Abstract
Nitric oxide (NO) is postulated to play a key role in the pathophysiology of renal failure in sepsis. Whether the renal effects of increased NO are beneficial or harmful remains unclear. In a porcine model of lipopolysaccharide (LPS)-induced shock, we evaluated the effect of LPS on glomerular filtration rate (GFR) and renal blood flow (RBF). We then administered the nonselective nitric oxide synthase (NOS) inhibitor N G-l-arginine methyl ester (l-NAME), and compared its effects on GFR and RBF with those of S-methylisothiourea (SMT), a selective NOS inhibitor, and those of saline. We postulated that SMT, by maintaining constitutive NO, would be more beneficial than either l-NAME or saline. LPS infusion decreased mean arterial pressure (MAP), and increased cardiac output, RBF, and medullary NO content. The increased RBF was diverted to the medulla. There was no evidence of renal dysfunction in the saline-resuscitated group. Both NOS inhibitors increased MAP but decreased RBF, but only l-NAME reduced GFR and increased sodium excretion and renal oxygen extraction. We conclude that NO in endotoxemia is beneficial because it maintains RBF and GFR. Additionally, selective NOS inhibition did not offer any advantages over saline resuscitation.
Keywords: nitric oxide; shock; glomerular filtration rate; renal blood flow; sodium excretion
Acute renal failure (ARF) in sepsis has been associated with reduction in the glomerular filtration rate (GFR) accompanied by a variable decline in renal blood flow (RBF). Both of these functions are influenced by the release of the vasorelaxing agent nitric oxide (NO) (1). Accordingly, NO is postulated to play a predominant role not only in sepsis-mediated hypotension, but also in sepsis-associated renal failure. However, studies of ARF in endotoxemia, the role of NO, and the prospect of NO inhibition as treatment have produced contradictory results (2-9).
The reasons for these discrepancies are multifactorial and include differences in species, time of measurements, adequacy of fluid resuscitation, and the use of anesthesia. Equally important is the type of model used, since many models have not reproduced the hyperdynamic state associated with human sepsis. Another factor is the use of nonselective NOS inhibitors that block both constitutive and inducible NO production. Since constitutive NO production (via endothelial NO synthase [eNOS]) has been shown to have important physiologic functions, its inhibition may be detrimental. Therefore, the use of selective inhibitors of inducible NOS (iNOS) has been proposed as a more fitting mode of treatment. Indeed, the use of the selective iNOS inhibitor S-methylisothiourea (SMT) in a rat model of sepsis improved renal function (10), although the study in which this was found evaluated only serum creatinine, which is not a sensitive marker of renal dysfunction. Moreover, there are few data on the effects of selective iNOS inhibitors on GFR and RBF in higher-order mammals.
In this study, we used a swine model, since the cardiopulmonary and renal systems of pigs are very similar to those of humans (11). Initially, we examined the early renal response to endotoxemic shock; we then assessed the role of NO in the pathophysiology of ARF by inhibiting NO synthesis with both selective and nonselective NOS inhibitors, and we compared these with saline resuscitation. We minimized the effects of anesthesia by using a heavily sedated, nonanesthetized model. We replicated clinical conditions by resuscitating animals only after shock had developed. We avoided excessive administration of either NOS inhibitor by not specifying a dosage a priori; rather, we infused each agent to a comparable increase in mean arterial pressure (MAP). We postulated that selective iNOS inhibition would result in improvement of renal function, whereas nonselective NOS inhibition would be detrimental.
This study was approved by the animal care and use committee of our institution, and was in keeping with National Institutes of Health guidelines. Female Yorkshire pigs, weighing 20 to 25 kg, were fasted for 12 h before the day of the experiment, but were allowed free access to water.
The animals were initially anesthetized with ketamine and xylazine (both at 2 mg/kg) for intubation and instrumentation. Animals were mechanically ventilated, their body temperature was maintained at 37 ± 2° C (mean ± SD), and supplemental oxygen was administered. A femoral pulmonary artery catheter and an arterial line were inserted. A second arterial catheter was introduced from a carotid artery into the left ventricle (LV), and its position was confirmed via pressure tracings. This second arterial line was used to administer microspheres into the LV and to measure left ventricular end–diastolic pressure (LVEDP). A urinary catheter was introduced into the bladder.
As described previously (12), colored microspheres (NuFlow fluorescent microspheres; Interactive Medical Technologies, Los Angeles, CA) were used to distinguish changes in renal medullary versus cortical blood flow.
One set of renal vessels was exposed and a 2- to 3-mm Doppler flow velocity probe (Crystal Biotech, Hopkinton, MA) was placed around the renal artery. A catheter was inserted into the renal vein and sutured in place. The skin and subcutaneous tissue were then closed in layers.
GFR was estimated by measuring the renal clearance of the soluble inulin polyfructosan-S (Inutest; Fresnius Pharma, Linz, Austria). A priming dose of inulin was administered, followed by an infusion (13). Venous blood and urine were collected in 30-min intervals starting from time zero (baseline), which was set at 60 min after the bolus dose. Inulin clearance (Clin), expressed in ml/min/kg, was calculated according to the manufacturer's kit (Fresnius Pharma). The filtration fraction was calculated as the ratio of GFR to RBF.
Nitrate/nitrite (NOx) levels in urine and in cortical and outer medullary tissues were determined with the chemiluminescence technique (14). Urine and sera were also collected for measurements of sodium and creatinine, from which the fractional excretion of sodium (FENa) was calculated.
Anesthetic agents were stopped after instrumentation, and the animals were sedated with alphaxolone/alphadolone (Saffan; Schering-Plough, Welwyn, UK) at a dose of 8 to 10 mg/kg. This dose has been shown to maintain heavy sedation and analgesia with minimal suppression of vagal and sympathetic baro- and chemoreceptor function (15). Sixty minutes after instrumentation, baseline measurements were recorded, and the animals were randomized into four groups. One group (n = 5) acted as a time control. This group received normal saline (NS) only, and was monitored for 6 h. In three other groups, an infusion of LPS (Escherichia coli O55:B5; Sigma Chemical, St. Louis, MO) was dissolved in 50 ml of normal saline and administered over a 30-min period at a dose of 100 μg/kg. Along with LPS, saline was administered to maintain a pulmonary artery opening pressure (Ppao) of once to twice the baseline value. Shock, defined as a decrease in MAP to ⩽ 60 mm Hg or a 30 mm Hg drop from baseline, occurred by 3 h in all animals receiving LPS. Animals in endotoxemic shock were then randomized into three groups (n = 8 in each group). The LPS/saline group received only saline. The LPS/l-NAME group was resuscitated with the nonselective NOS inhibitor N G-l-arginine methyl ester (l-NAME), and the third group was treated with SMT (LPS/SMT group), both from Sigma, and both dissolved in 5% dextrose in water (D5W) at a concentration of 25 mg/ml. At the start of resuscitation, l-NAME and SMT were titrated to elevate MAP to between 5 and 15 mm Hg for the duration of the experiment. All endotoxemic animals continued to receive saline as needed to maintain a Ppao of once to twice its baseline value. Animals were followed for 3 h after the development of shock and the start of resuscitation (5.5 to 6 h total).
We used SigmaStat software (Jandel, San Rafael, CA) to analyze the findings. Data were compiled and expressed as mean ± SD. Within-group differences in mean values were analyzed with analysis of variance (ANOVA) for repeated measures. Similarly, differences of means between groups were analyzed for the three time periods shown (i.e., baseline, shock, and final). If significance was found, a post hoc analysis (Newman–Keuls) was used to determine the source of the significance. The null hypothesis was rejected at the 5% level.
For ease of presentation, data are shown from three time periods: (1) baseline, after 1 h of stabilization, and before LPS administration; (2) shock, when MAP had decreased by the preset amount, which usually occurred from 2.5 to 3 h after LPS administration; and (3) final, when data were gathered at 3 h after the onset of shock, before termination of the experiment. The dose of l-NAME used to achieve the desired increase in MAP was 4 ± 2 mg/kg, the dose of SMT required was 6 ± 3 mg/kg.
Hemodynamic results are shown in Table 1. LPS administration resulted in a decrease in both MAP and systolic ventricular rate (SVR). Saline administration alone maintained this MAP for the duration of the experiment in the LPS/saline group. Cardiac output (CO) was increased at the time of shock, and remained elevated with saline. The NOS inhibitors had the same hemodynamic effects. There were no changes in central venous pressure (CVP) or Ppao in any group. To confirm that Ppao was representative of LVEDP in the face of pulmonary hypertension induced by LPS and NOS inhibitor, we also assessed LVEDP at the above time points described earlier. There was approximately a 1:1 relationship before and after either LPS or NOS inhibitor administration.
| MAP | CO | SVR | Ppao | CVP | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Baseline | ||||||||||
| Control | 97 ± 16 | 1.7 ± 0.4 | 4,400 ± 400 | 3 ± 1 | 1 ± 0.5 | |||||
| LPS/saline | 103 ± 16 | 1.6 ± 0.4 | 4,600 ± 1,120 | 3 ± 1 | 1 ± 0.2 | |||||
| LPS/l-NAME | 98 ± 6 | 1.8 ± 0.4 | 4,320 ± 880 | 3 ± 0.6 | 1 ± 0.3 | |||||
| LPS/SMT | 102 ± 15 | 1.7 ± 0.5 | 4,240 ± 960 | 3 ± 0.4 | 1 ± 0.4 | |||||
| Shock/3 h | ||||||||||
| Control | 99 ± 18 | 1.4 ± 0.4 | 5,120 ± 1,920 | 3 ± 1 | 1 ± 1 | |||||
| LPS/saline | 69 ± 9* | 2.0 ± 0.6* | 2,560 ± 1,360* | 5 ± 2 | 1 ± 0.6 | |||||
| LPS/l-NAME | 66 ± 4* | 2.0 ± 0.4* | 2,720 ± 720* | 4 ± 1 | 1 ± 0.4 | |||||
| LPS/SMT | 65 ± 7* | 2.4 ± 0.5* | 2,400 ± 640* | 5 ± 2 | 1 ± 0.7 | |||||
| Final/6 h | ||||||||||
| Control | 89 ± 3 | 1.6 ± 0.5 | 5,480 ± 1,520 | 3 ± 1 | 1 ± 1 | |||||
| LPS/saline | 64 ± 12* | 1.9 ± 0.6 | 2,880 ± 880* | 6 ± 4 | 1 ± 0.6 | |||||
| LPS/l-NAME | 71 ± 14* | 1.2 ± 0.3† | 4,880 ± 1,680† | 3 ± 2 | 2 ± 1 | |||||
| LPS/SMT | 75 ± 9* | 1.2 ± 0.2† | 5,200 ± 1,360† | 3 ± 1 | 2 ± 0.8 |
Microsphere measurement revealed that baseline cortical blood flow exceeded medullary flow by fivefold. In fact, medullary blood flow initially constituted 17% to 22% of total RBF, and remained at this proportion in the control group (Figure 1B). LPS infusion increased RBF as assessed by flow probe analysis (Figure 1A). Microsphere measurement then showed that the increased RBF was shunted to the medulla, with no increase in cortical blood flow (Figure 2). When animals were in shock, medullary flow had increased by about threefold, and constituted about 45% of total RBF (Figure 1B). It remained at this value in the LPS/saline group in spite of a low MAP. Renal artery flow decreased markedly with both NOS inhibitors as compared with RBF in the saline group (Figure 1A), and this decrease occurred in both the cortex and medulla (Figure 2). However, in both NOS inhibitor groups, medullary flow remained higher than at baseline (Figure 2), and still constituted about 40% of RBF (Figure 1B). The changes in RBF could not be fully explained by the changes in CO. For example, when the animals were in shock, LPS infusion resulted in a 16 ± 8% increase in CO over the baseline value. In this same time period, RBF increased by 53 ± 15%. In the final time period, CO remained 8 ± 2% higher than at baseline in the LPS/saline group, whereas RBF was 32 ± 7% higher (p < 0.05).
Fig. 1. (A) Normalized renal artery flow as assessed with a flow probe. RBF increased markedly when animals were in shock, but decreased to the same extent after administration of either l-NAME or SMT. (B) Medullary flow as percent of total RBF. Medullary flow constituted 17% to 21% of total RBF, and this increased to about 37% when animals were in shock. This percent remained constant despite RBF reduction after NO inhibition. *Significant versus baseline; †significant versus LPS/saline group; ‡significant versus LPS/SMT group; §significant versus control group.
[More] [Minimize]
Fig. 2. Changes in cortical and medullary blood flow measured with the microsphere technique. (Top) Cortical flow did not change with LPS infusion when animals were in shock, but dropped below baseline by the final measurement after administration of NOS inhibitors. (Bottom) Medullary flow increased markedly after administration of LPS, and decreased with NOS inhibition, but remained higher than baseline. There were no changes in the control group. *Significant versus baseline; †significant versus LPS/saline group; ‡significant versus LPS/SMT group; §significant versus control group.
[More] [Minimize]Renal functional values are shown in Table 2. GFR increased from baseline to the point at which the animals developed shock. From shock to the end of the experiment, this increase was sustained with saline administration. After l-NAME infusion, GFR decreased to baseline levels and GFR was unchanged in the LPS/SMT group. In the LPS/saline group, FENa did not change. However, SMT decreased FENa, whereas l-NAME caused a marked increase in excreted sodium during this period. The filtration fraction decreased with l-NAME administration during the interval from shock to the end of the experiment.
| Baseline to 3 h | 3 h to 6 h | |||
|---|---|---|---|---|
| Control group | ||||
| Clin | 1.9 ± 0.7 | 1.7 ± 1.1 | ||
| FENA | 2.0 ± 0.8 | 3.0 ± 1.6 | ||
| FF | 0.10 ± 0.04 | 0.10 ± 0.03 | ||
| LPS groups | ||||
| Clin | ||||
| Saline | 4.9 ± 2‡ | 4.0 ± 1.7‡ | ||
| l-NAME | 5.8 ± 3‡ | 2.1 ± 1*,† | ||
| SMT | 5.3 ± 1‡ | 3.7 ± 0.8‡ | ||
| FENA | ||||
| Saline | 2.3 ± 1 | 2.7 ± 1.3 | ||
| l-NAME | 2.5 ± 1 | 10.1 ± 3.3*,†,‡ | ||
| SMT | 1.9 ± 0.4 | 1.2 ± 0.3*,‡ | ||
| FF | ||||
| Saline | 0.19 ± 0.05‡ | 0.16 ± 0.03‡ | ||
| l-NAME | 0.17 ± 0.04‡ | 0.11 ± 0.04 | ||
| SMT | 0.18 ± 0.07‡ | 0.20 ± 0.1‡ |
Values of fluid administration and urine volume are shown in Table 3. As expected, the LPS/saline group received more saline between the shock and final time period than did either NOS inhibitor group. There were no differences in urine output between the groups. To further confirm that saline infusion did not cause intravascular dilution in the LPS/saline group, we measured hemoglobin (Hgb) concentration. We found no difference in Hgb between the groups.
| Baseline to Shock | Shock to Final | |||||
|---|---|---|---|---|---|---|
| Infusate | ||||||
| Saline | 600 ± 380 | 1,375 ± 660 | ||||
| l-NAME | 900 ± 600 | 660 ± 230† | ||||
| SMT | 550 ± 300 | 300 ± 124† | ||||
| Urine volume, ml | ||||||
| Saline | 280 ± 150 | 185 ± 100 | ||||
| l-NAME | 400 ± 200 | 140 ± 85 | ||||
| SMT | 320 ± 250 | 360 ± 240 | ||||
| Hgb, g/dL | Baseline | Shock | Final | |||
| Control | 9.1 ± 1.4 | 10.5 ± 1.5 | 10.4 ± 0.6 | |||
| Saline | 9.2 ± 0.8 | 10.8 ± 0.9 | 9.8 ± 2 | |||
| l-NAME | 11.1 ± 1.7 | 12.1 ± 1 | 11.2 ± 1 | |||
| SMT | 11.5 ± 1.3 | 12.1 ± 0.8 | 12.5 ± 0.7 |
Values for renal oxygen consumption and extraction are shown in Figure 3. In the control group, renal oxygen extraction remained at 20%. When animals were in shock, oxygen consumption increased due to increased blood flow. In the final time period, the LPS/l-NAME group had higher utilization than at baseline and as compared with the LPS/SMT group. The decrease in RBF resulted in increased oxygen extraction in both the SMT and l-NAME groups, and this increase was higher in the l-NAME group than in the SMT group.
Fig. 3. (A) Renal oxygen consumption (normalized to baseline) increased after administration of LPS, because of the animals' hypermetabolic state and increased RBF. In the final time period, oxygen consumption was higher in the l-NAME group than at baseline or in the SMT group. (B) Renal oxygen extraction increased markedly with l-NAME even though l-NAME and SMT decreased RBF to same extent. Extraction also increased with SMT, probably because of reduced flow. *Significant versus baseline; †significant versus LPS/saline group; ‡significant versus LPS/SMT group; §significant versus control group.
[More] [Minimize]Values of NOx are shown in Table 4. We did not detect a difference in cortical NOx content between the groups. However, medullary NOx increased after LPS administration, and remained increased in the LPS/saline group. The administration of NOS inhibitors decreased medullary NOx content, with l-NAME causing a return to baseline levels. Additionally, we did not detect changes in urine NOx levels after LPS administration when the animals were in shock. However, urine NOx levels did decrease, but only in the l-NAME group at the final measurement.
| Baseline | 3 h (Shock) | 6 h (Final) | ||||||
|---|---|---|---|---|---|---|---|---|
| Urine, μmol/L | ||||||||
| Control | 507 ± 66 | 452 ± 100 | 500 ± 90 | |||||
| LPS/saline | 529 ± 100 | 416 ± 200 | 753 ± 250 | |||||
| LPS/l-NAME | 556 ± 200 | 404 ± 110 | 154 ± 95*,†,‡ | |||||
| LPS/SMT | 486 ± 100 | 490 ± 60 | 285 ± 68*,† | |||||
| Renal tissue, μg/g wet tissue weight (measured at end of experiment) | ||||||||
| Control | LPS/Saline | LPS/l-NAME | LPS/SMT | |||||
| Medulla | 1.8 ± 0.3 | 4.6 ± 0.5§ | 1.7 ± 0.8‡,§ | 2.9 ± 0.4†,§ | ||||
| Cortex | 1.5 ± 0.1 | 1.5 ± 0.1 | 1.8 ± 0.5 | 2.0 ± 0.6 | ||||
The main findings of the study were that in early endotoxemia, renal autoregulatory mechanisms increase RBF, diverting it to the medulla, and increase GFR. These were maintained with saline resuscitation. Both NOS inhibitors used in the study had similar effects hemodynamically, and both caused comparable changes to medullary and cortical blood flow. However, SMT and l-NAME had remarkably different effects on renal function, indicating that the renal effects of NO on the kidney extend beyond its effects on RBF. l-NAME decreased medullary and urinary NOx content, decreased GFR, increased FENa, and increased higher renal oxygen extraction as compared with either saline or SMT. SMT decreased medullary NOx, although not to the same extent as l-NAME. SMT did not decrease GFR, and reduced FENa.
In rodent models of endotoxemia, RBF usually declines, resulting in a reduced GFR. However, in these models, LPS administration usually causes an initial drop in MAP and CO. Observations in hyperdynamic models (which may be more relevant for human septic shock) have been contradictory, with some investigators suggesting that the renal circulation does not participate in LPS-induced systemic vasodilatation (2-9). Others have found that the renal circulation does participate in systemic vasodilatation, so that blood flow to the kidneys does not diminish (16). Additionally, it remains unclear whether NO is harmful or beneficial to the kidney in sepsis. Increased NO production may protect the renal microcirculation by counteracting local vasoconstrictors such as endothelin and angiotensin II (1) and by inhibiting thrombosis (8). Although local vasodilatation could help maintain RBF during periods of hypoperfusion, excessive NO-induced vasodilatation certainly becomes detrimental if critical perfusion pressure is reduced. Additionally, excessive NO production may become injurious by augmenting cytotoxicity through increased oxidative or nitrosative stress (17).
In this study we chose a swine model, since there are considerable morphologic and physiologic similarities between porcine and human kidneys, whereas the dog and the rat kidneys differ in some important respects (13, 18). To our knowledge, the present study is the first study to compare the effects of selective versus nonselective NOS inhibitors on RBF and GFR in endotoxemic shock in a such a model. In this study, we reproduced the hyperdyanmic circulation usually associated with human septic shock, and began resuscitation only when shock was established. We avoided excessive administration of NOS inhibitors or saline, which might have affected CO or RBF.
Physiologically, RBF is directed mainly to the cortex to optimize reabsorption of solutes and glomerular filtration. By contrast, blood flow to the medulla is low, to preserve osmotic gradients and enhance urinary concentrations. Thus, the renal medulla exists in a state of hypoxia even under normal conditions, rendering it particularly susceptible to hypoxic insults caused by changes in RBF and oxygenation. Previous studies showed that the medullary blood supply is either preserved or even increased in hemorrhagic shock despite a reduction in intravascular volume and CO, thus protecting the medulla from further hypoxia and preserving its function (19, 20). Our study confirms that corticomedullary redistribution of blood flow occurs in early endotoxemic shock despite increased CO and RBF. NO has been postulated to cause this redistribution of flow. Our study demonstrated that this redistribution from the cortex to the medulla coincided with increased medullary NOx content. However, although NOS inhibition decreased medullary NOx content, it did not affect the percent of total RBF shunted to the medulla (Figure 1B). These results would indicate that NO is not exclusively involved in corticomedullary blood flow redistribution, and it is likely that other factors play a more important role in the increased medullary blood flow in endotoxemia.
In the normal kidney, tonic NO production controls glomerular afferent arteriolar tone, and has little effect on the efferent arteriole. Thus, NO plays a major role in the regulation of glomerular filtration pressure. Changes in filtration fraction (FF) are considered to represent changes in periglomerular arterial tone. This would be achieved by efferent arteriolar constriction or afferent arteriolar dilation (1, 16). Our results indicate that the administration of LPS and the resultant shock increased FF, probably as a result of increased NO, which preserved the glomerular hydrostatic pressure and filtration despite LPS-induced hypotension. The reduced FF after administration of l-NAME suggests a greater increase in afferent than in efferent arteriolar resistance, probably owing to abolition of the vasodilatory effects of NO on the afferent arteriole. This may be one mechanism whereby l-NAME administration caused the observed decrease in GFR. On the other hand, SMT preserved afferent arteriolar vasodilatation, presumably due to its selective inhibition of NOS. Therefore, it appears that some NO is essential in the preservation of GFR in endotoxemic shock, through its effect on the afferent arteriole.
NO also participates in tubuloglomerular feedback and sodium homeostasis, since it inhibits tubular transport, resulting in both natriuresis and diuresis. This action of NO is thought to be beneficial in that it reduces tubular work and hence medullary work, since sodium reabsorption is an active process requiring energy. This would be especially protective in instances in which medullary oxygen content or oxygen flow are reduced (1). Our data show that NO, in addition to influencing RBF, clearly affects tubular sodium handling.
Since NO has been shown to decrease tubular sodium reabsorption, inhibition of NOS should decrease sodium excretion and hence should reduce FENa. As expected, this occurred after administration of SMT. However, l-NAME caused a marked increase in FENa, the opposite of what was expected (Table 2). This inability of the renal tubules to reabsorb sodium points to tubular dysfunction after the administration of l-NAME but not of SMT. The etiology of this tubular dysfunction is not readily apparent. However, we noted greater oxygen extraction with l-NAME, verging on the reported maximum of 65% to 70% for the kidneys (1). Although we did not measure medullary oxygen content or medullary energy production, the greater oxygen extraction with l-NAME may have been due to excessive medullary work that could have led to tubular injury. Alternatively, this increased oxygen extraction with l-NAME could have been the result of microvascular renal ischemia that was not detectable by our assessment of overall RBF. Such ischemia could lead to tubular dysfunction as indicated by the inability to reabsorb sodium. Certainly others have suggested that LPS-induced NO synthesis contributes to the maintenance of microvascular integrity (16), which may have been disrupted in our study by the use of a nonselective NOS inhibitor. This may be another mechanism whereby l-NAME is detrimental to renal function. Furthermore, our data show that inhibition of NO does not interfere with the renal ability to extract oxygen when flow is reduced.
Hence, reduction of renal vascular NO content with SMT may be beneficial, since it increased MAP without negatively affecting renal function. However, further reduction of NO content with l-NAME becomes detrimental, as evidenced by a decreased GFR and declining tubular function. This may account for the sometimes contradictory effects of NOS inhibition on renal function in endotoxemia. The amount of local NO produced and the dosage and type of NOS inhibitor used will determine whether the net effect is harmful or beneficial (21). However, since both of the NOS inhibitors used in our study caused similar hemodynamic effects, yet different effects on GFR and FENa, it would be impossible to discern at which point NOS inhibition becomes deleterious unless GFR, renal sodium handling, medullary NO content, or medullary oxygenation are assessed continuously, none of which assessments are practical clinically.
We acknowledge that LPS-induced shock may not represent the entire spectrum of human septic shock, in which both the time interval and exposure to cytokines are longer, resulting in tissue changes not seen in acute models. However, the swine model used in our study does mimic the vasodilatation seen in septic shock, making it a reasonable model for studying the effects of NOS inhibition. We further recognize that SMT may inhibit eNOS, albeit at high doses (10). However, the two NOS inhibitors that we used, although having similar hemodynamic effects, had different effects on renal function and medullary NO content, indicating that at the organ level, these agents had dissimilar effects on NO inhibition. Our study also did not examine the role of other vasodilators, such as prostaglandins, which may also influence RBF during endotoxemia.
Perhaps of more potential relevance was the small increase in tissue NOx in this swine model despite the profound decrease in MAP. In fact, cortical NOx content did not change at all despite the significant changes in RBF. The unaltered cortical NOx in our study was similar to findings in another study of congestive heart failure and hypertention in which cortical NOx did not increase (22). The small increase in medullary NOx is in contrast to the early and much larger increase in NO in rodent models of endotoxemia (14). Thus, caution should be used when extrapolating from rodents to higher-order mammals. Additionally, given the small increase in medullary NOx content in the swine model, it is doubtful that highly selective inhibitors of iNOS would offer advantages over SMT, but this point requires further study. We should also note that our purpose was to study the early effects of endotoxemic shock and resuscitation with NOS inhibitors, and we followed the animals in our study for 6 h. It is conceivable that at later time periods NOx content would have increased appreciably.
We conclude that NO is not solely responsible for the redistribution of corticomedullary blood flow. Moreover, in endotoxemia, changes in RBF are not fully explained by changes in CO, thus pointing to the production of local vasodilating agents such as NO, which appears to be essential in maintaining GFR and RBF. Nonselective inhibition of NOS with l-NAME was clearly detrimental to renal function and in fact resulted in ARF (decreased GFR accompanied by decreased RBF and tubular dysfunction). However, apart from increasing MAP, a selective iNOS inhibitor (SMT) offered little advantage to the kidneys as compared with saline.
The authors thank Dr. Joseph Mattana for his review of the manuscript.
Supported by Research Grant RG-015-B from the American Lung Association.
| 1. | Sadayoshi I, Carretero OA, Keishi ANitric oxide in the regulation of renal blood flow. New Horizons31995615632 |
| 2. | Kikeri D, Penell JO, Hwang KH, Jacob AI, Richman AV, Bourgoignie JJEndotoxemic acute renal failure in awake rats. Am J Physiol2501986F1098F1106 |
| 3. | Kurovsky Y, Gispaann IAntioxidants attenuate endotoxin-induced renal failure in rats. Am J Kidney Dis2519955157 |
| 4. | Doursout MF, Kilbourne RG, Hartley CJ, Chelly JEEffects of N-methyl-l-arginine on cardiac and regional blood flow in a dog endotoxin shock model. J Crit Care1520002229 |
| 5. | Booke M, Hinder F, McGuire R, Traber LD, Traber DLNitric oxide synthase inhibition versus norepinephrine in ovine sepsis: effects on regional blood flow. Shock51996362370 |
| 6. | Hinder F, Brooke M, Traber LD, Matsumoto N, Nishida K, Rogers S, Traber DLNitric oxide synthase inhibition during experimental sepsis improves renal excretory function in the presence of chronically increased atrial natriuretic peptide. Crit Care Med241996131136 |
| 7. | Spain DA, Wilson MA, Garrison RNNitric oxide synthase inhibition exacerbates sepsis-induced renal hypoperfusion. Surgery1161994322331 |
| 8. | Shultz PJ, Raij L. Endogenously synthesized nitric oxide prevents endotoxin-induced glomerular thrombosis. J Clin Invest 992;90:1718–1725. |
| 9. | Vincent J-L, Zhang H, Szabo C, Preiser J-CEffects of nitric oxide in septic shock. Am J Respir Crit Care Med161200017811785 |
| 10. | Rosselet A, Feihl F, Market M, Gnaegi A, Perret C, Liaudet LSelective iNOS inhibition is superior to norepinephrine in the treatment of rat endotoxic shock. Am J Respir Crit Care Med1571998162170 |
| 11. | Olson NC, Kruse-Elliott KT, Dodam JR. Systemic and pulmonary reactions in swine with endotoxemia and Gram-negative bacteremia. J Am Vet Med Assoc200199218701884 |
| 12. | Kowallik P, Schulz R, Guth BD, Schade A, Paffhausen W, Gross R, Heusch GMeasurement of regional myocardial blood flow with multiple colored microspheres. Circulation831991974982 |
| 13. | Fahraeus B, Lorentzo RRenal clearance studies in the pig. Urol Res219746168 |
| 14. | Afulukwe IF, Cohen RI, Zeballos GA, Iqbal M, Scharf SMSelective NOS inhibition restores myocardial contractility in endotoxemic rats; however myocardial NO content does not correlate with myocardial function. Am J Respir Crit Care Med16220002126 |
| 15. | Cohen RI, Shapir Y, Davis A, Loona R, Scharf SMComparison between selective and nonselective nitric oxide synthase inhibition and phenylephrine in normal and endotoxic swine. Crit Care Med28200032573267 |
| 16. | Thijs A, Thijs LG. Pathogenesis of renal failure in sepsis. Kidney Int 1998;53 (Suppl. 66):S34–S37. |
| 17. | Yu L, Gengaro PE, Niederberger M, Burke TJ, Schrier RWNitric oxide: a mediator in rat tubular hypoxia/reoxygenation injury. Proc Natl Acad Sci USA91199416911695 |
| 18. | Sandin R, Wahlberg J, Modig JVariations in superficial renal cortical blood flow and tissue oxygenation: an experimental porcine model. Acta Anaesthesiol Scand351991411419 |
| 19. | Thadhani R, Pascual M, Bonventure JV. Acute renal failure. N Engl J Med 1996;334:1448–1460. |
| 20. | Brezis M, Rosen SHypoxia of the renal medulla-its implication for disease. N Engl J Med43199311601178 |
| 21. | Wink DA, Mitchell JBChemical biology of nitric oxide: insights into regulatory, cytotoxic, and cytoprotective mechanisms of nitric oxide. Free Radic Biol Med251998434456 |
| 22. | Zaid A, Gurbanov K, Rubinstein I, Better OS, Hoffman A, Winaver JRegulation of intrarenal blood flow in experimental heart failure: the role of endothelin and nitric oxide. Am J Physiol2741998F766F774 |
This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org
