studies measured lower markers of oxidative stress. Although ROS can signal host defense in low concentrations, the parallel finding of less oxidant stress and less organ dysfunction suggests a beneficial effect of reducing overwhelming ROS during critical illness. The largest study, however, using a mixture of micronutrients including oral vitamin C, found no effect on 28-day mortality or length of stay. Of note, the control group in Berger and colleagues’ study received 500 mg/day vitamin C.

36Combined administration with vitamin E and other micronutrients obscures the role of vitamin C. However, vitamin C regenerates vitamin E, and vitamin E is only consumed after depletion of vitamin C. Two small studies in burn patients studied a very high dose of vitamin C alone (66 mg/kg/hour) for about 24 hours and found a reduction in resuscitation volume, better gas exchange and less days on mechanical ventilation and increased urinary output, probably indicating less capillary leak. No signs of acidosis or renal insufficiency were found with this high dose. However, although vitamin C reduced morbidity in some studies, a mortality reduction was not found. We hypothesize that the effect of vitamin C can be improved by very early administration of a high intravenous dose as part of the resuscitation bundle in patients with shock.

37This narrative review summarizes the role of vitamin C in mitigating ROS-induced microcirculatory impairment and associated organ failure in ischemia/reperfusion or sepsis. Preclinical studies show that high-dose vitamin C can prevent or restore ROS-induced microcirculatory flow impairment, prevent or restore vascular responsiveness to vasoconstrictors, preserve endothelial barrier and augment antibacterial defense. These protective effects against oxidative stress seem to mitigate organ injury and dysfunction, and promote recovery in most but not all clinical studies after cardiac revascularization and in critically ill patients.

38Of note, many questions remain to be solved, including the optimal dose, timing, combination of vitamin C with other antioxidants and the inhibiting effect of vitamin C on the protection of ischemic preconditioning. However, high-dose vitamin C provides a cheap, strong and multifaceted antioxidant. Future research should answer the question of whether short-term high-dose intravenous vitamin C can mitigate the overwhelming oxidant cascade and thereby improve resuscitation of the macrocirculation and microcirculation and limit cellular injury in critically ill patients.