INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Acidosis is a response associated with several pathophysiologic conditions, including sepsis. In its advanced stage, sepsis is characterized by systemic hypotension and vascular hyporeactivity and is associated with metabolic acidosis and organ damage (1). Systemic hypotension and decreased vascular reactivity are presumably attributed to upregulation of inducible nitric oxide synthase and overproduction of nitric oxide (NO) (1). Indeed, endotoxin infusion into animals can lead to systemic hypotension, increased NO production, and acidosis (4, 5). The independent effects of NO and acidosis in such conditions are not well defined. The vasodilatory effects of extracellular acidosis are variably present in different vascular beds such as cerebral (6), coronary (7), skeletal muscle, and visceral circulation (8). Furthermore, extracellular acidosis and, consequently, intracellular acidosis, lead to depressed cell function in excitable and nonexcitable tissues (9).

NO is a potent vasodilator and is normally produced in picomolar concentrations, via calcium-dependent mechanisms, by a constitutive isoform of nitric oxide synthase (cNOS). NO diffuses from the endothelial cells into the adjacent smooth muscle cells where it causes relaxation. The endogenous production of NO at low levels is essential for regulation of blood pressure, blood flow distribution, and cell adhesion to the endothelium (10). Exposure to endotoxin or inflammatory cytokines (such as interleukin-1 or tumor necrosis factor) induces the expression of the inducible isoform (iNOS), which is calcium-independent, and generates large and sustained amounts of NO from many cell types, including endothelial cells (11). High levels of NO under such conditions may cause marked vasodilation, and may become proinflammatory and cytotoxic because of its reactive nature (10). The relationship between pH and NO production has not been studied, but NOS activity, and hence NO production, is likely to be influenced by intracellular pH. Carlin and colleagues (12) found that acute changes in extracellular pH produced increased levels of ENO, suggesting a possible role of pH on NOS activity. However, the effect of chronic changes in pH and NO production is unknown. Sepsis causes acidosis and increased NO production, suggesting a link between the two events.

In this study, we investigated the effect of acidosis on nitric oxide production in healthy anesthetized rats, and evaluated the relationship between acidosis, NO, hypotension, and organ damage. We hypothesized that acidosis could cause vasodilation and systemic hypotension by direct action on smooth muscle and, indirectly, by upregulating NO production. NO overproduction could be principally involved in subsequent tissue damage. To separate the independent effect of pH from the effect of NO, we infused acid into animals that were pretreated with aminoguanidine to inhibit NO synthesis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Surgical Preparation

After approval by the Institutional Committee for Human Use of Animals, 33 adult male Sprague-Dawley rats (Taconic, Germantown, NY) weighing 370 to 430 g were studied. Animals, fasted overnight with free access to water, were anesthetized with an intraperitoneal injection of sodium pentobarbital (50 mg/kg). The experiments were carried out at an ambient room temperature of 23 to 24° C, and no exogenous heating sources were used to alter the animals' temperature. In three rats, rectal temperature was monitored (Physitemp; Physitemp Instruments Inc., Clifton, NJ). A tracheostomy was performed and mechanical ventilation was established (Rodent Ventilator, Model 683; Harvard Apparatus, South Natick, MA) using room air. Ventilatory settings (respiratory rate = 40/min, tidal volume = 3.5 ml) were selected to provide blood gases in the normal range at the beginning of the experiment and were not changed for the duration of the study. Anesthesia was maintained with a continuous infusion of sodium pentobarbital (5 mg/kg/h, infused volume of 1.5 ml/h) through a catheter placed in the external jugular vein, while muscle relaxation was obtained by intravenous injection of pancuronium bromide (0.2 mg, as needed). The carotid artery was cannulated for continuous monitoring of systemic blood pressure and for blood samples. The other external jugular vein was cannulated and used for acid infusion. Animals were heparinized (1,000 U/kg), and baseline measurements, consisting of mean arterial pressure (), airway pressure (Paw), arterial blood gas determinations, and exhaled nitric oxide (ENO), were obtained. and Paw were recorded continuously (Grass Model 7D; Grass Instruments, Quincy, MA).

Experimental Protocol

Control animals (C, n = 10) received a volume of normal saline infusion comparable to all other groups. All other rats were randomly divided into one of three groups, each receiving an intravenous infusion of hydrochloric acid. The infusion rate was adjusted to cause a rapid (15 to 30 min) fall in the pH to 7.35, followed by a slow continuous infusion (maintained for 5 h) to bring the pH close to 7.1. One group (A1, n = 10) received 0.162 mmol of hydrochloric acid (1N; Fisher Scientific, Fairlawn, NJ) during the initial 15 to 30 min and then continued to receive acid at the rate of 0.058 mmol/h (0.132 ± 0.007 mmol/kg/h, infused volume of 0.5 ml/h). A second group (AG+A1, n = 6) received aminoguanidine (Sigma Chemicals, St. Louis, MO) as a bolus of 50 mg/kg, and 5 min later received acid infusion as in the previous group. A third group (A2, n = 7) received two to three times more acid; 0.162 mmol during the initial 15 to 30 min and then a continuous infusion of 0.123 mmol/h (0.307 ± 0.016 mmol/kg/h, infused volume of 0.7 ml/h). Thus, the total amount of fluids infused in all groups was between 2.0 and 2.2 ml/h. All the animals were monitored for 5 h.

, Paw, ENO concentration, arterial blood gases, and pH were measured every 30 min. Blood gases were analyzed using a blood gas analyzer maintained at 37° C (ABL 5; Radiometer, Copenhagen, Denmark). Additional blood (1 ml) was withdrawn during the baseline and at the end of the experiment for plasma nitrite/nitrate determination. After 5 h, animals were killed intravenously with an overdose of sodium pentobarbital. Lung tissue samples were harvested from one lung for myeloperoxidase activity as described below. The other lung was inflated with 10% formalin and immersion-fixed in formalin for histologic examination.

MPO Assay

Myeloperoxidase (MPO) activity was assessed in the lungs as an index of neutrophil aggregation. Briefly, about 100 mg wet weight tissue samples were placed in 2 ml of 50 mM phosphate buffer at pH 7.4. Samples were kept on ice, and then homogenized for 20 s and centrifuged at 20,000 rpm for 20 min at 5° C. The supernatant was discarded. The pellet was resuspended in 1 ml of homogenizing buffer (50 mM phosphate buffer at pH 6.0 containing 10 mM EDTA and 0.5% hexadecyl trimethylammonium bromide [HETAB]), and homogenized for 20 s. The samples were stored at 20° C overnight. Prior to assaying, each frozen sample was thawed and sonicated for 5 min. MPO activity was assayed according to Grisham and colleagues (13): 1 unit of activity was defined as the amount of enzyme present that produced a change in absorbance per minute of 1.0 at 37° C in a final reaction volume containing 0.2 M sodium acetate at pH 3.0.

Histology

After 24 h of submersion in formalin, two transverse sections along the cross apex-base axis were routinely processed. The sections were paraffin embedded, sectioned in 5 µm, and stained with routine hematoxylin-eosin. Vascular changes were scored from zero to 2 for congestion, fibrin-platelet thrombi, and PMN-rich thrombi depending upon the size of the vessels, their dilation, and the compactness of cellularity. Involvement of larger vessels, dilation, and cellular compactness received a score of 2. The individual scores for each section were tabulated, given a total score for each lung, and a group average was calculated.

Nitric Oxide Measurement

ENO was measured using a chemiluminescence analyzer (270 B; Sievers, Boulder, CO). The exhaled gas from the rats was collected for > 3 min using a 1-L polyvinyl bag. Exhaled NO was measured in parts per billion (ppb) using the NO analyzer, which was calibrated daily using nitrogen for a zero and a mixture of 248 ppb NO in nitrogen. Room air was monitored daily, and if NO concentration exceeded 5 ppb, rats were ventilated using air from a gas cylinder with zero NO. ENO was measured during baseline and, thereafter, every 30 min after acid infusion.

Plasma Nitrite/Nitrate (NOx) Concentrations

NOx was measured during baseline and at the end of the study. Heparinized arterial blood (1 ml) was centrifuged at 14,000 rpm for 3 min. Plasma samples were then stored at 78° C. The samples were thawed and 200 µl were pipetted into Eppendorf tubes fitted with Microcon-10 filters (Millipore, Burlington, MA). The samples were then concentrated by centrifuging at 13,000 rpm for 35 min. The filter was discarded and the supernatant was used for NOx determination using a colorimetric assay kit (Alexis Bioch Corp., San Diego, CA). Briefly, 10 or 20 µl of sample plus 70 or 60 µl of buffer (80 µl total volume) were assayed with 10 µl of enzyme cofactor mixture and 10 µl of nitrate reductase mixture. After 3 h of incubation at room temperature, 50 µl of Griess reagent 1 and 50 µl of Griess reagent 2 were added and the color was allowed to develop for 10 min. Absorbances were measured at 540 nm, using a microplate reader (EL310 Microplate Autoreader, Bio-Tek Instruments, Winooski, VT) and converted to NOx concentrations by using a nitrate standard curve.

Statistical Analysis

The results are presented as mean ± SEM. Statistical analysis for changes within each group was performed using analysis of variance (ANOVA) for repeated measures and the Bonferroni test was used as a post hoc test. Differences between the group means at any time point in the experiment were tested using one-way ANOVA followed by the Bonferroni test as a post hoc test. A p value < 0.05 was accepted as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline pH was not significantly different between the groups (average value, 7.44 ± 0.01) (Figure 1). In the control group, pH remained within a normal range for the duration of the experiment (7.41 ± 0.02 at 5 h). In contrast, rats receiving acid exhibited a fall in pH at 30 min, which continued gradually for 5 h. At the end of the experiment, the pH was 7.13 ± 0.05 in the first group (A1) and 7.16 ± 0.03 in the second group (AG+A1). The third group, which received a larger dose of acid (two to three times more), exhibited a decrease in pH that was not significantly different from the other two groups, reaching 7.17 ± 0.05 after 5 h. The changes in pH were not significantly different between the three groups receiving acid. Despite a lack of greater effect on pH in the third group, there were other effects, which are described below. Pa_CO2 did not change significantly in any group and there were no differences between them. Pa_O2 remained close to 100 mm Hg (data not shown). The blood gases and pH were measured at 37° C and were not corrected for body temperature because it was not monitored in all animals. In three rats, rectal temperature was monitored and was found to decrease gradually (from 36.4 ± 0.2° C to 33.3 ± 0.3° C) after 5 h.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Changes in arterial pH in the four groups. Data are expressed as mean ± SEM. *p < 0.05 versus baseline (ANOVA); §p < 0.05 versus control at the corresponding time. C = control; A1 = low dose of HCl; AG+A1 = low dose of HCl after aminoguanidine; A2 = high dose of HCl.

The effect of acidosis on ENO is shown in Figure 2. Baseline values were between 0.8 and 1.8 ppb. In the control group ENO rose slightly, reaching 15.8 ± 2.6 ppb after 5 h. Acidosis in A1 caused a marked increase in ENO, reaching 114.2 ± 22.3 in 5 h. Pretreatment with AG inhibited the rise in ENO for the duration of the experiment. The higher dose of acid (A2) blunted the production of ENO, such that the final concentration was only 38.3 ± 6.3 ppb after 5 h. This value was significantly less than the A1 value (p < 0.05).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Exhaled nitric oxide concentrations during 5 h of study. Data are expressed as mean ± SEM. *p < 0.05 versus baseline (ANOVA); §p < 0.05 versus control. Groups are the same as in Figure 1.

Because ENO is representative only of the amount of NO produced in the lungs, we measured plasma nitrite/nitrate concentration (NOx) as an index of NO production in the systemic circulation. The changes in NOx were consistent with those found from ENO. The baseline NOx ranged from 15.9 to 23.9 µM and was not significantly different between the groups (Figure 3). In the control group, NOx did not change significantly during the experiment (23.9 ± 3.6 µM during the baseline and 21.8 ± 2.4 µM after 5 h). Acid infusion at the low dose caused an increase in NOx from 17.3 ± 3.7 to 35.2 ± 4.3 µM (p < 0.05) that was abolished by AG. Infusion of the high dose of acid significantly increased NOx from 15.9 ± 3.2 during baseline to 29.1 ± 3.6 µM (p < 0.05); however, NOx final concentrations were not statistically different between the two doses of acid.

View larger version (11K): [in this window] [in a new window]   Figure 3.   Plasma nitrite/nitrate (NOx) at baseline and at 5 h. Data are expressed as mean ± SEM. *p < 0.05 versus baseline (ANOVA); §p < 0.05 versus control. Groups are the same as in Figure 1.

Systemic arterial pressure was 117.0 ± 3.5 mm Hg during baseline and was not significantly different between the groups (Figure 4). Blood pressure remained relatively unchanged in the control group (111.5 ± 8.2 mm Hg at 5 h). Infusion of acid at low dose caused a decline in pressure, which became evident after 2 h and reached statistical significance after 3.5 h. At the end of the experiment, the pressure in this group (63.3 ± 15.0 mm Hg) was significantly lower than that of all the other groups (p < 0.05). Aminoguanidine (AG) prevented the decline in blood pressure, despite the decrease in pH. Infusion of a higher dose of acid had only a small effect on pressure, but it was not different from the control group.

View larger version (17K): [in this window] [in a new window]   Figure 4.   Systemic mean arterial pressure changes during the 5 h of study. Data are expressed as mean ± SEM. *p < 0.05 versus baseline (ANOVA); §p < 0.05 versus control. Groups are the same as in Figure 1.

Lung damage was evaluated from MPO activity (Figure 5) and the injury score (Figure 6) from the histologic evaluation. Lung MPO increased significantly after acid infusion (282.2 ± 24.7 U/min/g in the control group versus 679.3 ± 57.3 U/min/g in A1, p < 0.05). The increase in MPO was attenuated by AG, but it remained above the control value (529.4 ± 36.7 U/min/ g). Moreover, infusion of the higher concentration of acid also caused a significant increase in MPO (515.0 ± 46.8 U/min/g), as compared with that in the control group. This increase in MPO activity was slightly attenuated compared with the effect of the low acid dose. The histologic evaluation revealed a significant increase in lung injury in the groups receiving HCl, as compared with the control group (Figure 6). Lung injury scores were not significantly different between the groups receiving acid (3.5 ± 0.6, 4.2 ± 0.6, and 3.8 ± 0.3 in A1, AG+ A1, and A2, respectively). Histologic analysis based on individual parameters, congestion, fibrin, or PMN thrombi also did not show differences between these three groups (data not shown).

View larger version (17K): [in this window] [in a new window]   Figure 5.   Lung myeloperoxidase activity in the four groups at the end of the study. Data are expressed as mean ± SEM. *p < 0.05 versus control. Groups are the same as in Figure 1. The value in the control group was not different from control values from rats killed immediately after anesthesia.

View larger version (8K): [in this window] [in a new window]   Figure 6.   Lung injury scores. Data are expressed as individual values (symbols) and mean (+). Groups are the same as in Figure 1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Systemic hypotension during sepsis is thought to be due to the increase in NO production with subsequent vasodilation (1). Acidosis is frequently associated with sepsis (1, 4, 5) and can have a direct vasodilatory influence on the systemic vasculature (14, 15). The relationship between blood pH and NO production has not been reported. This relationship was examined by infusing acid at a concentration that produced changes in pH comparable to those found during sepsis. We found that acid infusion into healthy rats produced an inflammatory state characterized by systemic hypotension, increases in NO production, lung injury, and increases in lung myeloperoxidase activity. These changes are characteristic of sepsis and septic shock (4, 16).

Acidosis is a common occurrence in sepsis and could have many deleterious effects such as systemic hypotension, bradycardia, ventricular dysrhythmias, impaired response to catecholamine, decreased cardiac contractility, generalized organ damage, and decreased cardiac output. Despite these associations, the effect of extracellular acidosis on hemodynamics and organ damage remains controversial. Furthermore, the effect of pH on NO production is unknown. Systemic hypotension during sepsis in animals (17) and humans (3) is thought to correlate with nitrite/nitrate (NOx) in the plasma. However, the use of NOS inhibitors as treatment for hypotension during sepsis have yielded conflicting results (2, 17, 18), thus suggesting that other factors may also be involved. Injection of endotoxin into animals is associated with acute systemic hypotension and metabolic acidosis, followed by a slow increase in NO production, with a greater decrease in blood pressure (4, 17). Therefore, we hypothesized that acidosis may be involved in the early phase of systemic vasodilation, whereas NO contributes to the delayed phase and to the organ damage caused by the oxidative property of NO and its derivatives (19). To study this hypothesis, we infused HCl intravenously into healthy rats. Hydrochloric acid has been extensively used in the past as a model to study the role of metabolic acidosis (3). The dose of acid was selected to lower the blood pH to values comparable to those seen in sepsis. Acid infusion (0.058 mmol/h) caused responses typical of the responses observed after endotoxin injection: acidosis, an increase in ENO and NOx, systemic hypotension, an increase in MPO, and lung injury. Furthermore, to see if these responses were dose-dependent, we injected twice as much acid, and found that the results were different from those expected: pH did not decrease much more, injury score did not become worse, and, in fact, ENO, NOx, and MPO were attenuated, and blood pressure remained nearly normal. We have no definite explanation for these findings. Doubling the dose of HCl is not likely to produce much change in pH because of the logarithmic relationship. Nevertheless, this small increase in acid concentration caused attenuation in some responses and provided a degree of protection against lung injury.

Exhaled NO represents NO production in the lungs and its levels have been shown to increase during inflammatory conditions such as sepsis, ARDS, and asthma (4, 20, 21). ENO may be the most sensitive index of inflammation in the lung (4, 22). Plasma nitrite/nitrate represents the end-product of NO produced throughout the body (19). Only a small fraction of this comes from the lung. Our data support previous findings (23) and show that the changes in ENO and NOx are tightly related. However, ENO is more sensitive and can be monitored continuously. Carlin and colleagues (12), using isolated rabbit lungs, showed that an acute increase in pH caused a small rise in ENO (within minutes) in a reversible manner. These rapid changes in NO production are probably due to changes in cNOS activity. In the present study, we found that chronic acidosis caused a substantial increase in NO production. The data do not allow us to determine precisely the source of NO; however, the results suggest that iNOS upregulation was involved because the time course was similar to that observed with endotoxin (5), and because pretreatment with AG blocked its production. AG, besides inhibiting iNOS, can have nonspecific effects (9, 11), but at the dose used (50 mg/ kg) its effects are thought to be specific to iNOS inhibition (17). Surprisingly, increasing the dose of acid infusion (without affecting extracellular pH) caused attenuation in NO production (ENO and NOx), lung injury (as indicated from the MPO level), and systemic hypotension. It is possible that the protective effects of the higher dose of acid may be related to changes in intracellular pH. In general, enzymes operate at an optimal intracellular pH. One study showed that cNOS activity is optimal at pH 7.5 in human endothelial cells (24). Because of the different experimental conditions, it is not clear how these data would apply to our findings. Current literature lacks information about the relationship between intracellular and extracellular changes in pH. It is conceivable that a rise in extracellular H+ ion concentration would cause influx into the cell, leading to an increase in intracellular concentration. Increased plasma membrane permeability to hydrogen ions, passive transport of protons, or intracellular production of acid may also be involved (14). We suggest that the optimal pH for iNOS is near 7.0. Intracellular pH is normally around 7.3; therefore, addition of acid would lower intracellular pH towards the optimal value, thus increasing iNOS activity and NO production. Further addition of acid would cause intracellular pH to fall below the optimal value leading to a decrease in NO production. More studies are necessary to support this hypothesis.

Although the data are not shown, we also did five experiments in which we infused lactic acid at a rate equal to that of the low dose of HCl. ENO, blood gases, and blood pressure were measured in these animals, but MPO and plasma NOx were not assessed. The effects of lactic acid infusion at this concentration were qualitatively similar to HCl infusion, but smaller. Lactic acid caused a marginal decrease in pH, moderate systemic hypotension, and a small increase in ENO, which was almost identical to the effect of the large dose of HCl (shown in Figure 2). Studies that compared the effects of lactic acid and HCl in awake animals (25) showed that the differences in the two are due to the respiratory response. However, our animals were sedated and paralyzed, and thus, could not compensate for the acidosis by adjusting their ventilation.

Upregulation of iNOS can occur in many cells, including macrophages, neutrophils, Kupffer cells, hepatocytes, and endothelial cells (10). Macrophages and neutrophils play an important role in the inflammatory response. The effect of acidosis may be secondary to the release of chemotactic agents and cytokines, leading to upregulation of iNOS and NO production, as well as an infiltration of neutrophils into the lung, and lung injury. NO can further act as a chemotactic agent for neutrophils and can be cytotoxic to the lung and other tissues (26, 27). Increased neutrophil accumulation in the lung is indicated by MPO activity (13). Acidosis in this study caused an increase in lung MPO that correlated with ENO levels. Stewart and colleagues (4) showed that increased ENO was associated with pulmonary injury after endotoxin injection in rats. At high concentrations, NO is potentially proinflammatory and cytotoxic because of its reaction with superoxide radicals and peroxinitrite formation (10). In an acidic environment, the reaction between NO and oxygen radicals becomes further enhanced, leading to more tissue damage (19). The correlation between lung MPO and NO levels suggests a cytotoxic role for NO. Our results are consistent with those of Nava and Salazar (28), who found a positive correlation between nitrite/nitrate levels and organ damage. Furthermore, studies on survival in mice and rats have shown that mortality was high when NOx in the plasma was elevated (29). The pretreatment with AG blocked NO production and partially reversed the increase in lung MPO. Some studies have shown that systemic hypotension correlated with NOx levels during sepsis in humans (3) and rats (17), whereas others did not find any correlation (28, 30). Our results are consistent with the latter. Perhaps, critical levels of acidosis, and/or NO may be necessary before hypotension develops. Acidosis can have a direct vasodilatory effect in the systemic vessels because of membrane hyperpolarization (31). Acidosis can also decrease vascular tone by activating the release of vasodilating mediators from endothelial cells (32), or by inhibiting the vasoconstrictive mediators (33).

During the course of the experiment, the animals slowly developed mild hypothermia, since they were allowed to regulate their body temperature without any exogenous source of heat. Hypothermia also occurs in anesthetized rats after endotoxin injection (5, 34). Mild hypothermia exerts a protective role on lung injury during sepsis, perhaps by decreasing cell metabolism and, consequently, oxygen consumption (35). Thus, the responses to acid infusion (lung injury, ENO, MPO) may be attenuated by the low temperature of the animals. The blood gas determinations reported in Figure 1 were measured at 37° C and were not corrected for body temperature. Correction for hypothermia would make Pa_O2 and Pa_CO2 values lower, and pH higher. In addition, metabolic rate is expected to fall gradually during the experiment, secondary to hypothermia.

In conclusion, chronic acidosis is associated with systemic hypotension. Acidosis can directly or indirectly trigger inflammatory responses in the lung leading to upregulation of nitric oxide production and tissue injury. Besides systemic hypotension, NO may be an important factor for tissue injury since iNOS inhibition is protective against inflammation, despite the development of acidosis. We suggest that iNOS inhibition could be protective against the cellular damage resulting from severe acidosis. More studies are needed on the effect of acid on NO upregulation in order to fully appreciate the importance of our results.