INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE ROLE OF ANGIOTENSIN II (ANG II) in the pathogenesis of glomerular hypertrophy and progressive renal injury is amply established in experimental and clinical renal disease (13). Accordingly, activation of the renin-angiotensin system (RAS) is believed to promote progressive renal scarring. Furthermore, therapeutic agents that attenuate the activity of the RAS offer promise in the prevention of progressive renal dysfunction (8, 10, 11, 17). In contrast, agents that promote activation of the RAS may have a deleterious effect on progressive renal injury. For example, diuretics have been widely used to manage hypertension and edema in the setting of renal disease; however, the potential adverse consequences of chronic volume manipulations coupled with the expected directional changes in the RAS have only recently been appreciated. Our laboratory recently demonstrated that chronic administration of oral furosemide increases glomerular volume and plasma renin activity in the normal rat (16). Importantly, glomerular enlargement was attenuated by simultaneous inhibition of ANG II converting enzyme (ACE), implicating the RAS in the pathogenesis of glomerular hypertrophy in this setting. These findings raise the disturbing possibility that chronic diuretic administration in vivo may promote progression of renal disease by activation of the RAS. Indeed, several clinical trials have suggested that diuretic monotherapy is associated with acceleration of renal dysfunction (9, 25).

Acute (<7 days) activation of the endogenous RAS induced via salt restriction or administration of exogenous ANG II raises glomerular capillary hydraulic pressure (P_GC) and arteriolar resistance in normal rats and rats with renal disease (14, 19, 22). Importantly, augmented P_GC, if sustained, has proven to be deleterious in the progression of renal disease (15). Nonetheless, the effects of chronic administration (>7 days) of diuretic monotherapy on glomerular hemodynamics have yet to be explored. The studies presented herein demonstrate that daily oral furosemide for more than 42 days results in an increase in P_GC and renal weight. We then examined the effects of selective antagonism of the ANG II type 1 (AT₁) receptor on glomerular hemodynamics to determine the potential contribution of ANG II to the pathogenesis of glomerular hypertension and renal hypertrophy in this setting. These latter studies suggest that the basis for our findings was secondary to an increase in ANG II activity.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Experimental design. All experiments were carried out in weanling male Munich-Wistar rats. Animals were pair fed a paste of Purina rat chow for 6-8 wk. Animals had free access to tap water. One group of animals received furosemide (n = 8), 40 mg per day, mixed into the chow; a second group received furosemide plus losartan, 5 mg per day, mixed into the chow (n = 9); and a third group received standard chow alone (n = 9). Glomerular micropuncture studies were performed after 6-8 wk of treatment as described below.

Micropuncture studies. Rats were anesthetized with pentobarbital sodium (50 mg/kg body wt), placed on a temperature-controlled operating table, and prepared for micropuncture as previously described (21). Briefly, a tracheostomy was performed, the left femoral vein was cannulated (PE-50), and a bolus of Ringer solution (0.5% body wt) was slowly administered over 15 min. Ringer solution containing 25 µCi/ml [³H]inulin was then infused at a rate of 0.5 ml · h¹ · 100 g body wt¹ for the remainder of the experiment. The femoral artery was cannulated (PE-50), and mean arterial pressure was monitored with a digital display pressure transducer. A bladder catheter (PE-50) was placed suprapubically. The left kidney was exposed by a subcostal incision, dissected free of perinephric tissue, immobilized in a plastic holder, and continuously bathed in mineral oil at 37.5°C. After a 45-min stabilization period, urine was collected in preweighed tubes for 30 min. During this interval, three to four timed (3-4 min) proximal tubular fluid collections were obtained from randomly selected superficial nephrons to determine single-nephron glomerular filtration rate (SNGFR). Samples of urine, plasma, and tubular fluid were added to scintillation cocktail, and radioactivities of the samples were measured in a liquid scintillation spectrometer (model LS 230; Beckman Instruments, Fullerton, CA).

Intratubular hydraulic pressures were measured under free-flow conditions in one group of tubules. Proximal tubular stop-flow pressures were obtained in the first surface convolutions distal to Bowman's space, after blockage of the tubular lumina with Sudan Black mineral oil. Pressures were also determined in randomly selected efferent vascular welling points. All pressure measurements were performed with a servo-null micropressure system (World Precision Instruments, New Haven, CT). P_GC was calculated as the sum of the proximal tubular stop-flow pressure and arterial colloid osmotic pressure (COP). Plasma COP was determined directly utilizing a colloid osmometer (model 4401; Wescor, Logan, UT). Blood samples were taken from efferent vascular welling points and analyzed, together with an arterial sample, for afferent (C_A) and efferent (C_E) arteriolar protein concentration using the Micro-Lowry technique. Single-nephron filtration fraction (SNFF) and single-nephron plasma flow (Q_A) were calculated as follows The glomerular ultrafiltration coefficient (K_f) was calculated using an iterative method as previously described (21). Renal afferent (R_A) and efferent (R_E) arteriolar resistances were calculated using the following relationships where P_E is efferent arteriolar pressure, MAP is mean arterial pressure, and Hct is hematocrit.

Statistics. The three treatment groups were compared using one-way analysis of variance followed by post hoc Scheffé testing. P < 0.05 was considered significant. All analysis was performed using the Statview package of statistical software (Abacus Concepts, Berkeley, CA).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Glomerular hemodynamic studies and laboratory data in rats subjected to micropuncture experiments are summarized in Table 1. Animals weighed less after treatment with furosemide, with or without losartan, compared with control animals (Table 1). In contrast, kidney weight was increased in the furosemide group compared with control rats or rats receiving the combination of furosemide and losartan (Table 1). Therefore, chronic administration of furosemide promotes renal hypertrophy in the normal rat. Moreover, these effects were attenuated by the simultaneous administration of the AT₁ receptor antagonist, losartan. These findings parallel those observed in our previous investigation (16), with the notable exception that the reduction in renal hypertrophy was minimal in rats receiving enalapril and furosemide compared with these experiments in which losartan was utilized.

Renal micropuncture studies revealed an increase in P_GC in furosemide-treated animals (52.1 ± 1.5 mmHg) compared with control rats (43.7 ± 1.5 mmHg) or rats receiving the combination of furosemide and losartan (41.3 ± 1.7 mmHg) (Table 1). These changes were accompanied by a relative increase in R_E compared with R_A (Table 1), a decrease in K_f, and an increase in SNFF (Table 1). Collectively, these hemodynamic findings are reminiscent of those obtained with salt restriction or exogenous administration of ANG II (14, 22). However, MAP at the time of micropuncture was significantly less in rats maintained on losartan and furosemide versus furosemide alone or controls (Table 1). Thus the decrease in P_GC in rats maintained on furosemide plus losartan was secondary to a fall in MAP as well as a relative decrease in R_E. Since there were no significant differences in proximal free-flow tubular pressure among these groups, the transcapillary hydraulic pressure determinations paralleled the changes in P_GC (Table 1). Unexpectedly, SNGFR and GFR were highest in rats receiving furosemide alone, although these findings failed to reach statistical significance (Table 1). Whether these latter observations represent adaptive changes in glomerular function induced by glomerular hypertrophy or were secondary to the rise in P_GC can only be inferred from these studies. The filtration fraction was highest in rats receiving furosemide alone and lowest in those animals maintained on combination therapy (Table 1). As anticipated, the ratio of R_E to R_A was highest in rats receiving diuretic alone (0.53 ± 0.08) and was lowest in rats maintained on combination therapy (0.38 ± 0.04). The latter may play an essential role in the attenuation of P_GC when coupled with the observed fall in systemic pressure in this group. There was a tendency for 24-h urine volume to be greater in either group of diuretic-treated animals; however, the results failed to achieve statistical significance (P = 0.08 vs. controls) (Table 1).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

An important finding in these investigations was the presence of increased P_GC during chronic administration of furosemide. Indeed, although several investigators have demonstrated an increase in P_GC following acute volume contraction, the present studies are unique in the demonstration of augmented P_GC after chronic administration (6 wk) of diuretic monotherapy. In addition, we observed an increase in filtration fraction, a decrease in K_f, and a relative increase in R_E versus R_A in rats receiving furosemide. Collectively, these hemodynamic changes parallel those observed after acute administration of ANG II or salt deprivation for short periods of time (<7 days), a manipulation that stimulates endogenous ANG II production (14, 22). An important finding in these studies was the attenuation of the glomerular hemodynamic changes induced by furosemide via concurrent treatment with the AT₁ receptor antagonist, losartan. The mechanism responsible for the attenuation of augmented P_GC included a fall in MAP coupled with a relative decrease in R_E. However, MAP in this group was occasionally below the range of autoregulation, which may lower P_GC independent of changes in arteriolar resistance. Thus these experiments do not prove conclusively that ANG II is mediating the intrarenal changes in hemodynamics observed in rats receiving chronic furosemide. Nevertheless, the observation that increases in P_GC occur with chronic administration of diuretic monotherapy is unique and raises the disconcerting possibility that chronic diuretic administration may engender a deleterious effect on progression of renal disease in vivo by unfavorably altering glomerular hemodynamics. Clearly, further studies in experimental renal disease are essential to investigate this hypothesis.

An important additional finding in these experiments was the induction of renal hypertrophy in rats maintained on oral furosemide for 6 wk. Previous studies from our laboratory revealed an increase in renal weight and glomerulomegaly in normal rats maintained on oral furosemide for a similar period of time (16). However, in those experiments, pretreatment with enalapril attenuated glomerular enlargement but did not abrogate renal hypertrophy. In contrast, in the present experiments the AT₁ receptor antagonist, losartan, completely arrested renal hypertrophy in rats maintained on diuretic therapy. Therefore, these studies suggest that endogenous activation of the AT₁ receptor is an essential step in the development of renal hypertrophy after chronic administration of oral furosemide and, accordingly, implies that AT₁ receptor antagonists may offer a therapeutic advantage in arresting renal hypertrophy compared with sustained inhibition of ACE. The potential mechanism underlying these differences may be inferred from emerging data which indicate that ACE-independent synthesis of ANG II represents a prominent synthetic pathway of ANG II production in rat and humans (12, 24). Thus AT₁ receptor antagonists should completely antagonize the effects of ANG II on the AT₁ receptor, whereas sustained inhibition of ACE would not. Other pharmacodynamic differences between ACE inhibition and antagonism of AT₁ receptors include augmented synthesis of bradykinin in the former and activation of ANG II type 2 (AT₂) receptors in the latter (23). Whether these actions also contribute to the disparate findings we observed in enalapril-treated animals (16) compared with losartan treatment cannot be answered by the present investigations. Nonetheless, the implications of this observation in the context of the treatment of progressive renal insufficiency are intriguing, since renal hypertrophy frequently accompanies progression of renal disease (2, 16).

Interestingly, several recent clinical investigations have suggested that diuretic monotherapy may accelerate progressive renal injury in the setting of hypertension (7, 9). For example, thiazide diuretics have been found to accelerate nephrosclerosis produced by nitro-L-arginine methyl ester (i.e., L-NAME) in spontaneously hypertensive rats (18). The adverse structural and functional consequences of thiazide treatment correlated with elevated P_GC. No evidence of glomerulosclerosis or tubulointerstitial fibrosis has been demonstrated in our experiments thus far; however, we have limited our studies to a relatively short time period (6-8 wk) in animals without coexisting renal disease or loss of renal mass.

Interestingly, there was a trend for an increase in SNGFR in rats receiving furosemide alone, despite the fact that these findings failed to achieve statistical significance. Statistics notwithstanding, this ostensive paradox could, in part, be ancillary to an alteration in tubuloglomerular feedback. For example, furosemide interrupts the sensing step at the macula densa, engendering a fall in R_A, which in turn would augment Q_A and SNGFR. Although there was a slight fall in R_A in the furosemide-treated animals, total renal vascular resistance increased slightly and Q_A fell accordingly. In contrast, the trend toward an increase in SNGFR could be incident to the rise in P_GC observed under these conditions. Interestingly, chronic but not acute diuretic monotherapy elicits an increase in renal plasma flow and GFR in humans (25). A biphasic response has generally been observed, e.g., initiation of diuretic therapy is accompanied by an early reduction in renal plasma flow and GFR; however, after several months of therapy, renal plasma flow and GFR either increase or return to baseline (25).

Diuretics have been included in multiple drug regimens to lower blood pressure in the rat model of renal ablation (27). Although such regimens typically lower systemic blood pressure, they do not consistently lower P_GC or alter progression of renal disease (1, 4). In contrast, similar degrees of blood pressure reduction with sustained inhibition of ACE ameliorates progressive renal dysfunction and lowers P_GC (2-4). These observations raise the attractive possibility that multidrug therapy that includes a diuretic agent may blunt the expected decrease in P_GC otherwise elicited by a fall in systemic blood pressure unless these regimens are accompanied by an agent that antagonizes the synthesis or action of ANG II. Admittedly, the results of other investigators would dispute this hypothesis. For example, Yoshida et al. (27) demonstrated a salutary effect of reserpine/hydralazine/thiazide on glomerular morphology in rats with renal ablation, which parallels the findings obtained in experiments using ACE inhibitors. Nevertheless, most studies have been unable to demonstrate a beneficial effect of multiple drug combinations unless accompanied by an agent that blocks ANG II action or synthesis (2, 3, 5, 27). It is plausible that the net effect of unique antihypertensive regimens on the behavior or synthesis of ANG II may underlie the potential of these regimens to alter progressive renal dysfunction in evolving chronic renal disease.

In summary, chronic administration of furosemide was accompanied by an increase in P_GC in the normal rat, most likely via activation of the RAS. The hemodynamic pattern observed in rats receiving sustained administration of oral furosemide was reminiscent of the effects of acute infusion of ANG II or salt deprivation for short periods of time. Although the AT₁ receptor antagonist, losartan, abrogated the hemodynamic changes accompanying furosemide administration, we cannot conclusively determine that these actions are secondary to intrinsic blockade of the AT₁ receptor. Nevertheless, the attenuation of P_GC was likely incident to a fall in MAP coupled with a relative reduction in R_E since a relative decrease in R_E is essential to elicit a fall in P_GC when MAP is reduced within the range of autoregulation (1, 4). These hemodynamic effects are reminiscent of those obtained following blockade of the synthesis of ANG II.

In addition, losartan completely arrested the development of renal hypertrophy induced by diuretic monotherapy. This contrasts with previous studies from our laboratory utilizing concurrent treatment with furosemide and the ACE inhibitor, enalapril (16). Although these contrasting findings could be related to variances in dose response, an alternative hypothesis is that the AT₁ receptor antagonists offer a selective advantage in arresting renal hypertrophy, perhaps by antagonizing ACE-independent mediated synthesis of ANG II (12). Regardless, persistent or intermittent activation of the RAS via diuretic monotherapy may prove to be deleterious in the treatment of progressive renal disease. Accordingly, it will be necessary to define the effects of diuretic monotherapy on glomerular hemodynamics, glomerular growth, and glomerulosclerosis in experimental models of renal disease.