Reactive oxygen species (ROS) are normally generated at basal amounts, but ischemia or oxygen re-supply during reperfusion lead to abnormally high ROS levels that cannot be cleared by normal anti-oxidizing systems, causing tissue reperfusion injury [1]. An oxygen (O₂) molecule generates the intermediate species, superoxide radical (O₂·), hydrogen peroxide (H₂O₂), and hydroxyl radical (OH·), during the monovalent reduction reaction, and it is eventually converted to water (H₂O) by catalase. In this process, the intermediates O₂· and OH· are important free radicals that cause peroxidation (LOO·) of intracellular lipids and tissue damage [2]. In particular, OH·, the most reactive ROS [1], is produced by the Fenton reaction (Fe²⁺ + HOOH → Fe³⁺ + OH· + OH^-) and the Haber-Weiss reaction (O₂·^- + NO· → OONO^- + H⁺ → OONOH → NOO· + OH·)[3]. ROS generated during reperfusion causes organ damage in the heart [3,4], brain [5,6], and blood vessels [7,8].

In 1980, Furchgott and Zawadzki [9] found an endothelium-derived relaxing factor (EDRF) later shown to be nitric oxide (NO) [10]. Other vasoactive substances, including endothelium-derived hyperpolarizing factor (EDHF) and the endothelium-derived contracting factor (EDCF), are released from the endothelium to regulate vascular tone. However, because the endothelium is composed of a cell monolayer, it is easily damaged by physical and chemical stimuli, including ROS during reperfusion. Inhalation anesthetics [1,4,11] and lidocaine, a local anesthetics, can prevent ischemia-reperfusion injury in the brain [12].

In this study, we tested the effects of lidocaine and procaine, two local anesthetics, on endothelial-dependent relaxation in the rabbit aorta to examine if their anti-oxidizing effects could suppress or reduce the endothelial injury by ROS.

We generated ROS by direct EL of a physiologic salt solution (PSS), such as the K-H solution [8], and the other researchers are directly injecting H₂O₂ to PSS [13]. ROS such as O₂·, H₂O₂, and OH· are generated during PSS EL [14-17]. We did not add normal levels of ascorbic acid (1.1 × 10^-4 M) to exclude its anti-oxidizing effect. A number of tiny air bubbles were generated at the positive and negative electrodes when the electric current was applied. In mouse thoracic aorta, 5 minutes of EL is sufficient to induce ROS injury [8]. However, 5 minutes of EL with 15 mA constant current caused too much injury in our model, so we used a 35-second treatment, indicating that there are species differences in the degree of injury by ROS.

Lecour et al. [18] verified that the function of the left ventricle declined when ROS was generated by 1 minute of EL in the K-H solution with a constant current of 1, 1.5, 3, 5, 7.5, and 10 mA, as measured by the Langendorff perfusion method. Superoxide radicals and hydroxyl radicals were also generated by the reaction of the 5, 5-dimethyl-1-pyrroline-N-oxide (DMPO)-OH signal, the adduct of DMPO EL, with superoxide dismutase (SOD : 100 IU/ml) and mannitol (50 mM), followed by a quantitative analysis with electron spin resonance. The DMPO-OH signal occurred as early as 10 seconds after starting EL and was maintained for 60 seconds, with diminishing levels after EL stopped. SOD pretreatement reduced the DMPO-OH signal by 60%, which showed that the hydroxyl radical generation process was partially by the Fenton reaction of a superoxide radical. Mannitol (50 mM) blocked the DMPO-OH signal but did not change cardiac parameters after EL [18]. Similarly, pretreatment of mannitol a scavenger of hydroxyl radicals, did not prevent endothelial dysfunction in our model.

ROS is also generated via cyclooxygenase during an inflammatory response [19]. Endotheliocytes, Kuffer cells, neutrophils, and macrophages also produce ROS and reactive nitrogen species (RNS) during an inflammatory response via antagonizing pathogenic bacteria in those cells. Antioxidant defenses are depleted by chronic exposure to ROS or RNS, leading to cancers, rheumatoid arthritis, or organ transplantation rejection. Large inflammatory responses also produce ROS, including peroxyl radical (ROO·), hydroxyl radical, superoxide radical, and hydrogen peroxide, and RNS, including ·NO and peroxynitrite anion (ONOO^-).

Vascular relaxation induced by ACh is caused by inducing NO release in endotheliocytes, and ROS inactivates NO [20,21]. In addition, ROS generated by ischemia or reperfusion causes endothelial damage [15,22]. We generated ROS with 35 seconds of EL and then replaced the K-H solution for measurements after 15 minutes of equilibrium. Thus, ACh should induce vascular relaxation via changes in endothelial injury and not from inactivation of NO [20]. The endothelial monolayer is easily damaged and dysfunctional during chronic hypertension, even to the point of detachment [23]. Smooth muscle may also be directly injured by ROS.

During oxygen depletion, such as ischemia, ATP in the endothelium is rapidly consumed, causing failure of the Na⁺-K⁺ ATPase pump and an increase of intracellular [Na⁺] i and [H⁺]i by inhibiting the removal of the metabolites and mitochondrial NADH oxidation. Increased intracellular [H⁺] i leads to an increase in [Na⁺]i by facilitating Na⁺-H⁺ exchange to maintain constant [pH]i and thus triggers the Na⁺-Ca²⁺ exchange, increasing [Ca²⁺]i as well [24]. Increased [Ca²⁺] i deforms proteins and phospholipids [25] via activation of phospholipase A₂ to generate arachidonic acid (AA), a free fatty acids. AA is then converted to prostanoids such as prostacyclin (PGI₂), a strong smooth muscle relaxant, and prostaglandin G, a constrictor, to thromboxane A₂ (TXA₂) or to leukotrienes (LT), which cause the synechia of leukocytes to endotheliocytes. In addition, it facilitates the conversion of xanthine dehydrogenase to xanthine oxygenase (XO) and increases the intracellular superoxide radical (O₂·) [23]. Oxygen re-supply after ischemia increases ROS: O₂·, OH·, and LOO· [26]. These ROS that are generated mainly in mitochondria of neutrophil [25] denature lipids and proteins and damage the cell membrane and mitochondria (mitochondrial permeability transition pore, MPTP) [27], and denature enzymes [25]. O₂· is the first ROS is generated by oxygen re-supply; it is produced by XO or by LT from neutrophils and nerve cells in the microvascular structure. Superoxide radical is converted by SOD to hydrogen peroxide, which eventually becomes water (H₂O) by catalase. Here, catalase (1,000 U/ml) pretreatment prevented endothelial injury by ROS and maintained endothelial-dependent aorta relaxation. However, excess O₂· is converted by the Fenton reaction and Haber-Weiss reaction to a hydroxyl radical, which causes the severe injury and eventually causes lipids peroxidation and cell and tissue injury by LOO· generation [26]. Pretreatment with mannitol (5 mM) as the OH· scavenger did not prevent endothelial injury, potentially because ROS sensitivity may shows species variation and higher doses may be needed.

Lidocaine protects against brain ischemia by reducing energy demand [28]. Anoxia-induced depolarization in the mouse hippocampus pretreated with lidocaine 1 µM and 0.25 µM 15 minutes before the ischemia-induction started after 49 seconds and 44 seconds, respectively, indicating a depolarization delay compared to the 27-second interval in the control group. The amount of ATP in the lidocaine pretreated group increased to 165% and 212% of controls at 1 minute and 2 minutes after ischemia, respectively. The activity of Na⁺-K⁺ ATPase and the activity of Ca²⁺ ATPase were not changed, and [Ca²⁺]i did not increase [28]. These findings imply that lidocaine maintains energy by delaying neuronal depolarization. In addition, [Ca²⁺] i increased in the left ventricle of the excised mouse heart where dysfunction was induced by hydrogen peroxide, but pretreatment with lidocaine 10 µM blocked this increase [29].

In human microvascular endothelial cells (HMVEC), lipopolysacharide in the culture solution caused necrosis of the endotheliocytes; amide local anesthetic pretreatment prevented this necrosis, but tetracaine and procaine, the esters, could not [30]. The protection by amide local anesthetics, including lidocaine, was blocked by 5-hydroxydecanoate (5HD), indicating that the mitochondrial ATP-sensitive K⁺ channel (mito[K⁺_ATP]) is involved in the effect, consistent with the mechanism of ischemia-reperfusion injury. Because EL-induced injury from ROS involves direct injury by ROS that have not undergone the cascade reactions caused in the preconditioned cells, 5-HD pretreatment did not block the endothelial-protecting effect of lidocaine and procaine, indicating that mito[K⁺_ATP] was not involved (data not shown).

We have tested whether ROS causes dysfunction in the rabbit aorta endothelium after lidocaine and procaine pretreatment. Lidocaine and procaine both dose-dependently protected endothelial-dependent relaxation by removing or anti-oxidizing ROS. Although lidocaine was less potent, both showed similar efficacy. In conclusion, pretreatment with lidocaine and procaine dose-dependently showed anti-oxidizing or ROS-removing effects to protect endothelial-dependent relaxation in the abdominal aorta after ROS exposure. Procaine could remove hydrogen peroxide as part of its protection mechanism.