INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

LUDWIG (17) was probably one of the first to discover the profound impact of blood pressure on urine flow. It was then Guyton and Coleman (8) who recognized the importance of this pressure diuresis and pressure natriuresis as a feedback-controlled system for long-term blood pressure regulation. According to this hypothesis, the pressure dependence of urine flow (UV) and sodium excretion constitutes the pivotal element of long-term blood pressure control (9). Although the organism, to some extent, can adapt to servo control of renal perfusion pressure (RPP) by attenuating renal fluid loss (25), the phenomenon of pressure diuresis as such is well established (5, 8, 14, 19). The underlying mechanisms, however, are still unclear. Recently, several mechanisms have been discussed as possible mediators of pressure diuresis (7), and recent studies have provided conflicting results as to whether changes in medullary blood flow of the kidney are involved in this process (18, 20). According to this hypothesis of pressure diuresis, the medullary portions of the kidney do not possess a pronounced autoregulatory capability, in contrast to the renal cortex. Hence, an increase in local perfusion pressure augments medullary blood flow, leading to a washout of the osmotic gradient found in this region of the kidney. Because the osmotic gradient is essential for retaining sodium, elevations in blood pressure increase urinary sodium excretion.

A second mechanism often suggested to influence renal hemodynamics and UV is the so-called tubuloglomerular feedback mechanism (TGF) (21, 29). Several studies have revealed that dynamic alterations of TGF over ~20-30 s are closely intertwined with fluctuations of not only arterial blood pressure (AP), but also renal blood flow (RBF) (2, 10). Consequently, physiological variations of blood flow, in addition to, or even independent of, changes in blood pressure, may modulate intrarenal mechanisms such as the TGF. In light of these studies, it appears likely that physiologically occurring short-term blood pressure oscillations (BPO), which can bypass the renal autoregulation, or spontaneous variations in RBF may have a profound impact on short-term UV.

It was the goal of the present study, therefore, to clarify the influence of short-term variations in RPP and RBF on UV in 13 resting conscious foxhounds on a normal-sodium diet. This was achieved by the following protocols. 1) A single-step reduction in RPP was performed to determine the response times of RBF and UV in the conscious dog. 2) RPP, RBF, and UV were measured under control conditions over a period of 4 h. After correction of the RBF and UV values for the response times, we investigated the extent to which spontaneous short-term variations in UV were coupled with fluctuations in RPP or in RBF. 3) To clarify the individual effects of RPP and RBF on UV at a reduced RPP, we superimposed sinusoidal pressure oscillations with a frequency of ~0.9 mHz on RPP. This induces a phase shift of ~180° between RPP and RBF in the conscious dog; thus RPP and RBF become uncoupled (30).

	METHODS

Top Abstract Introduction Methods Results Discussion References

All experiments were performed on 13 female, chronically instrumented, conscious, adult foxhounds, weighing 17.5-25 kg, that had free access to water and received a standard dog diet (Alma H 5003, 4 g/kg Na⁺). The daily sodium intake was ~110 mmol. Experiments were approved and performed according to the German Animal Protection Law.

Implantation Surgery

Measurements

5

View larger version (29K): [in this window] [in a new window]   Fig. 1.   A single-step reduction of renal perfusion pressure (RPP) to 80 mmHg (direct signal, top) induced a rapid fall in renal blood flow (RBF, middle) as well as in urine flow (UV, bottom). Thus such fast changes in RPP can bypass RBF autoregulation and even decrease UV.

During the experiments, AP, RPP, HR, RBF, and UV (analog output of the drop counter) were recorded directly with an analog recorder (model 2600, Gould). After analog-to-digital conversion, all data were stored on-line by a computer system (Labtech Notebook, IBM-compatible personal computer). HR and UV were sampled every 5th s. Direct signals of the AP, RPP, and RBF were concomitantly stored with a sampling rate of 10/s.

Experimental Protocols

Protocol 1: estimation of transition times between RPP, RBF, and UV. At resting blood pressure, AP, RPP, and RBF were continuously recorded with a sampling rate of 10/s (HR and UV with a sampling rate of 0.2/s). After stabilization, the extracorporeal control system was adjusted to reduce mean RPP to 80 mmHg. The pneumatic cuff was deflated, and ~10-20 min were allowed for recovery. Thereafter, the cuff was inflated brusquely using the electrical values from the memory of the control system, which were obtained during the first short period of pressure reduction. Thus RPP was reduced in a single step to 80 mmHg. The data obtained by the pressure reduction in a single step were used to calculate response times between RPP and UV as well as between RBF and UV. These values were then used to correct for the time lag of the RBF and UV response.

Protocol 2: influence of spontaneous variability of RPP and RBF on UV. After stabilization, RBF, RPP, HR, and UV were recorded for 4 h without intervention. Throughout the experiments the dogs lay quietly on their right side. In addition to the electrical data obtained via the pressure processors and the flowmeter, the analog output of the drop counter was used to determine the UV with high precision.

Protocol 3: dissociation between RPP and RBF. RPP was adjusted via the extracorporeal control system to 80 mmHg. Then the control system was used to superimpose sinusoidal blood pressure waves with an amplitude of about ±10 mmHg and a frequency of 0.9 mHz. Thus RPP varied continuously between ~70 and ~90 mmHg. A period of 30 min was allowed for stabilization before the data collection began. Thereafter, RBF, RPP, and UV were determined as described for protocol 2.

Data Analysis

Protocol 1. The last 10 min before the pressure step were used to calculate mean ± SD from the baseline data. A linear interpolation was used to connect data points.

Protocols 2 and 3. Mean values were derived from the original data by averaging within a 60-s window. Thus every 4-h recording resulted in 4 × 60 = 240 data points. Interindividual means were obtained by averaging the data triplets (RPP, RBF, UV) of every dog with the data triplets of the other dogs using RPP as a matching criterion. The interdependence among RPP, RBF, and UV was determined by a subsequent correlation analysis. The two-tailed Student's t-test was used for comparisons. P < 0.05 was taken to indicate significance. Values are means ± SE unless stated otherwise.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

No significant changes in HR (101 ± 4, 97 ± 4, and 94 ± 6 beats/min for control, step function, and induced BPO, respectively) or in AP (105 ± 3, 104 ± 3, and 119 ± 4 mmHg for control, step function, and induced BPO, respectively) were observed between the protocols.

The response time of RBF and UV vs. RPP was determined in seven dogs by single-step pressure reduction. Figure 1 shows original recordings from a single experiment. Inflation of the aortic cuff (dashed vertical line) led to a precipitous fall in RPP from 104 ± 3 to ~80 mmHg (Fig. 1, top). The sudden change in RPP caused RBF to decrease within a few seconds from 332 ± 38 to <100 ml/min. After this transient fall, RBF increased slowly but did not reach the baseline level within the following 50 s. UV, as measured via the drop counter at the end of the ureteral catheter, followed the precipitous reduction in RPP (Fig. 1, bottom) with a response time of ~10 s. The response time of the experiments was determined as indicated in Fig. 2: a 2 SD reduction from the baseline was compared with the corresponding points of RBF and UV. The resulting response time was <1 s between RPP and RBF (0.4 ± 0.1 s, n = 7; Fig. 2, middle). In contrast to this short latency, the response of UV to the sudden fall in RPP was calculated to be 8.1 ± 0.8 s (n = 7, P < 0.01 vs. RBF; Fig. 2, bottom). This sluggish response is not due to dead space, since the ureteral catheter was always filled and maintained on a constant suction via the extracorporeal pump.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   A deviation of 2 SD from baseline level (intersection of dashed line with line labeled "2SD") was used to calculate response times. Recordings of 7 dogs in this group revealed that UV follows the sudden fall in RPP (bottom) ~20 times slower than RBF. Time lag, however, was short enough to allow even fast oscillations in RPP to modulate UV within only a few seconds.

The 60-s mean values of the 13 dogs of the control protocol are shown in Fig. 3. To investigate the influence of spontaneous changes in RPP and RBF on UV, the UV data were plotted against RPP (x-axis) and RBF (y-axis). During the recording period, mean RPP varied spontaneously from ~95 to 115 mmHg, without any systematic change ("trend"). As indicated by the course of the surface grid, spontaneous changes in UV were more or less randomly related to increases or decreases in RPP. In contrast, the steep slope of the surface grid with respect to RBF revealed a close coupling of RBF on UV under spontaneous conditions. Mean RBF was 334 ± 36 ml/min during this control period (Fig. 4). As shown in Fig. 4, top, the physiological short-term variability of RBF was well correlated to the observed changes in UV (r = 0.94, P < 0.001). Thus a spontaneous, short-term increase in RBF by 10% of the baseline level induced an increase of ~23% in UV. Surprisingly, we observed no significant correlation of RPP with UV, as indicated in Fig. 4, bottom (r = 0.09, P = NS).

View larger version (108K): [in this window] [in a new window]   Fig. 3.   Interdependence of RPP (mmHg), RBF (ml/min), and UV (ml/min) in 13 resting dogs. Spontaneous short-term changes in UV were linked to spontaneous changes in RBF (steep slope of surface grid), in contrast to physiological oscillations in RPP, which were not closely coupled with changes in UV.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Under resting conditions, spontaneous short-term oscillations in RBF are correlated to spontaneous oscillations in UV (top, 60-s mean values from 13 dogs). Correlation analysis between RPP and UV of same data (bottom) revealed no comparable relationship between short-term changes in RPP and UV.

During protocol 3, RPP was forced with a frequency of 0.9 mHz around a mean value of 80 mmHg (Fig. 5, RPP). As indicated by the pulsatile signals of RPP and RBF, this dissociated the maxima in RPP and the maxima in RBF, in contrast to the maxima in UV (Fig. 5, RBF and UV), which coincided with the maxima in RPP.

View larger version (32K): [in this window] [in a new window]   Fig. 5.   Forcing RPP with a sine wave of 0.9 mHz (RPP) phase shifted RBF by ~180° (RBF), as indicated by minimum in RBF, which is reached around peak in RPP. In contrast, urine formation (UV) showed only a small phase shift with respect to RPP during this protocol.

Figure 6 shows mean values of the eight dogs from protocol 3, in which pressure oscillations were superimposed on RPP (81 ± 2 mmHg). As indicated in Fig. 6, top, RPP was forced with a period length of ~1,100 s (1,112 ± 4 s) and an amplitude of nearly ±10 mmHg. Because of the time constants of the systems involved in renal autoregulation, the minimum in RBF (Fig. 6, middle) was reached when the maximum in RPP was observed. In contrast to this dissociation between RPP and RBF, UV (Fig. 6, bottom) reached its maximum with a phase shift of only a few seconds to RPP (7 ± 3 s). The interdependence of RPP, RBF, and UV is demonstrated by a three-dimensional surface grid (Fig. 7). An overall steep slope of the surface with respect to RPP as well as with respect to RBF was observed. Thus the minimum in UV coincided with low RPP and high RBF, whereas the maximum in UV was observed at high RPP and low RBF. Figure 8 shows the correlation analysis of the 240 mean values of these experiments. The slope of the RPP-UV regression line (Fig. 8, top) was 6.38 µl UV · mmHg¹ · min¹. Thus a 10% increase in RPP led to an ~15% increase in UV (r = 0.95, P < 0.001). Because of the induced phase shift, changes in RBF were inversely correlated to the oscillations in UV (r = 0.85, P < 0.001, slope = 3.45 µl UV/ml).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Mean values of 8 dogs. Time course of UV indicates that pressure increase, and not an increase in RBF, enhances urine formation under these conditions.

View larger version (98K): [in this window] [in a new window]   Fig. 7.   Three-dimensional plot of UV (ml/min) vs. RBF (ml/min) vs. RPP (mmHg) shows that, during superimposition of RPP oscillations, changes in RPP are linked to UV (slope of surface grid). This pressure dependency was not overcome by inverse RBF oscillations.

View larger version (25K): [in this window] [in a new window]   Fig. 8.   When blood pressure was oscillated around 80 mmHg, increases in RPP were closely coupled with increases in UV (top), in contrast to phase-shifted RBF, which was inversely linked to UV (negative slope of regression line, bottom).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The data of the present study reveal a close relationship between spontaneous oscillations in RBF, i.e., changes around a normal baseline level of minutes to several hours, and UV. Surprisingly, we observed no coupling between the spontaneous variability of RPP and UV during these control experiments (Fig. 4). At first sight, this observation seems to be in contrast to the well-established phenomenon of pressure diuresis (3, 8, 9). However, as indicated by protocol 3, if RPP was manipulated, we observed a marked pressure dependency of UV.

These data describe only the short-term behavior of urine production. Changes in, e.g., vasopressin levels and alterations in RPP will most likely compensate for longer periods of inadequate urine formation. Several points must be kept in mind when the results of our experiments are compared with other studies that focus on pressure diuresis. In the present study we induced sinusoidal RPP oscillations to phase shift RBF from RPP. It is inherent to this approach that we control RPP and hence eliminate some of the physiologically occurring RPP variations at other frequencies. However, this protocol did not depulsate the signal, as seen in Fig. 5.

A problem of earlier approaches was the need for averaging. Rapid changes in urine formation are not readily measured in the awake animal. The determination of short-term UV suffers from biologic (bladder, renal pelvis) or experimental (catheter) dead spaces. A simple way to partly overcome these effects is the use of longer sampling periods.

In our experiments we used a ureteral catheter, which was advanced into the left renal pelvis together with a suction pump and a drop counter, to determine UV with the highest possible resolution and with minimum dead space. Thus we were able to investigate the interdependence between physiological oscillations in RPP, RBF, and UV, even in the range of several seconds. The data of the present study reveal very short response times of only several seconds between changes in RPP and subsequent modulations in RBF or UV (protocol 1, Fig. 1). This agrees with the work of Steele and co-workers (27), who used an anesthetized rat model to determine stimulus response times between RPP and UV by the use of pharmacologically or mechanically (aortic occlusion) induced changes in RPP. Under these conditions, the group observed a response time of ~5 s between induced changes in RPP and UV. The importance of RBF on UV, however, remains unclear.

A second issue that is important in the interpretation of this study is the potential masking of pressure diuresis by the sympathetic nervous system. Short-term variability in resting AP is, to a great extent, controlled by the action of the autonomic nervous system (26, 28, 32). An increase in sympathetic tone not only elevates AP but also enhances renin release and reduces UV, RBF, and renal sodium excretion (1, 5, 6, 13, 15, 16).

Despite these reservations, several studies support a role for RBF fluctuations in urine formation, and it is intriguing that spontaneously occurring RPP oscillations of minutes to several hours are poorly coupled with changes in UV (protocol 2, Fig. 4). In halothane-anesthetized rats, Yip and co-workers (31) observed spontaneous oscillations in single nephron blood flow as well as in tubular pressure. The spontaneous short-term variability in renal vascular resistance may change peritubular pressures and medullary blood flow or modulate the intrarenal transport of sodium and water (4) and, therefore, alter UV without any influence of the autonomic nervous system. In addition, changes in shear stress, as induced by changes in blood flow, increase NO production, which alters the plasma levels of renin in the conscious animal (24). This may also contribute to the observed modulations in UV (11, 23).

To clarify the observed lack of pressure diuresis and the short-term coupling of RBF on UV under baseline conditions, we used a common feature of feedback-controlled systems, the settling time. Every feedback-controlled system requires a certain period of time to compensate for initial changes in the controlled parameters. With respect to renal autoregulation of blood flow, the fastest feedback mechanism known so far is the myogenic response. Changes in RPP that are faster than the settling time of the vascular smooth muscle cells cannot, or at least not fully, be compensated for. Thus, in protocol 1, RBF and UV declined rapidly in response to the induced fall in RPP. On the other hand, very slow fluctuations in RPP, well within the range of the settling time of the RBF autoregulation, are effectively buffered (30). Between these very slow and the fast oscillations in RPP, there is a frequency range in which the settling time of the renal RBF autoregulation is similar to the half-period length of the oscillations in RPP. In this case, a decrease in RPP leads to an overshooting vasodilation. Thus minimum RPP coincides with maximum RBF, and, in turn, maximum RPP is observed at minimum RBF. In other words, there is a phase shift between RPP and RBF of 180°. In the conscious resting dog, BPO with a frequency of 0.9 mHz, corresponding to a wavelength of ~1,100 s (Fig. 6), led to this phase shift in RBF autoregulation (30). During these oscillations around a reduced RPP, an increase in RPP is extremely well coupled with an increase in UV (Fig. 7). In contrast to this observation, the effect of RBF on UV was totally overridden (Fig. 8). Hence, once blood pressure is perturbed, pressure diuresis is reestablished. It must be emphasized, however, that these data do not provide insight into the long-term importance of RBF variability for urine formation. Further studies using longer observation periods may help clarify the overall importance of "blood-flow dependent diuresis." Additionally, it must be kept in mind that the RPP occasionally exceeded the lower limit of RBF (~75 mmHg) and glomerular filtration rate (~80 mmHg) autoregulation (12) during protocol 3. This may have contributed in part to the observed pressure dependence of UV. We assume that the phase shift between RPP and RBF is not dependent on the lower limit of autoregulation; otherwise this phase shift should disappear at RPP above ~80 mmHg.

Our results reveal that when RPP fluctuates spontaneously around a normal baseline level, changes in RBF rather than in RPP coincide with changes in urine production. Conversely, experimentally induced 0.9-mHz oscillations in RPP of ~80 mmHg lead to a profound pressure-dependent response of UV, irrespective of RBF fluctuations.

Perspectives