Effects of an Angiotensin-Converting Enzyme Inhibitor, a Calcium Antagonist, and an Endothelin Receptor Antagonist on Renal Afferent Arteriolar Structure

Animals

Four-week-old male SHR and WKY were obtained from the Møllegaard Breeding Center (Ll Skensved, Denmark) and kept in pairs in polycarbonate cages (0.15×0.25×0.40 m) with sawdust bedding at a constant room temperature of 20°C to 21°C and light on from 7 am to 7 pm.

Medication

Rats were divided into 10 groups (see Table 1⇓). Groups 1 through 6 (n=16 in each group) and groups 7 through 10 (n=12 in each group) were as follows: Group 1, untreated WKY; group 2, untreated SHR; groups 3 and 4, SHR treated with 5 to 10 and 1 mg/kg per day cilazapril, respectively; group 5, SHR treated with 20 to 30 mg/kg per day mibefradil; group 6, SHR treated with 100 mg/kg per day bosentan; group 7, untreated SHR; groups 8 and 9, SHR treated with 5 to 10 and 1 mg/kg per day cilazapril, respectively; and group 10, SHR treated with 20 to 30 mg/kg per day mibefradil. An additional eight 4-week-old SHR were obtained and distributed to group 3 (n=4) and group 5 (n=4), because three rats from group 5 died. The drugs were supplied by F Hoffmann–La Roche. The rats were treated from age 4 to 20 weeks. The initial dose of cilazapril in groups 3 and 8 was 10 mg/kg per day, but this reduced SBP below 100 mm Hg; therefore, the dose was reduced to 5 mg/kg per day after 4 weeks of treatment. We increased the initial dose of mibefradil in groups 5 and 10 from 20 to 30 mg/kg per day for the last 4 weeks of treatment to keep SBP at normotensive levels. The dose of the remaining treatment groups was kept constant throughout the treatment period. Drugs were administered in normal sodium chow to which the rats had free access. Fresh chow mixtures were prepared once a week, and the rats were weighed once a week during the treatment period. Drug concentrations were adjusted so that the doses (calculated as milligrams per kilogram per day) were kept constant regardless of food intake and body weight. At age 20 weeks, the treatment was stopped. Kidneys in groups 1 though 6 were prepared for histomorphometric examination as described below. In groups 7 through 10, the rats were killed 12 weeks later with carbon dioxide, after which the heart and left kidney were removed, cleaned of fat, gently blotted, and weighed.

BP Measurements

SBP was monitored noninvasively by the indirect tail-cuff method, and heart rate was recorded simultaneously every 1 to 4 weeks according to a predetermined schedule (see Figs 1 and 2⇓⇓). Each measurement was the mean of a minimum of five determinations made during a 20-minute period while the rats were quiescent. In some rats (four rats randomly taken from each of groups 1 through 6), intra-arterial pressure was measured under conscious, resting conditions at 16 to 18 weeks of age as described elsewhere.²⁵

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Effect of treatment with cilazapril, mibefradil, and bosentan on SBP of SHR for groups 1 through 6 (n=14-19). Control groups were untreated WKY (▴, group 1) and untreated SHR (•, group 2). □ indicates 5 to 10 mg/kg per day cilazapril (group 3); ⋄, 1 mg/kg per day cilazapril (group 4); ▵, 20 to 30 mg/kg per day mibefradil (group 5); and ○, 100 mg/kg per day bosentan (group 6). Cilazapril had a dose-dependent effect on SBP. Cilazapril at the lower dose and mibefradil kept SBP to WKY levels throughout treatment. Bosentan had no effect on SBP in treated compared with untreated SHR. Bars show SEM when it exceeds the size of the symbol.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Effect of cilazapril and mibefradil on SBP of SHR for groups 7 through 10 during 16 weeks of treatment and 12 weeks after treatment withdrawal (n=10-12). • indicates untreated SHR (group 7); □, 5 to 10 mg/kg per day cilazapril (group 8); ⋄, 1 mg/kg per day cilazapril (group 9); and ▵, 20 to 30 mg/kg per day mibefradil (group 10). Cilazapril had a dose-dependent effect on BP during treatment, as for groups 3 and 4. Low-dose cilazapril and mibefradil kept SBP to WKY levels throughout the treatment period. After withdrawal of cilazapril treatment, the effect on BP persisted. After withdrawal of mibefradil treatment, the persistent effect was modest and not significant after 12 weeks. Bars show SEM when it exceeds the size of the symbol.

Microfil Infusion and Morphometry

The concept of the method for determination of afferent arteriolar diameter is to allow fixation of the afferent arterioles in situ while they are relaxed and subjected to a transmural pressure of 100 mm Hg. Afferent arteriolar dimensions are then measured histologically (lumen diameter and media thickness) as described previously.¹¹

In brief, the rats from groups 1 through 6 were anesthetized with methohexital (1% Brietal, Eli Lilly & Co; 7.5 mL/kg body wt IP). After the abdominal wall was opened and the major arteries were cleaned free of fat, the right renal artery was catheterized with a polyethylene tube. The kidney was transfused with plasma solution (human plasma obtained from a local hospital) containing bovine albumin (2.5 mg/mL plasma, Sigma Chemical Co) and heparin (0.004 mL/mL plasma, 5000 IU, DAK) for 10 minutes under constant pressure (100 mm Hg). Ligatures around the aorta, proximal and distal to the origin of the right renal artery and the right suprarenal artery, were tied in rapid succession. Ligatures were tied around the inferior vena cava proximal and distal to the right renal vein, and the vena cava was cut below the right renal vein to allow escape of blood and perfusate. After plasma perfusion was started, the rat was killed while still anesthetized by opening of the thorax and removal of the heart. The atria were removed from the heart, and the ventricles were cleaned and weighed.

With perfusion pressure maintained at 100 mm Hg, the vasculature was relaxed by changing the perfusate to a papaverine solution (2 mg papaverine per milliliter saline, Mecobenzon) for 5 minutes and then to a silicone rubber solution (Microfil MV-130, Flowtec Ltd) containing ultrasonically dispersed microspheres (latex beads, Sigma; diameter, 11.9±1.9 μm [SD]; Sigma specification, approximately 200 000 microspheres per milliliter Microfil) for 5 minutes. At this time, flow was decreased mainly because of the lodging of microspheres in the glomerular capillaries. Then the right renal vein was ligated to stop residual shunt flow through the outer medullary and subcortical zones. This procedure resulted in all afferent arterioles being filled with Microfil, relaxed, and inflated under the same pressure (100 mm Hg) but without hyperfiltration. The Microfil was allowed to harden for 2 hours. The kidney was then removed and immersion-fixed in 3% (wt/vol) formaldehyde plus 1% (wt/vol) glutaraldehyde in 3/4 Tyrode's buffer for a minimum of 3 days.

The perfused kidneys were split longitudinally into halves. The dorsal half of each kidney was cut into six pieces at right angles to the corticomedullary junction for optimization of the number of afferent arterioles cut in cross section. The tissue pieces were preembedded in agar for maintenance of orientation, dehydrated through a graded series of ethanol solutions, and embedded in glycol methacrylate (Historesin, LKB). Resultant blocks were cut in serial 2-μm-thick sections on a microtome (Supercut 2065, Leica, Reichert-Jung), placed on glass slides, stained with Giemsa, and coded so that the examiner was unaware of the tissue source. With the use of two microscopes (BH2, Olympus; oil immersion lenses, ×100) equipped with mirrors, pictures of adjacent sections were projected onto a tabletop. Inner diameter (ID) and diameter of the border between the media and adventitia of the vessel (OD) were measured with a ruler at a total magnification of ×1650 and resolution of 0.6 μm. Media thickness and media cross-sectional area were calculated as (OD−ID)/2 and π[(OD/2)²−(ID/2)²], respectively. We included only vessels that were situated in the renal cortex, contained Microfil, had an inner lumen diameter less than 35 μm, and fulfilled at least one of the following criteria for being afferent arterioles: (1) presence of a microsphere in the lumen or in a glomerulus, providing that it was possible in serial sections to show that the microsphere had passed through the vessel to that glomerulus; (2) presence of an internal elastic lamina, because this lamina is absent in efferent arterioles even where many smooth muscle cells are seen in the vascular wall³⁴ ; (3) a close relation to an interlobular artery, if necessary demonstrated in serial sections; (4) presence of epithelioid cells; and (5) presence of Microfil in the arteriole but absence in the glomerular capillaries, indicating that the efferent part in that region had not been filled (for further details, see Reference 11). The arterioles were categorized as either distal, which were observed to be in close relation to the glomerulus, or proximal, which were all other afferent arterioles measured. On average, 32 (range, 23 to 46) profiles of afferent arterioles were measured in each kidney. Branching and irregular vessels were excluded.

At the end of the study, 140 of 152 rats remained. Reasons for exclusion were as follows: 1 rat died at the age of 5 weeks for unknown reasons in group 5; 2 rats died during intra-arterial measurements in group 5; 1 rat was killed at age 12 weeks because of rotatio misiofacialis in group 9, and the partner to that rat subsequently had a large variation in BP measurements and was therefore also taken out of the study; and for 7 rats, kidney perfusion was poor, so the rats were taken out of the study without further analysis (2 rats from each of groups 1 and 2, and 1 rat from each of groups 3, 4, and 5).

Statistics

All results are presented as mean±SE. Differences in rat characteristics, BP values, and effect of treatment on renal afferent arterioles were tested among all groups by one-way ANOVA (Bonferroni test, INSTAT). The null hypothesis that slope and intercepts equaled zero for tail-cuff SBP and intra-arterial mean BP was evaluated with Student's t test. Probability levels less than 5% were considered significant.

Effect of Treatment on BP

SBP in the low-dose cilazapril– and mibefradil-treated groups was close to SBP for WKY throughout the treatment period (Fig 1⇑, Table 2⇓, groups 1 through 6). Bosentan did not lower BP remarkably but followed the curve for untreated SHR. SBP was kept below that of WKY in the high-dose cilazapril group. Fig 2⇑ and Table 2⇓ (groups 7 through 10) show the persistent effect of cilazapril and mibefradil treatment on SBP after treatment withdrawal. Cilazapril had a persistent and maintained effect. For the low dose of cilazapril, BP was only slightly increased after withdrawal, such that from 3 to 12 weeks after treatment withdrawal, the high- and low-dose cilazapril groups did not differ. After withdrawal of mibefradil, SBP rose slowly but was not significantly different from SBP in untreated SHR at age 32 weeks (P>.05).

We measured intra-arterial SBP, diastolic BP, and mean BP (only mean BP is shown) directly at ages 16 to 18 weeks in four rats from each of groups 1 through 6. Fig 3⇓ shows the relation between intra-arterial mean BP and tail-cuff SBP for these rats. Intra-arterial mean BP measurements were similar to noninvasive tail-cuff SBP measurements (r=.82, P<.0001), but intra-arterial measurements in the mibefradil group tended to be higher than tail-cuff measurements.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Relation between intra-arterial mean blood pressure (MBP) and tail-cuff SBP measured in four rats of each of groups 1 through 6. Mean BP was taken at age 18 weeks, except for group 4, in which it was taken at 16 weeks. SBP was taken as the mean of SBP at 16 and 20 weeks. Line shows linear regression (r=.82, P<.0001). Symbols are as defined in Fig 1⇑ legend.

Afferent Arteriolar Structure

Table 3⇓ presents the results of morphometric measurements of the afferent arterioles. Lumen diameter was significantly smaller in untreated SHR than in WKY in both proximal and distal afferent arterioles. Bosentan treatment did not alter afferent arteriolar dimensions compared with untreated SHR. There was a tendency for a reduced lumen in distal afferent arterioles in mibefradil-treated rats (P=.014 by direct t test against untreated SHR), but this was not significantly different from values in untreated SHR using our standard statistical analysis (one-way ANOVA and Bonferroni test). Cilazapril treatment caused a dose-dependent increase in distal and proximal afferent arteriolar lumen diameters compared with untreated SHR, and this was associated with a decreased media-lumen ratio in each case. Media cross-sectional area was increased in distal afferent arterioles at the high cilazapril dose compared with the remaining five groups, in which no difference in media cross-sectional area was found. A similar tendency was seen in proximal afferent arterioles.

Fig 4⇓ shows distal afferent arteriolar diameter as a function of SBP for both individual rats (top) and the rat groups (bottom). Fig 4⇓ confirms that in general, a reduction in SBP was associated with an increase in distal afferent arteriolar lumen diameter but that for equal hypotensive effect, cilazapril treatment caused a greater increase in lumen diameter than mibefradil treatment.

• Download figure
• Open in new tab
• Download powerpoint

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Distal afferent arteriolar lumen diameter and SBP for rats in groups 1 through 6. Top, Individual rats; bottom, group averages, where bars show SEM when it exceeds the size of the symbol. Symbols are as defined in Fig 1⇑ legend.

BP Control During Treatment

The dose-dependent reduction of BP with cilazapril (where BP reduction was maintained throughout the study; 1 mg/kg per day lowered SBP to the WKY level, whereas 5 to 10 mg/kg per day lowered SBP below WKY values) confirms the antihypertensive effect of cilazapril shown in previous studies on SHR.^{23 35} Also, mibefradil had an antihypertensive effect, in accord with other reports^{36 37} ; in the present study, 20 to 30 mg/kg per day mibefradil lowered SBP to WKY levels throughout the treatment period. The lack of effect of orally administered bosentan on BP is in agreement with other findings in SHR.³² In contrast, when bosentan is given to deoxycorticosterone acetate–salt hypertensive rats³⁸ or sodium-depleted squirrel monkeys,³⁹ BP values are reduced 10% (SBP) and 30% (mean BP), respectively.

Measurement of Afferent Arteriolar Structure

The conditions under which arteries are fixed and investigated are crucial if morphological comparisons between vessels are to be valid.^{40 41} Since arteries are contractile elastic structures whose dimensions will consequently depend on the distending transmural pressure, the classic way of investigating vessel dimensions in situ is with perfusion fixation at a constant flow^{15 42} or pressure.^{8 10} However, because the transmural pressure to which a given vessel is subjected during fixation will depend on the vascular resistance upstream and because that resistance is obviously unknown, vessel structure measurements made under the conditions given by the simple perfusion fixation technique will not be unequivocal. Furthermore, any contractile effect of the fixative on the smooth muscle cells used during fixation must also be considered in making a distinction between structural and functional influence on vascular structure. The technique we have used¹¹ differs in a number of important respects from those used by other researchers. First, the use of microspheres to lodge in the glomerular capillaries means that the afferent arterioles are fixed under a known intravascular pressure. Second, the use of Microfil in the perfusate ensures that no filtration occurs, so that interstitial pressure will not rise and thus transmural pressure during fixation is close to intravascular pressure. Third, the use of papaverine ensures that the vessels are relaxed. Fourth, the technique allows measurements of both lumen diameter and the corresponding media thickness, thus differing from the vascular casting method,^{8 10 15} which allows measurement of only diameter. The method therefore provides information in one process about the structural change of the vascular wall under well-defined conditions.

Renal Afferent Arteriolar Structure

We and others^{8 9 10 11 15} have previously shown that the afferent arteriolar diameter in SHR is smaller than that in WKY. Furthermore, SHR have fewer but similar-sized glomeruli compared with WKY.⁴³ The narrowing of the afferent arterioles is in full agreement with the hemodynamic findings that renal vascular resistance at maximal vasodilatation is increased by 15% to 36%.^{44 45} It is likely that this narrowing plays a crucial role in the pathogenesis of the hypertension in SHR, because it has recently been shown that a narrowed afferent arteriole in 7-week-old F₂ SHR×WKY contributes to the later development of high BP.¹³

Generally, the lumen diameter of resistance vessels is found to be reduced in hypertension in both humans⁴⁶ and animals.⁴¹ The reduction was originally thought to be the result of a growth process in which the media encroached into the lumen.⁴⁷ More recently, however, it has been recognized that the narrowed lumen need not be associated with growth but can be due to a rearrangement of the wall material around the smaller lumen.^{46 48} Indeed, this appears to be the case in renal afferent arterioles, for media cross-sectional area is reported to be either similar^{8 15} or decreased^{11 13} ; in the present study, a tendency toward a reduction was seen in the distal afferent arterioles (P=.07).

Effect of Treatment on Renal Structure

In extrarenal arteries, ACE inhibitors dose-dependently increase lumen diameter and reduce media-lumen ratio.^{22 23 49 50} Similarly, in renal afferent arterioles, Notoya et al¹⁵ showed that lisinopril treatment caused an increase in lumen diameter, consistent with the findings of our study using cilazapril and with the general finding that ACE inhibitor treatment results in increased renal blood flow.^{14 51 52} More surprisingly, we found that with the high dose of cilazapril, the media cross-sectional area of the afferent arterioles was increased. This was also a finding in the study of Notoya et al, and they suggested that since the vascular wall in afferent arterioles is composed of only one or two layers of vascular smooth muscle cells, widening of the lumen would cause an abnormal vessel without an increase in the amount of vascular wall material. In our experiments too, the 29% increase in lumen diameter was not accompanied by any change in media thickness, thus giving a 22% decrease in media-lumen ratio but an increase in media cross-sectional area. Since it can be expected that the ACE inhibitor treatment reduced glomerular pressure,⁵³ some reduction in media-lumen ratio is also to be expected; nevertheless, geometry dictates that an increase in lumen must be accompanied by an increase in media cross section. Thus, although the widening of the lumen may play a role in the hypotensive action of ACE inhibitors, the increase in media cross section can probably be accounted for by the normal adaptation mechanisms of the vasculature.⁴¹

Our findings and those of Notoya et al¹⁵ that ACE inhibitor treatment causes increases in the lumen diameter of afferent arterioles are in contrast to a report by Kett et al,⁵⁴ who found that enalapril treatment from age 4 to 10 weeks failed to alter either media-lumen ratio or media cross-sectional area in arcuate and interlobular arteries from perfusion-fixed SHR. The reason for the discrepancy is unclear but might be related to the relatively short treatment period in their study. Alternatively, the difference may be that we and Notoya et al have investigated more distal vessels than Kett et al.

The effects of calcium antagonist treatment on renal blood flow vary. Thus, although benidipine treatment results in increased renal blood flow²¹ and diltiazem treatment causes decreased glomerular pressure,⁵⁵ nitrendipine treatment does not affect renal blood flow in SHR⁵⁶ and nifedipine does not do so in humans.⁵¹ Dworkin et al⁵⁷ showed that nifedipine reduces glomerular pressure to the same level as ACE inhibitors in uninephrectomized SHR. However, the beneficial effect of nifedipine on glomerular injury appears not to be mediated through the reduction of glomerular capillary pressure in remnant-kidney⁵⁸ or deoxycorticosterone acetate–salt⁵⁹ hypertension, which is the case for ACE inhibitors. Our findings, in which cilazapril was more effective than mibefradil in increasing afferent arteriolar lumen diameter, would be in accord with cilazapril treatment having a greater effect on renal blood flow than mibefradil treatment, but hemodynamic measurements are required to confirm this.

The relatively modest effect of mibefradil treatment on the diameter of renal afferent arterioles in our study (Fig 4⇑) suggests that the antihypertensive effect of mibefradil is not mediated primarily through an effect on renal afferent arteriolar structure. However, it should be noted that the antihypertensive effect of mibefradil may not have been as marked as shown in Figs 1 and 2⇑⇑, even though these are based on more than 200 noninvasive measurements. As indicated in Fig 3⇑, the intra-arterial measurements in four of the rats treated with mibefradil suggested a more modest antihypertensive effect. We cannot therefore exclude the possibility that with mibefradil treatment, the tail-cuff method underestimates BP, in which case there would be a better correlation between BP and afferent arteriolar diameter in this rat group. Unfortunately, there is no ideal method for measuring BP over long periods in rats.

The lack of effect of bosentan treatment on renal afferent arteriolar structure in our experiment is consistent with the lack of effect of this drug on BP. However, the result does contrast with the reported ability of bosentan to increase renal blood flow acutely.³¹ Long-term measurements of the effect of endothelin receptor antagonists on renal blood flow are clearly required.

Persistent Effect of Cilazapril on BP

The persistent effect of ACE inhibitors on BP after treatment withdrawal in SHR has been reported previously.^{14 22 25 27 60} Numerous causes of this persistent effect have been suggested based on cardiac,^{61 62} sympathetic,⁶⁰ and renal^{14 53} mechanisms. It has also been suggested that the failure of BP to rise after ACE inhibitor treatment is withdrawn may reflect a persistent reduction of structural cardiovascular damage as a result of low BP.^{14 47} However, a study of the small arteries from four different vascular beds has concluded that the effect on the structure of small arteries is not directly related to the persistent effect on BP.²²

The present results support the possibility that renal mechanisms are involved in the persistent effect of ACE inhibitor treatment. It is striking that the persistent effect of cilazapril was associated with a substantial increase in afferent arteriolar diameter during treatment, whereas the modest persistent effect of mibefradil was associated with a modest increase in afferent arteriolar diameter. The findings are thus consistent with the findings of Harrap et al,¹⁴ who showed that glomerular filtration rate and renal blood flow remained increased and renal vascular resistance remained reduced in rats previously treated with perindopril. It would therefore be of interest to determine afferent arteriolar morphology in the period after withdrawal of ACE inhibitors.

In conclusion, this study demonstrates that the ACE inhibitor cilazapril is more effective than the calcium antagonist mibefradil in altering the structure of renal afferent arterioles in SHR. The endothelin receptor antagonist bosentan did not affect BP or afferent arteriolar morphology. Moreover, cilazapril had a persistent effect on BP after treatment withdrawal, whereas the persistent effect of mibefradil was modest. The association of increased afferent arteriolar diameter and lower BP level after treatment withdrawal may suggest a pathogenic role for afferent arteriolar diameter in the development of high BP in the SHR.