| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Hypertension. 2007;49:932.)
© 2007 American Heart Association, Inc.
Original Articles |
From the Baker Heart Research Institute (S.L.B., G.A.H., G.W.L.), Melbourne, Australia; and the Department of Physiology (R.G.E.), Monash University, Melbourne, Australia.
Correspondence to Roger G. Evans, Department of Physiology, PO Box 13F, Monash University, Victoria 3800, Australia. E-mail Roger.Evans{at}med.monash.edu.au
| Abstract |
|---|
|
|
|---|
Key Words: kidney circulation renal medulla renal sympathetic nerves reninangiotensin system sodium excretion
| Introduction |
|---|
|
|
|---|
The role of the sympathetic nervous system in secondary hypertension remains controversial. Sympathetic vasomotor drive may be increased in hypertension induced by chronic infusion of angiotensin II (Ang II; Ang IIinduced hypertension)35 and in 2 kidney, 1 clip (2K1C) hypertension,6,7 because depressor responses to sympatholysis or ganglionic blockade are augmented. Furthermore, muscle SNA and total norepinephrine spillover are increased in human renovascular hypertension.8 Renal denervation can delay or blunt 2K1C and Ang IIinduced hypertension,911 suggesting a contribution of increased RSNA or increased renal responsiveness to RSNA. However, RSNA is not necessarily increased in secondary hypertension,6,12,13 and responsiveness of total renal blood flow (RBF) to electrical stimulation of the renal nerves (RNS) in vivo is not augmented.14,15
Neuroeffector gain is the responsiveness of renal neuroeffectors (eg, cortical and medullary perfusion, glomerular filtration rate [GFR], sodium excretion, and renin release) to given levels of RSNA or RNS. We hypothesized that the gains of specific neuroeffector mechanisms are increased in secondary hypertension. Therefore, rabbits with stable hypertension were anesthetized and instrumented for measurement of neuroeffector responses. Graded RNS was applied and renal norepinephrine spillover was measured to enable dissection of changes in postjunctional responsiveness of neuroeffectors from changes in postganglionic nerve function.
Changes in sympathetic neuroeffector function in secondary hypertension may differ depending on the experimental setting of the kidney. Renal catecholamine content16 and response of RBF to RNS14 are diminished in renovascular hypertension. In contrast, renal innervation density and neurally evoked vasoconstriction of arcuate arteries are augmented in Ang IIinduced hypertension in rats.17 Therefore, we studied both the clipped and nonclipped kidneys in 2K1C hypertension and the kidneys of rabbits with Ang IIinduced hypertension. This allowed the chronic influences of Ang II to be dissected from the effects of unilateral renal artery stenosis on the clipped and nonclipped kidney.
| Methods |
|---|
|
|
|---|
Terminal Experiment
Surgery
Four to 6 weeks after initial surgery, rabbits were anesthetized with pentobarbital (90 to 150 mg plus 30 to 50 mg h1 IV; Sigma) and artificially ventilated. Extracellular fluid volume was maintained by intravenous infusion (0.18 mL kg1 min1) of a 4:1 mixture of compound sodium lactate and polygeline/electrolyte solution.19 The left kidney was exposed via a flank incision and placed in a cup secured to the operating table. Catheters were placed in the ear arteries and veins, left renal vein,20 and left ureter. The (left) renal nerves were cut, and the distal end placed across a stimulating electrode. The kidney was instrumented for measurement of RBF (transit-time ultrasound) and cortical and medullary laser Doppler flux (CLDF and MLDF, respectively).19
Measurements
MAP, heart rate, RBF, CLDF (corrected for background 6±1 U), and MLDF (background 16±1 U) were acquired as 2-s averages. GFR was measured as clearance of [14C]-inulin (10 µCi plus 45 nCi kg1 min1; Perkin Elmer). Renal norepinephrine spillover measurements were facilitated by infusion of 90 nCi kg1 mL1 ring labeled [3H]-norepinephrine (Perkin Elmer) and calculated as described previously.21 PRA overflow (ng min1) was defined as the PRA arteriovenous difference (ng of angiotensin I mL1 per 60-minute incubation) multiplied by renal plasma flow (mL min1).
Experimental Protocol
Before sectioning the renal nerves, we tested responses to 20 minutes of hypoxia, but these results will not be described herein. The effects of hypoxia were completely reversible, so all of the variables reported herein returned to their control levels by 20 minutes after hypoxia. Renal denervation reduced renal norepinephrine spillover and PRA overflow but did not significantly alter renal hemodynamics or excretory function. RNS was applied22 for 20-minute periods at 0, 0.5, 1, and 2 Hz (supramaximal voltage; 5 to 9 V), followed by 3-minute periods at 4 and 8 Hz. A final 20-minute control period followed a 30-minute recovery period. Urine was collected from the left ureter during the final 15 minutes of each period. At the midpoint of each urine collection, arterial (3 mL) and renal venous (2 mL) blood samples were taken for measurement of renal excretory function,19 PRA, and norepinephrine spillover. Blood was replaced from a donor rabbit.
Statistics
Data are expressed as mean±SEM. Split-plot repeated-measures ANOVA allowed for within- and between-subject contrasts.12 PRA and norepinephrine values were log10 transformed to reduce heteroskedacity. Type 1 error was controlled with Bonferroni and GreenhouseGeisser corrections.23 Lines of best fit were determined by the least-products method.24 Two-sided P<0.05 was considered statistically significant.
| Results |
|---|
|
|
|---|
0.05). By 4 weeks after commencing Ang II infusion or clipping the left or right renal artery, MAP had increased by 24±6%, 17±1%, and 22±6%, respectively (Figure 1), but heart rate had not changed significantly. Hematocrit increased similarly in all groups of hypertensive animals, to average 41.2±0.8% 4 weeks after surgery. PRA had decreased by 1.5±0.7 ng mL1 after 4 weeks of Ang II infusion but did not change significantly after renal artery clipping (Figure 1).
|
Baseline Variables in Anesthetized Rabbits
Before RNS, these were similar across the 4 groups with only a few exceptions (Table). RBF, GFR, urine flow, and sodium excretion were, respectively, 35%, 44%, 74%, and 64% less in the clipped kidney of 2K1C rabbits than in sham-operated rabbits. Urine flow was 85% greater in the nonclipped kidney of 2K1C rabbits than in sham-operated rabbits. Basal PRA overflow was similar in the clipped kidney and the kidney in sham-operated rabbits but was considerably less in the nonclipped kidney in 2K1C rabbits and in the kidney of Ang IItreated rabbits. After recovery from RNS, all of the variables returned to control levels except for CLDF and hematocrit, which were slightly (13±3% and 2.5±0.6%, respectively) less than control in all 4 of the experimental groups.
|
Effects of RNS
RNS produced frequency-dependent reductions in RBF, CLDF, and MLDF (P<0.001; Figure 2) but little change in MAP (P=0.9). Responses to brief high-frequency RNS (4 to 8 Hz) did not differ significantly according to group, but responses to prolonged stimulation at lower frequencies (0.5 to 2 Hz) did. RBF was reduced less in the nonclipped kidney of 2K1C rabbits, and CLDF was reduced less in the clipped kidney than in the kidney of sham-operated rabbits. RNS at 2 Hz reduced RBF and CLDF by 58% and 59%, respectively, in sham-operated rabbits but by 43% and 52%, respectively, in the nonclipped 2K1C kidney and by 51% and 36% in the clipped 2K1C kidney. RBF and CLDF responses to RNS in Ang IItreated rabbits resembled those in sham-operated rabbits. The reductions in MLDF were greater in Ang IItreated rabbits and the nonclipped kidney of 2K1C rabbits than in sham-operated controls (P<0.05). For example, 2-Hz stimulation reduced MLDF in Ang IItreated rabbits (by 25%) and in the nonclipped kidney in 2K1C rabbits (by 15%) but not in sham-operated rabbits (Figure 2).
|
RNS reduced GFR, urine flow, and sodium excretion similarly in Ang II-treated rabbits and sham-operated rabbits (Figure 3). However, reductions in sodium excretion in both the clipped and nonclipped kidney in 2K1C rabbits were less than in sham-operated rabbits. Urine flow followed a similar pattern, although differences between the nonclipped kidney in 2K1C rabbits and sham-operated rabbits were not statistically significant (P=0.08). Reductions in GFR tended to be less in the nonclipped kidney in 2K1C rabbits than in sham-operated rabbits (P=0.06).
|
RNS increased renal PRA overflow and renal norepinephrine spillover (Figure 4). Increases in renal norepinephrine spillover tended to be blunted in the clipped kidney of 2K1C rabbits (P=0.08) but were otherwise similar in all 4 of the groups. Increases in renal PRA overflow were greater in the sham-operated rabbit kidney (+439±79 ng min1 at 2 Hz) than in the clipped (+134±78 ng min1 at 2 Hz) and nonclipped (+156±85 ng min1 at 2 Hz) kidney of 2K1C rabbits or the Ang IItreated rabbit kidney (+211±112 ng min1 at 2 Hz).
|
There were strong predictive relationships between (Log10) renal norepinephrine spillover and RBF, CLDF, GFR, urine flow, and sodium excretion (P<0.001 in all cases). The slopes of these relationships were indistinguishable in the 4 groups of rabbits (P
0.12). Renal PRA overflow correlated strongly with (Log10) renal norepinephrine spillover (P<0.001), but the slope of this relationship was less in hypertensive rabbits than in sham-operated rabbits (P<0.001; Figure 5). Across all 4 of the groups, there was no statistically significant relationship between MLDF and (Log10) renal norepinephrine spillover (P=0.38). However, a highly significant relationship was present in hypertensive rabbits (P=0.003) but not sham-operated rabbits (P=0.10). Thus, the slope of this relationship differed in hypertensive rabbits compared with sham-operated rabbits (P=0.001; Figure 5).
|
Postmortem Measurements
The left ventricle of sham-operated rabbits (1.31±0.03 g kg1) weighed less than that of Ang IItreated rabbits (1.56±0.04 g kg1) and of rabbits with a clip on the right or left renal artery (average 1.47±0.03 g kg1; P<0.01 for comparison of hypertensive and sham-operated rabbits). Kidney weight was similar in sham-operated (3.37±0.11 g kg1) and Ang IItreated (3.61±0.10 g kg1; P=0.21) rabbits. However, in 2K1C hypertensive rabbits, the clipped kidney had atrophied (2.71±0.21 g kg1) and the nonclipped kidney had undergone hypertrophy (3.95±0.15 g kg1; P<0.01 compared with sham-operated).
| Discussion |
|---|
|
|
|---|
2 Hz) were enhanced in kidneys of hypertensive rabbits. Prolonged RNS at 2 Hz did not reduce MLDF in sham-operated rabbits, consistent with our previous findings,25 but reduced MLDF in all groups of hypertensive rabbits. However, RNS-induced reductions in GFR, urine flow, and sodium excretion were not enhanced in kidneys from hypertensive animals. Indeed, these responses also appeared to be blunted, at least in the nonclipped kidney in 2K1C hypertension. We also found that RNS-induced renin overflow was greatly attenuated in all of the hypertensive groups studied.
Thus, of all of the renal neuroeffectors we studied, only neural control of medullary perfusion appears to have increased gain and is, therefore, the only one that could be considered prohypertensive. Reductions in medullary perfusion induced by RNS are likely mediated chiefly by constriction of juxtamedullary arterioles and/or outer medullary descending vasa recta, which are densely innervated.26 The medullary circulation is pivotal in long-term blood pressure control, chiefly through its influence on tubular sodium reabsorption,27 such that increased sensitivity of medullary vascular elements to RSNA could contribute to development of hypertension. Similar responses of MLDF to RNS were observed in the kidneys of rabbits with Ang IIinduced hypertension and in both the clipped and nonclipped kidney in 2K1C hypertension, indicating that factors common to all 3 of the conditions are responsible for enhancing responsiveness of the medullary circulation to RNS. These putative factors are unlikely to include hypertension, per se, because MAP under anesthesia was similar in all 4 groups of rabbits, and acute changes in renal perfusion pressure have little effect on the responses of MLDF to RNS.28,29 There is good evidence, however, that the activity of the intrarenal reninangiotensin system is increased in Ang IIinduced hypertension and also in both the clipped and nonclipped kidney in 2K1C hypertension.30 This is likely to have at least 2 actions that could enhance neurally evoked vasoconstriction in the medullary circulation. First, endogenous Ang II appears to enhance neurally evoked vasoconstriction in the medullary circulation through activation of both Ang II type 1 receptors31 and Ang II type 2 receptors.32 Second, Ang II increases superoxide production and, thus, reduces NO bioavailability.33 Endogenous NO acts to blunt responses of MLDF to low- (
2 Hz) but not high- (
4 Hz) frequency RNS,31,34,35 so reduced NO bioavailability would be expected to enhance responses of medullary perfusion to RNS at low but not high frequencies, just as we observed in the hypertensive animals in the current study. Consistent with this notion, oxidative stress blunts NO-dependent attenuation of sympathetic vasoconstriction in the hindlimb of rats with angiotensin-dependent and 2K1C hypertension.36
Responses of the cortical circulation and of GFR, urine flow, and sodium excretion to RNS were not enhanced in hypertension. Indeed, in the case of 2K1C hypertension, the gains of these renal neuroeffector mechanisms were, if anything, reduced compared with normotensive rabbits. These observations are consistent with those of Fink and Brody,14 who demonstrated impaired neurogenic control of the renal vasculature in 2-kidney 1-wrap hypertension in rats. RNS-induced reductions in RBF, CLDF, GFR, urine flow, and sodium excretion in rabbits with Ang IIinduced hypertension were indistinguishable from those in normotensive rabbits, as was found previously for RBF alone.15 Collectively, these data indicate that changes in the sensitivity of cortical perfusion and sodium excretion to RNS in secondary hypertension depend on the precise experimental setting of the kidney. Exposure of the kidney to elevated circulating levels of Ang II, per se, appears to have little impact, but clipping one renal artery appears to set in train changes in the gains of these renal sympathetic neuroeffectors in both the clipped and unclipped kidney.
The fact that RNS-induced reductions in sodium excretion were not enhanced in hypertensive animals may seem at odds with the enhanced response of MLDF to RNS in hypertension, because reduced medullary perfusion should have an antinatriuretic effect.27 However, the effects of changes in medullary perfusion on salt and water reabsorption likely depend on slowly evolving changes in renal interstitial hydrostatic pressure.37 The relatively short clearance periods (15 minutes) in the current study may, therefore, not have allowed full expression of the effects of altered medullary perfusion on renal excretory function. Nevertheless, prolonged reductions in medullary perfusion induced by ongoing RSNA would be expected to enhance tubular sodium reabsorption37 and, thus, increase extracellular fluid volume and the long-term set point of arterial pressure.27
RNS increased PRA overflow less in all of the hypertensive rabbit groups than in sham-operated controls. Interactions between hemodynamic factors and neural control of renin release have been studied,38,39 but effects of Ang II-negative feedback on neurally mediated renin release have received little attention. Our present observations are, therefore, novel and indicate that Ang II negative feedback largely overrides neural stimulation of renin release.
Arterial PRA was greatly reduced in Ang IIinduced hypertension, consistent with the negative feedback influence of Ang II on renal renin release40 and our finding of reduced basal PRA overflow from the denervated kidney in Ang II-treated rabbits under anesthesia. Our finding that arterial PRA was not increased in 2K1C hypertension may seem surprising, because this model is considered "renin dependent." However, hypertension without elevated PRA is frequently observed in established 2K1C hypertension in rabbits and other species.41,42 Nevertheless, increased activity of the intrarenal reninangiotensin system is often observed in both the clipped and nonclipped kidney in 2K1C hypertension,30 which could act to inhibit renal renin release.
A major contribution of the current study has been to distinguish between presynaptic and postsynaptic mechanisms in renal neuroeffector function by using renal norepinephrine spillover. We could not detect significant differences in the effects of RNS on renal norepinephrine spillover between the normotensive and hypertensive groups of rabbits. However, RNS tended to increase renal norepinephrine spillover less in the clipped kidney in 2K1C hypertension than in the kidney of normotensive rabbits, consistent with the notion that renovascular hypertension is associated with a defect in norepinephrine storage in the renal nerves.16 Nevertheless, our data indicate that norepinephrine release from renal sympathetic nerves, for any given frequency of RNS, is not altered in Ang IIinduced hypertension or in the nonclipped kidney in 2K1C hypertension. Our data also suggest that the major changes in sympathetic neuroeffector function that we observed are mediated chiefly at the postjunctional level. Consistent with this, we found that the relationships of renal norepinephrine spillover to renal PRA overflow and MLDF were significantly different in hypertensive compared with normotensive rabbits. Thus, the increased sensitivity of medullary perfusion and reduced sensitivity of renal renin release to RNS, observed in hypertensive rabbits, appears to be because of changes in the sensitivity of these neuroeffectors to sympathetic neurotransmitters, such as norepinephrine.
In the current study, we examined the renal neuroeffector mechanisms under anesthesia, which may be considered a disadvantage. However, as has been observed previously in Ang IIinduced hypertension,15 differences in MAP between the 3 groups of hypertensive rabbits and sham-operated controls were largely abolished by pentobarbital anesthesia. This may reflect the sympatholytic effects of pentobarbital.43 Regardless, it aids interpretation of our experiment by removing potentially confounding effects of systematic differences in MAP between the rabbits. We confirmed in the present study that the 2K1C procedure and the Ang II infusion did indeed produce moderate hypertension, which was similar in the 3 groups.
Perspectives
Our present results provide a new perspective of how changes in renal sympathetic neuroeffector function could contribute to the pathogenesis of Ang IIinduced and 2K1C hypertension. We can reject the hypothesis that the sensitivity of renin release to renal nerve activation is enhanced in these forms of hypertension, because RNS-induced increases in PRA overflow were blunted in the kidneys of all of the hypertensive rabbits that we studied. Our results also do not support the hypothesis that the sensitivity of the renal cortical circulation to renal nerve activation is increased in these models of hypertension. However, our results are consistent with the hypothesis that the sensitivity of medullary perfusion to renal nerve activation is increased in both 2K1C hypertension and Ang IIinduced hypertension. The precise mechanisms underlying this phenomenon remain unknown, but available evidence suggests that reduced NO bioavailability secondary to Ang IIinduced oxidative stress and direct effects of endogenous Ang II on responses of the medullary circulation to neural activation are prime candidates that should be further investigated. Future studies should also investigate the contribution of neural control of medullary perfusion to the development of these forms of secondary hypertension.
| Acknowledgments |
|---|
Sources of Funding
This work was supported by grants from the National Health and Medical Research Council of Australia (317821, 143785, 182816, and 384101) and the National Heart Foundation of Australia (G04M1550).
Disclosures
None.
Received December 5, 2006; first decision December 21, 2006; accepted January 25, 2007.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. L. Burke, R. G. Evans, J.-L. Moretti, and G. A. Head Levels of Renal and Extrarenal Sympathetic Drive in Angiotensin II-Induced Hypertension Hypertension, April 1, 2008; 51(4): 878 - 883. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. G. Evans, S. L. Burke, G. W. Lambert, and G. A. Head Renal responses to acute reflex activation of renal sympathetic nerve activity and renal denervation in secondary hypertension Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2007; 293(3): R1247 - R1256. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Hypertension Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2007 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |