Effect of Salt Intake and Inhibitor Dose on Arterial Hypertension and Renal Injury Induced by Chronic Nitric Oxide Blockade

Methods

One hundred and ninety-six adult male Munich-Wistar rats, obtained from an established colony at the University of São Paulo and weighing 239 to 295 g, were used in this study. Rats received tap water ad libitum and commercially available rat chow (Harlan Teklad) containing 25% protein and sodium chloride at either 0.06% sodium (LS diet), 0.5% sodium (SS diet), or 3.2% sodium (HS diet). All experimental procedures were conducted in accordance with our institutional guidelines.

Results

Pressure-natriuresis and pressure-diuresis lines obtained in groups SS0, SS3, and SS25 are shown in Fig 1⇓. In all groups, urine flow and sodium excretion rate were best described as linear functions of renal perfusion pressure. Fitting to nonlinear functions such as two-phase exponential growth was also obtained, although the r values were lower and the differences between groups less apparent than with linear regression. In rats treated with low-dose L-NAME (group SS3), the slope of these regression lines was significantly lower than in untreated rats (group SS0). In rats receiving high-dose L-NAME (group SS25), the pressure-natriuresis and pressure-diuresis lines were both shifted to the right compared with SS0. However, their slopes were not significantly different compared with untreated rats. In rats receiving HS diet, the pressure-natriuresis and pressure-diuresis lines exhibited exactly the same behavior as in rats given an SS diet (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Pressure-natriuresis (left) and pressure-diuresis (right) lines at 2 weeks of L-NAME treatment in rats receiving SS (0.5% NaCl, wt/wt) diet. U_NaV indicates urinary excretion of sodium; V, urine volume; and RPP, renal perfusion pressure.

Renal and systemic hemodynamic parameters measured at 10 to 15 days of treatment in LS and HS groups are shown in the Table⇓. Body growth was similar among groups. Left kidney weight was not significantly altered by either L-NAME treatment or dietary salt intake, except for group HS3, which had slightly larger kidneys. Mean arterial pressure was not significantly affected by dietary sodium in the absence of L-NAME treatment. However, sodium intake was decisive for the establishment of arterial hypertension in rats receiving low-dose L-NAME. LS diet significantly attenuated arterial hypertension in rats receiving high-dose L-NAME, although it failed to normalize BP in these rats. Low-dose L-NAME (groups LS3 and HS3) did not significantly affect glomerular filtration rate. However, glomerular filtration rate was moderately decreased in groups LS25 and HS25, although only in the latter did the differences with the respective control achieve statistical significance. Filtration fraction was similar among groups, except for groups LS25 and HS25. Accordingly, renal plasma flow was lowered in these two groups compared with their respective controls. Renal vascular resistance was insensitive to salt intake in untreated rats. Low-dose L-NAME treatment increased renal vascular resistance by 41% in rats receiving an LS diet and by 48% in those given an HS diet. Vasoconstriction was more severe in rats receiving high-dose L-NAME.

Long-term treatment with L-NAME was associated with normal body growth in all groups except group HS25, in which body weight (262±9 g) was low compared with rats receiving either low-dose L-NAME and salt overload (HS3, 307±11 g) or high-dose L-NAME and salt restriction (LS25, 322±5 g). No limitation to body growth was seen in the other groups. The evolution of tail-cuff pressure during 6 weeks of treatment is shown in Fig 2⇓. Tail-cuff pressure increased steadily in group HS3 compared with untreated controls, reaching 152±8 mm Hg after 6 weeks of treatment, compared with 124±4 mm Hg in group HS0 (P<.05). Salt restriction prevented BP elevation in group LS3, characterizing the development of hypertension at this L-NAME dose as salt dependent. Rats receiving high-dose L-NAME and HS intake (group HS25) developed more severe and rapidly progressive hypertension, reaching 170±10 mm Hg after 6 weeks of treatment. Since the mortality rate in this group was 20%, as opposed to 0% in the remaining groups, and since it is likely that the BPs of the rats that died would have been even higher at 6 weeks, the tail-cuff pressure measured at the end of the study may have underestimated the severity of the hemodynamic derangement occurring in this group. Salt restriction attenuated but did not prevent hypertension in group LS25 (145±4 versus 118±1 mm Hg in group LS0, P<.05).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Tail-cuff pressure as a function of time in rats receiving no treatment, low-dose L-NAME (3 mg/kg per day), or high-dose L-NAME (25 mg/kg per day) and varying amounts of dietary sodium.

As expected, plasma renin levels were persistently higher in rats receiving an LS diet than in those given salt excess (Fig 3⇓). Low-dose L-NAME treatment had little effect on PRA in rats receiving an HS diet but promoted a steady decline in renin levels in rats given an LS diet. High-dose L-NAME treatment depressed PRA even faster in rats receiving an LS diet. In rats receiving an HS diet, high-dose L-NAME further lowered circulating renin levels at 3 weeks but markedly increased PRA at 6 weeks of treatment.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

PRA as a function of time in rats receiving no treatment, low-dose L-NAME (3 mg/kg per day), or high-dose L-NAME (25 mg/kg per day) and either HS or LS diet. AI indicates angiotensin I.

Albumin excretion rate (Fig 4⇓) was increased at 6 weeks of treatment in group HS3, reaching 30.2±18.3 mg/24 h, although this variation did not reach statistical significance compared with groups HS0 (1.3±0.3 mg/24 h) and LS3 (1.2±0.2 mg/24 h). Urinary albumin excretion rate was numerically increased in group SS25, reaching 19.9±7.0 mg/24 h, although this value did not differ statistically from those observed in groups LS25 (2.7±0.6 mg/24 h) and SS3 (1.3±0.3 mg/24 h). In group HS25, however, albuminuria rose steadily from the beginning of the study, reaching 168±22 mg/24 h at 6 weeks (P<.05 versus groups HS0, HS3, and LS25).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Urinary albumin excretion rate as a function of time in rats receiving no treatment, low-dose L-NAME (3 mg/kg per day), or high-dose L-NAME (25 mg/kg per day) and varying amounts of dietary sodium.

Fig 5⇓ shows the three most frequently occurring modalities of renal injury encountered in L-NAME–treated rats.²⁷ Collapsed glomeruli (Fig 5⇓, top) appeared reduced in size, with prominent thickening of the basement membrane and extreme narrowing of capillary loops, without adhesions to Bowman’s capsule. Glomerulosclerosis (Fig 5⇓, middle) was defined as tuft deposition of PAS-positive material with loop occlusion and/or adhesions to Bowman’s capsule. Interstitial expansion (Fig 5⇓, bottom) consisted of focal enlargement of interstitial tissue, with infiltration by fibroblasts and deposition of collagen-like material, frequently associated with tubular atrophy and vacuolization. Confluence of these areas was seen in more severely affected rats. Fig 6⇓, left, depicts the frequency of glomerular collapse in the groups studied. In group SS25, 15.9±4.3% of glomeruli presented with tuft collapse (P<.05 versus SS0 and SS3), whereas sodium excess increased the frequency of these glomeruli in group HS25 (28.5±3.4%, P<.05 versus groups HS0 and HS3). Sodium restriction attenuated glomerular collapse in group LS25 (7.0±1.4%, P<.05 versus HS25 and P>.1 versus LS0). Parallel results were obtained regarding glomerulosclerosis (Fig 6⇓, middle) and interstitial expansion (Fig 6⇓, right). Focal and segmental necrosis of the glomerular tuft appeared in nearly 1.2% of glomeruli examined in group HS25 but was virtually absent in the other groups. Arteriolar wall thickening was observed in group HS3 and, more often, in group HS25. Fibrinoid necrosis of the arteriolar wall was seen only in group HS25.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Top, Global collapse of the glomerulus, with widespread occlusion of capillary loops and thickening of basement membrane. PAS in 3-μm-thick section, magnification ×160. Middle, Segmental sclerosis of the glomerular tuft, consisting of deposition of PAS-positive material with occlusion of capillary loops and adhesion to Bowman’s capsule. PAS in 3-μm-thick section, magnification ×160. Bottom, Focal expansion of the interstitial area. Fibroblast nuclei, matrix expansion, and tubular atrophy are readily visible. Masson’s trichrome in 3-μm-thick section, magnification ×100.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Bar graphs show frequency of structural abnormalities in rats receiving L-NAME treatment for 6 weeks. Open bars represent LS diet; shaded bars, SS diet; and solid bars, HS diet. Left, collapsed glomeruli; middle, glomerulosclerosis; and right, relative interstitial area. The following comparisons were statistically significant: left, glomerular collapse: HS25 vs HS0, SS25 vs SS0, HS25 vs SS25, HS25 vs LS25, SS25 vs LS25, HS25 vs HS3, and SS25 vs SS3; middle, glomerulosclerosis: HS25 vs HS0, HS25 vs SS25, HS25 vs LS25, and HS25 vs HS3; right, interstitial area: HS25 vs HS0, HS25 vs SS25, HS25 vs LS25, and HS25 vs HS3.

Discussion

The present observations indicate a rather complex pathogenetic interaction between salt intake and chronic NO inhibition. At low L-NAME doses, the development of hypertension depended entirely on concomitant dietary salt overload, an observation in agreement with those made by Salazar and coworkers,³ who showed induction of salt-dependent hypertension in dogs by chronic infusion of subpressor doses of L-NAME, and by Tolins and Shultz,⁴ who showed marked aggravation of hypertension induced by low-dose L-NAME in rats also receiving dietary salt excess. In addition, these results are consistent with those reported by Chen and Sanders,⁶ who suggested that spontaneous NO deficiency may underlie the salt-dependent hypertension observed in Dahl rats. Collectively, these studies support the view that low-grade NO inhibition subtly impairs the renal ability to excrete sodium, predisposing the animals to the development of hypertension on salt overload. This effect was predicted by the pressure-natriuresis and pressure-diuresis relationships observed in the present study. Regardless of salt intake, the slopes of the regression lines were decreased compared with untreated rats, indicating that (1) relatively small reductions in salt intake would be expected to promote substantial BP lowering and (2) salt overload is necessary for hypertension to become manifest. Previous studies showed that the pressure-natriuresis relationship is blunted in rats and dogs acutely treated with NO inhibitors.^{7 8} The present observations indicate that this functional abnormality can persist for at least 2 weeks of NO inhibition and contribute to the maintenance of hypertension in salt-loaded animals.

Salt sensitivity was less pronounced at higher L-NAME doses. Although massive salt overload did elevate final BP by 25 mm Hg compared with rats receiving an SS diet, reduction of salt intake to very low levels had little effect on BP. These results are in keeping with those reported previously in rats treated with relatively high doses of L-NAME^{10 11 12} and characterize a relatively salt-insensitive hypertension in this setting. Again, these effects of salt intake were predicted by the pressure-natriuresis/diuresis relationship obtained in this group. As in rats treated with low-dose L-NAME, the regression lines were shifted to the right, indicating that higher BPs were required for excretion of a given amount of sodium and water. However, the slopes of these lines were indistinguishable from those of control, indicating a resetting of the pressure-natriuresis relationship and predicting (1) no more salt sensitivity in these rats than in untreated controls and (2) little BP effect of salt deprivation compared with SS intake, as effectively observed in this group.

Although the analysis of the pressure-natriuresis/diuresis lines helps the understanding in phenomenological terms of the complex interplay between salt intake and different levels of NO inhibition in the generation of hypertension, the mechanistic aspects involved in this setting are unclear. Administration of low-dose L-NAME in the present study promoted a 40% increase in renal vascular resistance, whereas higher-grade NO inhibition increased renal vascular resistance by more than twofold. Although these hemodynamic effects may contribute to the discordant responses of the pressure-natriuresis relationship to differing doses of L-NAME, they may also reflect simple autoregulation. On the other hand, previous studies showed that selective NO production at the inner medulla in response to variations in renal perfusion pressure may strongly influence sodium excretion by modifying local blood flow^{9 31 32} and/or by directly influencing tubular sodium transport.³³ Further investigation is necessary for determination of whether L-NAME has a dose-dependent effect on these mechanisms, thus explaining the differing functional effects of high and low doses of the inhibitor. Additional limitation to renal sodium excretion may have originated from an enhanced renal sympathetic influx,^{21 22 23} as well as from activation of the renin-angiotensin system.^{10 19}

Of particular interest was the behavior of plasma renin levels in these experiments. As expected, PRA was depressed by sodium excess and augmented by sodium restriction in rats not receiving L-NAME, compared with levels previously observed in rats receiving an SS diet.^{1 27 34} Low doses of L-NAME promoted a progressive decline in PRA in both salt-depleted and salt-loaded rats, whereas the initial depression of renin levels was even faster with high L-NAME doses, suggesting that whatever the level of salt repletion, renin production at least partially depends on NO availability. These results are consistent with several recent studies from this laboratory and elsewhere in which chronic NO inhibition, at least in the initial weeks of treatment, was reported to lower circulating renin levels.^{1 10 11 27 34} They are also in accordance with the observed depression of renin levels by acute L-NAME treatment in vivo^{35 36} and in isolated perfused rat kidneys^{37 38} and by the observation that endothelium-derived NO enhances renin secretion in cultivated juxtaglomerular cells.³⁹ Later in the course of treatment, renin levels were markedly and paradoxically elevated in rats receiving high-dose L-NAME and dietary sodium overload, confirming previous observations of this laboratory.^{1 27} In these rats, hyperreninemia is unlikely to reflect a primary effect of NO on renin production. Rather, enhanced renin production appears to result from the profound structural derangement of the renal parenchyma, especially the extensive glomerular ischemia associated with this model. Hyperreninemia is likely to have contributed to aggravation of hypertension in this group at the late stages of the study. However, the initiation and maintenance of hypertension at the initial phase of this process appear to bear no direct relationship to the circulating renin levels, which were elevated in the face of normal BPs in LS rats receiving low-dose L-NAME and depressed in HS rats given the same dose of the inhibitor that nevertheless were always hypertensive. Since administration of renin-angiotensin antagonists has been shown to largely prevent hypertension in this model,^{1 27 40} the beneficial effects of these drugs must depend on some aspect of the renin-angiotensin system not directly related to circulating renin levels.

The renal parenchymal injury observed at 6 weeks in L-NAME–treated rats not subjected to salt restriction may have contributed additionally to the development of hypertension. As described previously,^{1 27} the most frequent structural abnormality encountered in these rats was a global collapse of the glomerular tuft, consistent with severe glomerular ischemia. Correa-Rotter and coworkers⁴¹ reported evidence that these glomeruli, also found in the renal ablation model,^{41 42} may release excessive amounts of renin, which in turn might promote the development of both systemic hypertension⁴³ and renal structural injury.⁴² Exacerbation of this abnormality by salt overload coincided with aggravation of systemic hypertension, especially at later stages, suggesting a pathogenetic role for inadequately “high” circulating and/or local renin levels. Sodium restriction prevented both hypertension and glomerular collapse, further suggesting a pathogenetic association between these two phenomena. The mechanisms whereby salt intake levels might influence the development of glomerular collapse are unclear. NO inhibition and sodium overload might interact to promote severe afferent constriction, an abnormality shown in this and previous studies²⁷ to associate with glomerular collapse. Sodium overload has been shown to increase cytosolic free calcium concentrations in erythrocytes.⁴⁴ Although direct experimental evidence in this regard is lacking, excess sodium might influence smooth muscle cells similarly, thus favoring vasoconstriction. Conversely, sodium restriction might promote smooth muscle relaxation by lowering cytosolic calcium concentration, thus exerting a protective effect akin to that previously obtained in this model with chronic nifedipine treatment.⁴⁵ Alternatively, glomerular collapse may have resulted directly from systemic hypertension. Although this hypothesis would also explain the observed prevention of glomerular collapse by salt restriction, it must be noted that hypertension as severe as that observed in this study is not associated with glomerular collapse in models such as the spontaneously hypertensive rat,⁴⁶ renal ablation,⁴⁷ or the Milan hypertensive rat.⁴⁸

The present study confirms previous observations of this laboratory that renal interstitial fibrosis is a prominent feature of the structural derangement observed in rats receiving chronic L-NAME treatment and that this process is worsened by simultaneous sodium overload.²⁷ In addition, sodium restriction largely prevented renal interstitial expansion in L-NAME–treated rats, regardless of the dose used. Given the antimitogenic properties of NO,^{49 50 51} renal interstitial enlargement in this model may be related at least in part to abnormal cell proliferation. Sodium overload has been shown to elevate cytosolic calcium concentration,⁴⁴ whereas the latter has been consistently linked to activation of mitogenesis.^{52 53 54} Thus, it is possible that in the chronic NO blockade model, sodium overload aggravates interstitial expansion by enhancing cell proliferation, whereas sodium restriction has the opposite effect. In the absence of direct experimental evidence, however, this possibility must remain speculative.

In summary, the present observations indicate that chronic NO inhibition promotes salt-dependent hypertension. However, salt dependence can be demonstrated only at a relatively low dose of the inhibitor. Hypertension resulting from higher doses of the inhibitor is aggravated by dietary salt overload but cannot be prevented by salt restriction. Whatever the dose of NO inhibitor used, sodium restriction largely prevents the renal parenchymal injury associated with this model. Elucidation of the intimate mechanisms dictating the pathogenetic interaction between chronic NO inhibition and salt intake levels awaits further investigation.

Selected Abbreviations and Acronyms