INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

PROXIMAL TUBULAR REABSORPTION and glomerular filtration are closely coordinated by the integrated actions of glomerulotubular balance (GTB) and tubuloglomerular feedback (TGF). GTB is a physiological process that affects tubular reabsorption, bestowing a forward dependence of late proximal flow (V_LP) on nephron glomerular filtration rate (SNGFR). TGF is a physiological process through which a signal derived from tubular fluid at the macula densa engages the glomerular microvasculature, resulting in a negative feedback dependence of SNGFR on V_LP. The relationships among SNGFR, V_LP, GTB, and TGF can be depicted in a single plane defined by V_LP and SNGFR. In this plane TGF is typically represented by a sigmoid curve with a negative slope and GTB is typically represented by a straight line with a positive slope. These two curves intersect at a single point. This is the "operating point" of the nephron. Neither SNGFR nor V_LP can deviate from their values at the operating point without a change in the behavior of GTB, TGF, or the TGF-independent inputs to SNGFR (10, 11).

One important property of this GTB-TGF system is its tendency to stabilize its own internal variables, thus minimizing the effects of hemodynamic fluctuations on the rate of salt and water excretion. The efficiency with which the system performs this task is monotonically linked to the slope of the TGF curve at the operating point and is typically optimized by alignment of the operating point along the steep portion of the TGF curve. Changes in the systemic milieu that affect TGF-independent influences over SNGFR and alter proximal reabsorption will be partially compensated by GTB-TGF but will also result in movement of the operating point along the TGF curve. If this movement places the operating point at a shallow point on the TGF curve, then the ability to compensate for subsequent perturbations will be reduced. This loss of TGF efficiency would be mitigated if TGF were to shift over time to realign the steep portion of the TGF curve with the new operating point. This phenomenon of shifting is referred to as "TGF resetting." In a previous study (11) we augmented early proximal flow by 20 nl/min in individual free-flowing nephrons by microperfusion with artificial tubular fluid and demonstrated that a sustained supraphysiological flow is sufficient to cause TGF resetting in that nephron. In the present study, we demonstrate TGF resetting in response to reduced proximal reabsorption elicited by systemic administration of the proximal tubular diuretic benzolamide.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Rather than attempting to characterize the entire TGF function, we derived evidence of TGF resetting by correlating changes in V_LP with changes in homeostatic efficiency of the integrated GTB-TGF system as determined by perturbation analysis in free-flowing nephrons (10, 11).

Studies were conducted in euvolemic adult male Munich-Wistar-Frömter rats weighing ~250 g. Animals were prepared for micropuncture according to standard protocols used in our laboratory (3). Briefly, animals were anesthetized with Inactin (100 mg/kg body wt ip), and their body temperature was maintained at 37°C. After tracheostomy (PE-200) and placement of catheters in the jugular vein, femoral artery, and urinary bladder (PE-50), the left kidney was exposed by flank incision, immobilized in a Lucite cup, and covered with warm Ringer saline, and the left ureter was cannulated (PE-50). In addition to Ringer saline (1.5 ml/h), animals were made euvolemic by infusion of donor rat plasma as replacement for surgical losses (1% body wt over 1 h followed by continuous infusion at 0.15% body wt/h). For microperfusion of late proximal tubules, we used artificial tubular fluid (ATF) with the following composition (in mM): 130 NaCl, 10 NaHCO₃, 4 KCl, and 2 CaCl₂ as well as 45 mg/100 ml urea and 0.1% FD & C, pH 7.4. Arterial blood pressure was monitored continuously (23Db Gould-Statham pressure transducer) throughout each experiment.

The homeostatic efficiency of the TGF-GTB system was determined by perturbation analysis in free-flowing nephrons, with the use of videometric flow velocitometry and a theoretical paradigm as previously described (11).

Theoretical Paradigm

Videometric Flow Velocitometry

Experimental Protocols

Closed-loop perturbation studies. Seven experiments were conducted in which V_LP was perturbed in free-flowing nephrons by addition or subtraction of fluid (V_H) using a Hample microperfusion apparatus as previously described (10, 11). Flow rate (V_M) was measured immediately upstream from the perturbation by videometric flow velocitometry (VMFV). Perturbations were applied in cyclical fashion. A perturbation cycle was completed every 7.5 min. In one-half the nephrons, the cycle ran 0, , +, 0. In the other one-half, the order was reversed (0, +, , 0) for  = 8 nl/min for the first four nephrons, = 5 nl/min for the fifth nephron, and  = 7.2 nl/min for the sixth and seventh nephrons. After each change in V_H, 2 min were allowed for reequilibration, followed by a 30-s recording. The hypothesis that a sustained increase in tubular flow induces resetting of TGF was tested by examining V_M/V_H repeatedly before and during the continuous infusion of the carbonic anhydrase inhibitor benzolamide (BNZ, 10 mg/kg iv over 4 min followed by 5 mg · kg¹ · h¹ continuous iv infusion). During BNZ infusion, 300 mM NaHCO₃ (25 µl/min iv) was infused in addition to standard Ringer saline and plasma replacements. This protocol has previously been verified to maintain volume and serum bicarbonate concentrations equal to the pre-BNZ condition (14). Tubular pressure (P_T) was monitored in an adjacent tubule with a servo-null micropressure apparatus (IPM, San Diego, CA).

Open-loop flow determinations. To examine the importance of TGF in blunting the effect of BNZ on V_LP, four comparison experiments were conducted in which TGF was inactivated by placement of an obstructing wax block to administration of BNZ in the late proximal tubule of an index nephron. An oil-tipped collecting pipette (12-14 µm OD) was placed in the tubule immediately upstream from the wax block. A pipette containing rhodamine B dextran was placed immediately upstream from the collection pipette. VMFV was used to continuously monitor the uptake of fluid into the collecting pipette before and during BNZ administration. P_T was monitored in the same loop and was manually held constant with a glass syringe connected to the collection pipette to minimize artifactual transients.

Statistical Analysis

For parameters measured more than once during an experiment, univariate analysis of variance with design for repeated measures was used for group statistics. Proprietary software (SYSTAT, SPSS, Chicago, IL) was used to compute P values for the univariate repeated-measures F test and a series of single degree-of-freedom polynomial contrasts. The number of terms in these polynomials is equal to n  1, where n is the number of repeats. Special attention was given to linear, quadratic, and tertiary terms because these were pertinent to the hypotheses being tested. A significant linear term implies a linear trend over time. A significant quadratic term implies an increase followed by a decrease, or vice versa. A significant tertiary term implies the existence of a true inflection point (increasedecreaseincrease or decrease increasedecrease).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Systemic blood pressure remained stable throughout all experiments. As illustrated in Fig. 1, BNZ was an effective diuretic with a rapid onset of action. Urine flow remained relatively constant throughout the duration of BNZ infusion. The effect of BNZ on proximal P_T in free-flowing nephrons is also illustrated in Fig. 1. The initial bolus infusion of BNZ (250 µl over 4 min) elicited a transient increase in P_T. For the ensuing ~40 min, P_T remained stable at ~1.1 mmHg above pre-BNZ baseline despite concomitant gradual increases in V_LP and TGF efficiency (see below). After ~45 min of BNZ infusion, P_T began to increase, stabilizing ~30 min later at ~7.5 mmHg above pre-BNZ levels. During this phase of increasing P_T, V_LP and TGF efficiency were no longer increasing and TGF appeared to become less efficient.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Urine flow and tubular pressure (PT) as functions of time before and during benzolamide (BNZ) infusion.

Closed-Loop Studies in Free-Flowing Nephrons

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Ambient late proximal flow (VLP) and fractional compensation for perturbations applied to VLP as functions of time during BNZ infusion. Results of significance tests given in Table 1.

TGF efficiency was reduced immediately after starting BNZ [P = 0.023 for linear term of C on 15 < time (min) < 7.5]. This is consistent with our hypothesis that the TGF response is saturated soon after the start of BNZ. In all experiments, BNZ induced an initial nadir in TGF efficiency followed by a gradual increase in compensation during the next 45-60 min. This is consistent with our expectation of TGF resetting during continued infusion of BNZ [P = 0.004 for quadratic term of C on 15 < time (min) < 45 and P = 0.002 for linear term of C on 0 < time (min) < 45]. The time course of this resetting was remarkably similar to that observed when high flow was imposed on single nephrons in prior studies (11). Six of the seven nephrons yielded data beyond 60 min. In three of these six, TGF exhibited a late decline. This is similar to the desensitization observed with high flow in the single nephron (11). For the group overall, however, the decline did not reach statistical significance [P = 0.067 for the quadratic term of C on 0 < time (min) < 75 and P = 0.102 for the linear term of C on 60 < time (min) < 75].

Ambient V_LP increased steadily with time during the first 45-60 min of BNZ infusion [P = 0.011 for the linear term of V_LP on 0 < time (min) < 45]. This would be expected from a pure rightward shift in the TGF function unaccompanied by any change in SNGFR₀. When supraphysiological flow was imposed on single early proximal nephrons in the past, we did not observe V_LP to increase further during TGF resetting. This required SNGFR₀ to decrease coincidently with TGF resetting (11).

Open-Loop Studies in Blocked Nephrons

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Changes in ambient VLP from baseline during BNZ infusion under closed-loop and open-loop conditions. These experiments demonstrate that in absence of tubuloglomerular feedback (TGF), VLP increases more rapidly. When TGF is operative, increase requires TGF resetting.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

For GTB-TGF to participate in renal autoregulation, ambient tubular flow must lie within the narrow range where the gain of the TGF function is concentrated. Any major change in steady-state flow, such as occurs after acute volume expansion, growth, or contralateral nephrectomy, etc., must be accompanied either by the loss of TGF efficiency or by TGF resetting. In fact, TGF resetting has been documented during acute volume expansion (10) and normal physiological growth (5) and after contralateral nephrectomy (2). For the case of acute volume expansion, TGF resetting is known to correlate with preserved homeostatic efficiency of GTB-TGF (10). However, the fact that changes in ambient tubular flow and TGF resetting tend to occur together does not resolve the question as to whether these two events occur as parallel consequences of changes in the systemic milieu or whether they occur in series. If changes in SNGFR, GTB, and TGF are linked in parallel, then the subsequent orientation of the operating point relative to the TGF function would essentially be left to chance. For this reason, we favor the hypothesis that, when TGF resets in response to systemic events, this resetting is actually mediated by a prior sustained alteration in tubular flow such that TGF efficiency is optimized by alignment of the TGF function with ambient flow in each nephron. The present data fulfill the main prerequisite of this hypothesis by demonstrating that sustained activation of TGF is sufficient to induce TGF resetting.

Rather than attempting to detect changes in the TGF function during precise manipulation of the signal at the macula densa, we conducted these experiments with the natural operating point as frame of reference. This "closed-loop" approach has been applied previously and has been found more useful for detecting subtle changes in the operational characteristics of the TGF system than the more traditional technique of open-loop microperfusion (10, 11). In the present experiments, TGF resetting was manifested by a gradual increase in ambient V_LP linked to a gradual restoration of the ability of TGF to compensate for perturbations in ambient tubular flow. These events began within 10 min after imposition of a sustained reduction in proximal reabsorption and continued to evolve for ~60 min. The gradual increase in V_LP in free-flowing nephrons during this interval of time probably reflects a gradual relaxation of the TGF-mediated difference between SNGFR₀ and ambient SNGFR. Alternative explanations, including an isolated increase in TGF-independent SNGFR₀ and/or a secondary slow effect of BNZ on proximal reabsorption, are not tenable because an isolated increase in SNGFR₀ would not restore TGF efficiency and the slow increase in V_LP was only observed when TGF was allowed to operate.

The tonic influence of TGF over SNGFR (in other words, the difference between SNGFR₀ and SNGFR) will be reduced during TGF adaptation whether that adaptation takes the form of a rightward resetting or an overall flattening of the TGF function. In the present case, adaptation not only allowed tubular flow to increase but also restored the efficiency with which TGF compensated for perturbations in ambient flow. Because flattening of the TGF function would lessen TGF efficiency, TGF adaptation during the first hour of BNZ must assume the form of rightward resetting, which restores the operating point to the steep portion of the TGF function while relaxing the tonic influence on SNGFR. This point is illustrated in Fig. 4.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Paradigm for changes in TGF during BNZ infusion. Dotted hyperbolic tangent (dotted curved line) and dotted straight line refer to TGF and glomerulotubular balance (GTB) functions, respectively, before start of BNZ. Solid straight line refers to GTB function during BNZ infusion. Two solid hyperbolic tangents correspond to TGF functions immediately after start of BNZ (left solid curved line) and after TGF resetting (right solid curved line). A-C, sequential operating points. To generate this figure, hypothetical GTB functions were derived by setting f2' = 0.5 for control state, setting initial operating point A at steepest part of TGF function, and assuming 60% reduction in fractional reabsorption with BNZ. Hypothetical TGF functions were calculated from fractional compensations measured in the present study with the above assumptions regarding f2' and from allowing f1 to saturate at 12 nl/min, typical of euvolemic rats in our lab. Shift from A to B, immediate effect of BNZ on basis of assumption that 20% of effect of BNZ on afferent stimulus to TGF is mediated distal to where VLP is measured. Note that ambient VLP is very sensitive to this effect and decreases slightly from A to B despite a major reduction in upstream reabsorption. As TGF resets rightward, operating point gradually drifts from B to C, where ambient VLP exceeds pre-BNZ baseline. Single-nephron glomerular filtration rate (SNGFR) at C is intermediate between SNGFR at A and B. Fractional compensation is suppressed on moving from A to B and restored on moving from B to C.

When flow is perturbed in single nephrons, the compensatory change in SNGFR is mediated principally by TGF, with a negligible contribution from direct effects of the perturbation on P_T (7). However, when flow out of the proximal tubule is simultaneously increased in all nephrons, the ensuing increase in interstitial pressure can increase distal flow resistance to the point where a significant portion of the inverse dependence of GFR on tubular flow might be mediated by changes in P_T (8). In the present study, a transient 3-mmHg surge in P_T accompanied the initial bolus infusion of BNZ. Subsequently, P_T relaxed to 1.1 mmHg above the pre-BNZ baseline and remained stable throughout the ~40-min period of TGF resetting, despite steady increases in V_LP. A gradual increase in P_T occurred from ~45 to 60 min into BNZ, commensurate with termination in the rise of V_LP. This delayed rise in P_T at relatively constant V_LP requires a gradual increase in distal flow resistance. Leyssac et al. (8) have shown that, when applied to the whole kidney, distal flow resistance increases with distal flow. This behavior is opposite to that expected of a Starling resistor and must be accounted for by changes in interstitial pressure impinging on the distal nephron. The present observations, in consideration of the time course of P_T, may reflect gradual filling of the interstitium, a greater tendency to fill as flow increases during TGF resetting, and/or nonlinear compliance of the interstitial compartment.

Data from the open-loop experiments suggest that the increase in distal flow resistance that coincides with equilibration of V_LP is not the main reason why V_LP did not continue to increase beyond 50 min in free-flowing nephrons. In the open-loop experiments in which V_LP was immune to the effects of both TGF and distal flow resistance, the increment in V_LP achieved with BNZ was not different from that observed after TGF resetting in closed-loop nephrons that were subject to the influence of distal flow resistance. If changes in distal flow resistance were a major factor in determining the ultimate V_LP in the closed-loop experiments, then a lesser increment in flow should have been observed in the free-flowing vs. the wax-blocked nephrons. Therefore, it may be pure coincidence that the onset of the rise in P_T and the end of the rise in V_LP occurred simultaneously.

The effects of BNZ on proximal reabsorption and the role of TGF in the reduction in GFR with BNZ have been previously established (13). Because TGF is not perfectly efficient, an error signal must persist at the macula densa to reflect the portion of the stimulus for which TGF fails to compensate. If BNZ had affected reabsorption only proximal to the site where flow was measured, then this error signal would have translated into an increase in V_LP throughout the experiment. However, a slight decrease in tubular flow was observed immediately after starting BNZ. This implies that BNZ altered the relationship between proximal tubular flow at the point where it was measured and the signal at the macula densa. Most proximal tubular carbonic anhydrase is located in the S1 segment, and our measurements were made downstream from the S1 segment. However, if our compensation data are combined with reasonable estimates for the effects of BNZ on GTB, only ~20% of the effect of BNZ on the TGF signal need have been mediated at sites distal to where flow was measured to account for the initial decrease in flow that we observed (Fig. 4). Also, this downstream effect of BNZ need not have required downstream carbonic anhydrase inhibition per se, because the BNZ bolus was administered in a hypertonic solution which may have transiently increased tonicity along the length of the nephron, thereby amplifying the TGF signal (9). Furthermore, water retained in the tubule during inhibition of early bicarbonate reabsorption from the S1 segment would decrease the transepithelial gradient that drives chloride transport in the later proximal segments. Regardless of the mechanism, the effect of BNZ is sufficient to shift the operating point far enough rightward along the TGF curve to substantially reduce TGF efficiency. Furthermore, inhibiting tubular reabsorption increases the slope of the GTB function, thereby tending to increase TGF efficiency. Therefore, the decrease in fractional compensation observed during the course of BNZ infusion actually underestimates the true effect on the slope of the TGF function at the new operating point.

In a prior study, we demonstrated TGF resetting when early proximal flow was supplemented by 20 nl/min in a single nephron (11). The time course of resetting in those experiments was similar to that now reported during BNZ. However, the present findings do differ from the former in at least one respect. In the present study, V_LP increased throughout the course of TGF resetting until the tonic influence of TGF over V_LP approximated control values. In the former study, there was an early 7-nl/min increment in V_LP that constituted the residual after immediate buffering by TGF. However, no further increase in V_LP was noted during rightward resetting of TGF. For the TGF profile to have gradually recentered about the higher ambient flow without a gradual increase in ambient flow, we postulated a reduction in SNGFR₀ to coincide with TGF resetting in those experiments (11). By arbitrary definition, SNGFR₀ must remain independent of TGF. With restriction of the definition of TGF to include only interactions between the tubule and glomerulus with time constants on the order of 30-60 s, a slow reduction in SNGFR₀ mediated over 30-60 min by a sustained increase in tubular flow is permitted by our theoretical paradigm. A similar secondary reduction in SNGFR during sustained supraphysiological distal flow has also been reported by Briggs et al. (5). In contrast, the events that occurred during the first 60 min of BNZ infusion in the present studies appear more straightforward and consistent with a pure rightward shift in the TGF function. The increase in distal flow imposed on the index nephrons in the prior study (5) was of greater magnitude than could be achieved by reducing proximal reabsorption with BNZ. Perhaps a lesser stimulus is required to induce TGF resetting than to induce a reduction in SNGFR₀, such that a threshold for the latter effect may not have been achieved in the present study. As a caveat, beginning from an operating point on the steep portion of f₁, fractional compensation is less sensitive to reduction from a shift in SNGFR₀ than from a shift in ambient V_LP. Therefore, although the present data are accounted for without a change in SNGFR₀ during TGF resetting, they do not preclude such a change. With or without a change in SNGFR₀, the data still imply TGF resetting.

One main advantage of the closed-loop perturbation analysis is that the gain of the system does not depend on where in the circuit the perturbation is applied (10). However, the technical constraint that perturbations be applied upstream from the loop of Henle leads to a paradigm in which events occuring in the loop of Henle are incorporated into the process of TGF. Therefore, the present data do not permit us to distinguish whether TGF resetting during BNZ is mediated by increased loop transport, decreased macula densa transport, or an altered effector response. These are subjects for further study.

Perspectives