The Renal Medulla and Hypertension

Reason for Studying the Medullary Circulation in Hypertension

In many experimental models of hypertension, there is no reason to believe that changes in blood flow to the renal medulla participate in the resetting of the pressure-natriuresis relationship (Fig 1⇑). For example, immediately after renal artery stenosis (Goldblatt hypertension), it is clear that pressure and flow throughout the kidney are compromised. Similarly, in reduced renal mass, renoprival, and perinephric models of hypertension, it is obvious that global reductions in RBF and GFR are responsible for the reduced renal excretory capacity. In the endocrine models of hypertension produced by exogenous administration of sodium- and water-retaining hormones such as Ang II, aldosterone, or deoxycorticosterone acetate, the resetting of the pressure-natriuresis relationship can largely be attributed to enhanced tubular reabsorption of sodium and water. So even if medullary blood flow is reduced in these models of hypertension, such reductions are not necessary to explain the rise of arterial pressure.

It is also apparent that renal functional abnormalities exist in all of the genetic models of hypertension, as demonstrated by the shift of the pressure-natriuresis-diuresis relationships to higher pressures in SHR,^{12 28} Lyon hypertensive rats,¹⁴ Dahl salt-sensitive rats,¹³ and transgenic rats overexpressing the mouse Ren-2 gene.¹⁵ However, the nature of the renal dysfunction in these forms of hypertension, as in human essential hypertension, has not been apparent. RBF and GFR are similar in SHR and normotensive rat strains^{12 29} before and during the development of hypertension. This is also the case early in the development of human essential hypertension.^{30 31 32} In these situations, which represent most hereditary and clinical hypertension, it was reasonable to suspect that more subtle changes in renal tubular or vascular function contributed to the resetting of renal function.

For these reasons, we explored the possibility that alterations of renal medullary blood flow might be responsible for the shift in the pressure-diuresis relationship toward higher pressure in SHR. The first convincing evidence that medullary blood flow was indeed reduced in hypertension was obtained with the use of laser-Doppler flowmetry, as described below. These studies indicated that medullary blood flow was reduced before the development of hypertension in SHR.^{28 33} Further studies revealed that the reduction in medullary flow in SHR was related to enhanced vascular tone in afferent arterioles of juxtamedullary nephrons.³⁴ These studies confirmed earlier work by Ganguli et al³⁵ indicating that papillary blood flow is reduced in SHR. They also were consistent with the original studies by Muirhead,³⁶ who suggested that an impaired release of antihypertensive lipids (medullipins) from interstitial cells of the medulla may contribute to the development of hypertension in SHR.

Although our studies suggested that reductions of medullary blood flow might be responsible for the development of hypertension in the SHR, this type of experiment could not establish cause-and-effect relationships. Therefore, we directed our efforts toward developing a method whereby we could selectively alter medullary blood flow independent of changes of renal cortical blood flow and systemic vascular resistance to determine whether a primary reduction in medullary flow is sufficient to produce hypertension. Related to this goal, we also had to develop methods to chronically measure cortical and medullary blood flow in unanesthetized rats. The remainder of this review describes the studies carried out in our laboratory over the last 3 years to examine the hypothesis that a primary reduction in medullary blood flow is sufficient to produce hypertension, whereas increases in medullary blood flow can lower arterial pressure.

Measurement of Medullary Blood Flow With the Use of Laser-Doppler Flowmetry

Testing concepts regarding the role of the renal medulla in hypertension has proved to be challenging because blood flow to the inner medulla accounts for less than 1% of total RBF. Few people tried to study medullary blood flow throughout the late 1970s and 1980s because it had become apparent that the techniques available for the study of this circulation were fraught with difficulties and probably invalid.^{37 38 39} These techniques included measurement of the clearance of diffusible indicators (H₂, ⁸⁵Kr, ¹³³Xe, heat), measurement of albumin or red blood cell accumulation in the papilla, and radiolabeled microspheres. Blood flow in the vasa recta capillaries at the tip of the papilla of immature rats could be surgically exposed and studied with the use of fluorescent videomicroscopy.^{38 39} This technique has been criticized, however, because removal of the ureter itself alters papillary blood flow by stimulating the release of prostaglandins. So each of these techniques had problems, and most of them also required anesthesia, surgery, exposure of the kidney, and death of the animal. None of these methods could provide continuous measurements of medullary flow or could be adapted for repeated use in conscious animals.^{37 38}

In the early 1980s, Roman and coworkers^{13 40 41} began to explore the use of laser-Doppler flowmetry to measure changes in blood flow in the renal cortex and tip of the exposed papilla of anesthetized rats. These studies yielded important new information indicating that papillary blood flow increased after elevations in renal perfusion pressure. Remarkably, the circulation of the inner medulla failed to autoregulate blood flow. These responses could even be demonstrated in young Munich Wistar rats in which the optical probe could be used to measure papillary blood flow without exposure of the renal papilla.⁴²

A further refinement of the laser-Doppler technique that has since evolved in our laboratory now allows us to chronically implant small optical fibers into various regions of the kidney for simultaneous measurement of cortical and medullary blood flows in conscious rats under a variety of experimental conditions.^{43 44 45} As illustrated in Fig 2⇓, the 0.5-mm-diameter fibers connected to an external probe are implanted to various depths in the renal cortex and medulla of the kidney through a small hole made in the renal capsule with the use of a 25-gauge needle. Implantation of these fibers results in minimal bleeding in the rat and has no effect on RBF, GFR, urine concentrating ability, or sodium and water excretion. Histological damage is confined to within 200 μm of the fiber track, and no disruption has been observed to the microcirculatory region beyond the tip of the implanted fiber where flow is determined.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Photomicrograph shows the distribution of [¹⁴C]clentiazem after its infusion into the renal medullary interstitium of a rat kidney. The illustrations show the position of the interstitial catheter used for compound delivery into the renal medulla and the position of the two optical fibers used to measure outer cortical blood flow and outer medullary blood flow.

All of the methods previously used to measure regional blood flow in the kidney have severe limitations, so direct validation of the laser-Doppler technique for measurement of tissue blood flow in the kidney has been a problem. However, a number of comparisons have been made which indicate that implanted optical fibers can provide reliable measurements of changes of regional blood flow in kidney, albeit not in absolute flow units. Specifically, changes of cortical blood flow measured with implanted fibers in the superficial cortex closely followed changes observed in whole-kidney blood flow measured with an electromagnetic flowmeter.⁴³ Changes of inner medullary blood flow measured with an implanted fiber closely correlate to changes in papillary blood flow measured with an external probe focused on the exposed papilla of the rat⁴³ or to changes in blood flow in vasa recta capillaries measured with videomicroscopy.⁴⁶ The most direct validation of laser-Doppler flowmetry for the measurement of papillary blood flow in the rat compared the laser-Doppler flow signal with the rate of accumulation of ⁵¹Cr-labeled erythrocytes in the papilla.⁴¹ In this study, the laser-Doppler signal was linearly related and highly correlated (r=.92) to red blood cell flow into the papilla.

The use of acutely implanted fibers has recently enabled us to confirm the original observations by Roman et al⁴⁰ showing that medullary blood flow is not autoregulated as well as cortical flow in volume-expanded rats. The use of multiple optical fibers enabled simultaneous determination of changes in blood flow in the inner and outer medullas as well as the superficial and inner cortices³³ in response to changes in perfusion pressure. These studies showed that as renal perfusion pressure was increased above 100 mm Hg, total RBF, superficial cortical blood flow, and deep cortical blood flow were all very well autoregulated. In contrast, blood flow to the inner and outer medullas were autoregulated poorly in volume-expanded rats. In hydropenic animals in which the plasma levels of AVP are high and the renin-angiotensin system is activated, medullary blood flow does not rise when renal arterial pressure is increased above normal levels. That is, blood flow in the medulla is autoregulated and the pressure-natriuresis relationship markedly attenuated. The mechanism responsible for altering medullary flow and the sensitivity of pressure natriuresis in hydropenic versus volume-expanded rats is under intense investigation in our laboratory but remains to be defined.

Development of Techniques for the Infusion of Vasoactive Compounds Into the Renal Medulla

For the evaluation of the consequences of selective alterations of medullary blood flow on the long-term control of arterial pressure, special techniques had to be developed that enabled the delivery of vasoactive agents directly into the renal medulla of the rat.⁴⁷ These techniques were first used in acute anesthetized rats studies and later adapted for the continuous long-term infusion of compounds into the renal medulla for weeks.^{45 48 49} Renal interstitial catheters made of polyethylene and with a tip diameter extruded to 100 μm were inserted into the white inner medulla and compounds continuously infused into the medullary interstitium at a rate of 5 to 10 μL/min. Substances infused into this region of the kidney are selectively accumulated because of the efficient countercurrent exchanger in the vasa recta circulation. As shown in Fig 2⇑, infusion of the radiolabeled (¹⁴C) calcium antagonist clentiazem for 20 minutes resulted in the distribution of more than 92% of the total radioactivity in the infused kidney to the medulla (outer zone plus inner zone plus papilla). The total radioactivity retained in the infused kidney was 47 times that found in the contralateral kidney.⁴⁷

This technique has enabled us to deliver vasoactive agents in the medulla to selectively alter medullary vascular tone both acutely (hours) and chronically (weeks), as determined by laser-Doppler flowmetry. As summarized below, a number of mechanistically different vasoconstrictor and vasodilator agents now have been infused into the renal medulla to determine the effects of preferential changes of papillary blood flow on sodium and water excretion as well as the long-term control of arterial pressure.

Influence of NO on Medullary Blood Flow, Pressure Natriuresis, and the Long-term Control of Arterial Pressure

A number of studies have now indicated that NO plays a major role in the regulation of arterial pressure. Short-term inhibition of NO synthase is associated with a fall in RBF^{50 51 52 53} and GFR^{50 51 54} and increases both preglomerular and postglomerular vascular resistance.^{55 56} Long-term administration of NO synthase inhibitors produces sustained hypertension^{57 58 59 60 61 62} that appears to be related to a shift of the renal pressure-natriuresis relationship to higher levels of arterial pressure.^{52 55}

We carried out studies to determine whether we could selectively alter renal medullary blood flow and to determine the effects of such modulation on sodium and water excretion. In these studies, infusion of N^G-nitro-l-arginine (L-NAME, 120 μg/h) into the medullary interstitium produced a 24% fall in papillary blood flow without altering cortical flow.⁶² Renal interstitial pressure was reduced by 23%, and sodium and water excretion was reduced by approximately 35%. GFR, fractional sodium and water excretion, and total RBF were unchanged, and no changes of systemic arterial blood pressure occurred in these short-term, 1-hour studies. These results demonstrated for the first time that NO is tonically active in the medullary circulation and that changes in medullary blood flow alter sodium and water excretion (see Fig 3⇓). These observations were consistent with previous reports that the renal papilla tissue slices have a greater capacity to synthesize NO than the renal cortex in vitro⁵⁶ and that the vascular and tubular segments of the renal medulla have a large capacity for producing NO.⁶⁴

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Line graph (left) shows changes in renal cortical and papillary blood flows in response to infusion of N^G-nitro-l-arginine methyl ester (L-NAME) at 120 μg/h into the renal medullary interstitium of anesthetized normotensive rats. Bar graph (right) shows changes in urine volume (UVol) and sodium excretion (UNaV) accompanying medullary interstitial infusions of L-NAME in the infused left kidney and noninfused contralateral kidney. *P<.05. (Data from Mattson et al⁶² ; reprinted with permission from Cowley et al.⁶³ )

Other studies in our laboratory have also demonstrated an important role for NO in mediating the effects of kinins on medullary blood flow. Previous studies have suggested that bradykinin shifts the distribution of renal blood^{65 66 67} to the renal medulla and promotes the excretion of sodium and water.⁴² The influence of endogenously produced kinins on medullary blood flow was determined by assessing the effects of converting enzyme inhibitors and other inhibitors of kinin degradation and by determining whether these effects could be reversed with a kinin receptor antagonist.^{42 68} Kininase II inhibition with enalaprilat increased papillary blood flow by nearly 50%, and infusion of the kinin antagonist returned papillary blood flow to control levels.⁴² Outer cortical blood flow and arterial pressure were not altered, and GFR was unchanged. Associated with the increase of papillary flow, urine flow and sodium excretion increased.^{42 69}

We also carried out studies to examine the effects of infusion of bradykinin directly into the medullary interstitium.⁶⁹ Papillary blood flow increased by 20% during bradykinin infusion, and total RBF, GFR, and renal interstitial hydrostatic pressure were unaffected. Urine flow and sodium excretion increased by 100% to 120% in the absence of changes in renal tubular or vascular function in the contralateral kidney. Pretreatment of the medullary interstitium with the NO inhibitor L-NAME blocked the effects of bradykinin on the medullary circulation. L-NAME also blocked the effects of captopril on papillary blood flow, whereas this effect was not blocked by pretreatment of the rats with a cyclooxygenase inhibitor.¹⁶ These studies show that the intrarenally generated kinins may play an important role in the regulation of medullary blood flow and that the effect of kinins on the medullary circulation is mediated by an NO-dependent mechanism.

In these same studies, we also determined the role of NO on the vasodilator effects of acetylcholine, demonstrating that medullary infusion of acetylcholine at a dose of 200 μg/h produced increases in medullary blood flow similar to those produced with bradykinin (34%), with only small increases (13%) of cortical blood flow.⁶²

Role of NO in Long-term Regulation of Medullary Blood Flow

Having obtained evidence in anesthetized rats for an important role of NO control of medullary blood flow, we turned our attention to studies designed to determine whether we could chronically reduce medullary blood flow and produce a sustained form of hypertension. Rats were instrumented with a catheter for long-term L-NAME infusion into the renal medulla, and optical fibers were implanted for measurement of changes in cortical and medullary blood flows. As summarized in Fig 4⇓, long-term L-NAME infusion into the renal medulla selectively reduced papillary blood flow by nearly 30%.⁴⁵ This reduction was apparent as early as 2 hours after L-NAME infusion was started and clearly preceded the rise of mean arterial pressure (MAP). The reduction of medullary flow was maintained throughout the experiment. Cortical blood flow was unaltered. After L-NAME infusion was stopped, blood flow in the renal medulla and arterial pressure gradually returned to control levels. Sodium excretion decreased significantly on the first day of L-NAME infusion and remained reduced throughout 5 days of drug infusion. MAP rose in parallel with the retention of sodium excretion. Negative sodium balance, which paralleled the return of arterial pressure to control levels, was seen after L-NAME infusion was stopped.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Line graphs show time course of changes in mean arterial pressure (MAP), renal medullary blood flow (medullary flow signal), and renal cortical blood flow (cortical flow signal) in conscious, uninephrectomized Sprague-Dawley rats during renal medullary interstitial infusion (r.i.) of N^G-nitro-l-arginine methyl ester (L-NAME) for 5 days. Blood flow in the renal cortex and medulla is represented by the raw voltage signal recorded from optical fibers implanted in the kidney with the use of laser-Doppler flowmetry. *P<.05 from third control day. (From Mattson et al.⁴⁵ )

These data demonstrate that NO plays an important role in the regulation of blood flow in the renal medulla. Moreover, these studies indicate that selective reductions of medullary blood flow can produce a sustained elevation of arterial blood pressure. The rise of arterial pressure could not be explained on the basis of a systemic action of recirculated L-NAME because minimal systemic inhibition of NO synthase was observed in these studies. Specific reductions in medullary blood flow would be expected to increase the reabsorption of sodium in the deep nephrons, especially in the loop of Henle,²⁷ which could explain the retention of sodium and hypertension. However, removal of the inhibitory effects of endogenous NO on sodium transport in the medullary collecting duct cannot be excluded as part of the mechanism involved in the resetting of the pressure-natriuresis relationship in this form of hypertension. Nevertheless, as discussed below, we have observed a consistent relationship between changes in medullary blood flow and the level of blood pressure with the use of a variety of vasoactive compounds, some of which have no known action or even opposite effects on tubular function. Thus, the common element in our studies related to blood pressure appears to be changes of medullary blood flow.

Long-term Intravenous L-NAME Infusion

We have performed studies to determine the extent to which changes in renal medullary flow participate in the development of hypertension produced by long-term intravenous L-NAME administration.⁷⁰ Intravenous L-NAME infusion produced sodium and water retention and a sustained elevation of blood pressure, as expected. This was accompanied by a preferential 22% fall in medullary blood flow, which remained at this level throughout the experiment. After L-NAME treatment was stopped, medullary blood flow and arterial pressure rapidly returned to control levels. The L-NAME dose chosen was low enough so that it had no effect on cortical blood flow throughout the study. The reduction of medullary flow and development of hypertension were associated with retention of sodium and water. Previous studies have assumed that systemic vasoconstriction and central neural actions were largely responsible for the development of hypertension produced by systemic blockade of NO synthase. However, our results indicate that changes in renal medullary blood flow probably are equally important in resetting the pressure-natriuresis relationship, which is essential. Taken together, the results of both the intrarenal and intravenous L-NAME infusion studies indicate that in the absence of measurable changes in RBF at the whole-kidney level, reductions of medullary flow alone are sufficient to reset the pressure-natriuresis relationship and promote sodium retention and the development of hypertension.

Effects of Long-term Medullary Interstitial Infusion of a V₁ Agonist

The renal medullary circulation has long been thought to play an important role in maintaining the osmotic gradient, but the possible role of changes in medullary blood flow in determining maximal urinary concentrating ability has not been directly studied. Acutely implanted optical fibers enabled us recently to study the influence of AVP on medullary blood flow.⁷¹ These studies, performed in renal denervated rats, also determined the relative contribution of V₁ and V₂ receptors to the response to AVP. Infusion of a V₁ receptor agonist into the renal medulla selectively reduced blood flow in the outer medulla by 15% and to the inner medulla by 35%. Equimolar doses of AVP also decreased outer medullary blood flow by 15%, but the fall in inner medullary flow (17%) was significantly less than that observed with the V₁ agonist. Stimulation of V₂ receptors by medullary interstitial infusion of the V₂ agonist 1-desamino-8-d-AVP or infusion of AVP in rats pretreated with a V₁ receptor antagonist increased medullary blood flow by 16% and 27%, respectively. These studies demonstrate that AVP has two diametrically opposed actions on medullary blood flow resulting from stimulation of the V₁ and V₂ receptors. Stimulation of V₂ receptors attenuates the vasoconstrictor actions of AVP and the V₁ receptor on the medullary circulation and likely accounts for the inability of the endogenous hormone to induce hypertension or sodium retention.

Given that infusion of the V₁ agonist lowers medullary blood flow, we then studied the long-term effects of such infusion on arterial pressure. We found that the continuous intravenous infusion of small doses of a specific V₁ receptor agonist produced sustained hypertension.⁷² Since it was evident from our short-term studies in anesthetized rats⁷¹ that this compound resulted in a significant reduction of medullary blood flow, we undertook several studies to examine the role of V₁-mediated medullary vasoconstriction in long-term arterial pressure control. In one of these studies, we were able to prevent the development of hypertension produced by systemic infusion of the V₁ agonist by simultaneous administration of a V₁ receptor antagonist (equimolar amounts) into the renal medulla.⁴⁹ Moreover, as soon as the infusion of the V₁ antagonist into the renal medulla was discontinued, the continued intravenous administration of the V₁ agonist produced systemic hypertension. These studies indicate that the renal medullary vasoconstrictor actions of the systemically administered AVP agonist were required to produce the hypertension.

As reported in the above study, infusion of the V₁ agonist in the renal medulla reduced inner medullary blood flow by 35% (see above and Reference 70⁷⁰ ). We performed studies to determine whether infusion of the V₁ agonist into the renal medulla alone was sufficient to produce sustained hypertension.⁴⁹ As shown in Fig 5⇓, we observed a sustained rise in MAP of nearly 20 mm Hg over the 14 days of the experiment. After infusion was terminated, arterial pressure rapidly returned to the normal control levels of 100 mm Hg. The same degree of hypertension was seen by renal medullary infusion in renal denervated rats. Since renal medullary infusion of AVP does not produce hypertension, it appears that simultaneous stimulation of the V₂ receptor in some way modulates the hypertensive effects of the endogenous hormone.^{49 72} Preliminary studies (A.W.C., unpublished results, 1994) in five rats have indicated that medullary interstitial infusion of these amounts of the V₁ agonist results in a sustained reduction of medullary blood flow in the complete absence of changes in cortical flow over a period of at least 5 days.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Line graph shows time course of changes in mean arterial pressure (MAP) in conscious Sprague-Dawley rats given a V₁ agonist into the renal medulla (r.i.) for 14 days. *P<.05 from the final control day. (From Szczepanska-Sadowska et al.⁴⁹ )

Once again, the renal medullary vasoconstrictor actions of a compound appear to account fully for the development of this model of hypertension. Interestingly, we found no evidence of sodium retention in these studies, which conforms to short-term observations in anesthetized rats that medullary interstitial infusion of the V₁ agonist does not result in increased reabsorption of filtered sodium.⁷³ Nor did the long-term infusion of the V₁ agonist systematically lead to volume retention or a rise of plasma renin or aldosterone.^{49 72} Measurements of systemic plasma levels of AVP during long-term infusions into the medullary interstitial space (7 days) showed no changes in circulating levels in the blood. Since AVP and the V₁ agonist are metabolized identically and can be measured by radioimmunoassay, this indicated that recirculation of the infused compound does not account for the rise of arterial pressure.

It is recognized that hypertension can occur in the absence of sodium and water retention because sodium and water balance can be achieved after renal excretory ability is impaired by a rapid rise of arterial pressure that offsets the lower excretion rate through the pressure-natriuresis-diuresis mechanism. Nevertheless, the mechanism whereby medullary vasoconstriction led to the rise of arterial pressure in these particular V₁ agonist studies raises important questions as to whether the rapid rise in arterial pressure seen in these experiments was due to suppression of the release of a vasodilator substance such as medullipin³⁶ or the release of an unidentified vasoconstrictor substance.

Antihypertensive Effects of a Selective Increase of Medullary Blood Flow in SHR

We and others^{3 12 28 74} have shown that kidneys of SHR require a higher level of arterial pressure than kidneys of normotensive rats to excrete a given amount of sodium and water and that this is associated with a reduced papillary blood flow.^{12 19 33} As might be predicted from the reduction of medullary blood flow, parallel reductions of renal interstitial hydrostatic pressure for a given level of perfusion pressure were observed.^{19 74} Importantly, the relationship between papillary blood flow and renal perfusion pressure was shifted to a higher pressure level early in the development of hypertension. For example, a reduction in the slope of the pressure-natriuresis relationship was clearly evident in 3- to 4-week-old SHR compared with age-matched Wistar-Kyoto rats (WKY).²⁸ At this age, it is difficult to demonstrate a significant difference in MAP with indwelling catheters. Thus, the fall in papillary blood flow and the resetting of pressure natriuresis are probably not consequences of preexisting hypertension. Since the reduction in renal papillary blood flow is one of the earliest abnormalities in renal function that occurs during the development of hypertension in SHR, it appeared that this dysfunction might contribute to the resetting of the pressure-natriuresis relationship and the development of hypertension in these animals.

Recently, we tested this concept by long-term infusion of captopril, a potent medullary vasodilator, into the renal medullary interstitial space of uninephrectomized SHR.⁴⁸ As shown in Fig 6⇓, this increased medullary blood flow by 40%, whereas renal cortical blood flow remained unchanged throughout the 5 days of medullary infusion of captopril. MAP fell 20 mm Hg while captopril was infused, and this was associated with a negative sodium balance. Medullary blood flow returned to control after captopril infusion, and arterial pressure rebounded to levels that exceeded those seen at the beginning of the study. In another rat group exposed to chronically low and high levels of sodium intake during medullary interstitial captopril infusion, it was seen that the chronic renal function curve (the steady-state relationship between sodium excretion and MAP) was shifted to a lower level of arterial pressure compared with that seen in vehicle-treated SHR. Captopril infused intravenously at the same low dose given into the renal medulla had no effect on arterial pressure. This finding excluded the possibility that the antihypertensive effect of captopril given into the renal medulla of SHR was due to recirculation of the infused compound.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Line graphs show time course of changes in mean arterial pressure (MAP), renal medullary flood flow (medullary flow signal), and renal cortical blood flow (cortical flow signal) in conscious spontaneously hypertensive rats receiving captopril into the renal medulla (r.i.) for 5 days. *P<.05 from final control day. (From Lu et al.⁴⁸ )

These results provide the first direct evidence in unanesthetized rats that the renal medulla plays an important role in the development of hypertension in SHR. The reduction of arterial pressure in this model once again demonstrates a striking parallelism with the increase in medullary blood flow.

Escape From Long-term Actions of ANP and Calcium Channel Inhibitors

Several other studies carried out in our laboratory have contributed to our understanding of some of the determinants of medullary flow and their relationship to the long-term control of arterial pressure. In one of these studies, we evaluated the effects of the ANP atriopeptin III (APIII) on pressure natriuresis and papillary blood flow. Previous studies have demonstrated that ANP can chronically lower blood pressure without causing sodium retention, indicating that renal excretory function must be reset in some way to enable sodium balance to be achieved at a lower renal perfusion pressure.⁷⁴ Since our studies had shown that changes in renal medullary hemodynamics are associated with the pressure-natriuresis response, we first examined the effects of APIII on the renal response to changes in renal perfusion pressure in anesthetized rats.⁷⁶ We found that APIII infusion altered the relationships between the fractional excretion of sodium and water and renal perfusion pressure without producing sustained alterations of GFR or total RBF. APIII infusion also produced a 15% increase of papillary blood flow, whereas cortical flow was not significantly altered. In other studies, we found that the natriuretic response to APIII was associated with a rise in renal interstitial pressure and a marked increase in the distal delivery of sodium.^{62 76} Removal of the renal medulla to prevent elevations in papillary blood flow completely blocked the changes in renal interstitial pressure and the natriuretic response to APIII.⁷⁷ These studies suggested that small increases of circulating ANP could influence the long-term control of arterial pressure by preferentially altering renal medullary hemodynamics, raising renal interstitial pressure, and promoting the elimination of sodium and water in the absence of changes in total RBF or GFR.

We have since carried out long-term studies infusing APIII (50 ng/kg per minute) into the medullary interstitial space for 5 days in four uninephrectomized Sprague-Dawley rats, as shown in Fig 7⇓.⁷⁸ MAP decreased 20 mm Hg from normal control levels for the first 3 days of infusion but then began to return toward control levels. At the end of the 5-day infusion period, pressure had returned halfway (10 mm Hg) toward control levels. Although we did not carry out daily measurements of medullary blood flow in this study, the APIII dose used was the same as that found in short-term studies to be sufficient to increase medullary blood flow when infused into the renal medulla in the absence of changes of cortical blood flow or MAP.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Line graph shows changes in mean arterial pressure (MAP) in conscious, uninephrectomized Sprague-Dawley rats (n=4) receiving an infusion of atriopeptin III (APIII) into the renal medulla (r.i.) for 5 days. *P<.05 from final control day. (From Lu.⁷⁸ )

Short- and long-term studies with calcium antagonists also support an important role of the medullary circulation in the long-term control of arterial pressure. In one study, responses to intravenous nisoldipine were compared in SHR and WKY.¹⁹ These studies showed that papillary flow increased to a greater extent in SHR than in WKY and normalized the pressure-natriuresis relationship between the SHR and WKY. In another study, the effects of the calcium channel antagonist diltiazem on papillary blood flow were evaluated.⁴⁷ Studies were carried out to examine whether calcium antagonists would increase renal medullary blood flow and shift the relationship between sodium excretion, papillary blood flow, and renal perfusion pressure. Infusion of diltiazem into the medullary interstitial space at a dose that produced no significant changes of arterial pressure, total RBF, or renal cortical blood flow⁴⁷ increased papillary blood flow by 26%, and this was associated with a 65% rise in sodium and water excretion. GFR was unchanged after an initial rise during the first 20 minutes of diltiazem infusion, whereas the fractional excretion of sodium remained 80% above control values, paralleling the changes in papillary blood flow that remained elevated for more than 1 hour after drug administration.

We then carried out long-term studies in which we infused diltiazem chronically into the medullary interstitial space (100 μg/kg per minute) of six uninephrectomized SHR.⁷⁸ As observed with captopril, MAP decreased by nearly 25 mm Hg from a control level of 165 mm Hg during the first day of infusion. However, this reduction was not sustained, and after 5 days of medullary infusion of this compound, arterial pressure returned to control levels. The diltiazem dose chosen for this study was the same as that studied in normal anesthetized rats (see above) and which increased papillary blood flow by 26%.

These studies once again indicate that arterial pressure can be chronically influenced by the dilation or constriction of the medullary circulation. However, these results also demonstrate that as with many control systems, counterregulatory systems exist that may allow for escape from the primary or initial effects of certain stimuli. It is evident that the mechanisms which regulate medullary blood flow and deep nephron function are complex and likely involve interactions with many of the other factors regulating sodium excretion. Depending on the stimuli used to modulate blood flow to this region, it is evident that different counterregulatory mechanisms such as activation of the renin-angiotensin system or sympathetic nervous system secondary to volume contraction could either reinforce or override the initial response.

Control of Medullary Blood Flow

The studies summarized above suggest that changes of renal medullary blood flow have the ability to influence the long-term control of arterial blood pressure. Specifically, it is clear that blood flow to this region is an important component of the pressure-natriuresis mechanism, which is of utmost importance in the long-term control of arterial pressure. As discussed in detail elsewhere,^{1 27} it is also evident that factors shown to reduce medullary blood flow are those that are commonly associated with long-term elevations of arterial pressure and salt-sensitive forms of hypertension (sympathetic nerve stimulation, cyclooxygenase inhibition, kinin antagonists, NO synthase inhibition, Ang II, and AVP). Conversely, the factors that increase medullary blood flow are those which have been often associated with the lowering of blood pressure (ANP, acetylcholine, prostaglandins, converting enzyme inhibitors, bradykinin, and calcium antagonists). The studies reviewed in the preceding sections extended these studies to determine the long-term consequences of changing medullary blood flow by manipulating some of these systems.

Summary

The evidence summarized above establishes that the renal medulla plays an important role in sodium and water homeostasis and in the long-term control of arterial blood pressure. Many of the potentially important physiological and pathophysiological determinants of medullary blood flow, ranging from intrarenal paracrine systems, circulating hormones, and renal sympathetic nerve activity, have been characterized. Although blood flow to the inner renal medulla comprises less than 1% of the total RBF, changes in blood flow to this region can have an important effect on the set point of the pressure-natriuresis relationship and thereby influence the long-term control of arterial pressure. Studies in SHR suggest that the initial changes of renal function begin as a subtle preferential reduction of medullary blood flow in the absence of more global changes in renal hemodynamics such as cortical flow, RBF, and GFR. Many studies have shown that with prolonged established hypertension, the renal circulation throughout the entire kidney eventually becomes involved with the development of generalized glomerular sclerosis and end-stage renal disease. Results of studies in which medullary blood flow was preferentially reduced by blockade of NO synthase or medullary infusion of AVP clearly demonstrate that reduction of medullary blood flow alone is sufficient to produce systemic hypertension. However, considerable work remains to find the genes and establish the genetic abnormalities that increase medullary vascular resistance and promote the development of hypertension in hereditary forms of hypertension in humans and genetic animal models such as the SHR.