Upsala J Med Sci 97: 107-1 14 Tubuglomerular feedback control in long-looped nephrons Topical Minireview Hans R Ulfendahl, Ann Goransson-Nyberg, Peter Hansel1 and Mats Sjoquist Department of Physiology & Medical Biophysics, University of Uppsala, Biomedical Center, Uppsala, Sweden ABSTRACT The tubuloglomerular feedback mechanism is highly activated in juxtamedullary nephrons and considered to play a major role in intrarenal regulation of glomerular filtration rate. The vasculature of juxtamedullary nephrons is highly vasoreactive with a high ability for vasodilation. This vasoreactivity is a prerequisite for an important influence of the tubuloglomerular feedback mechanism on the medullary blood flow and its regulation. The renal medulla is of particular interest since the processes which regulate the osmolar concentration of the extracellular fluid by forming a dilute or concentrated urine primarily occur in this region. The nephrons, which constitute the functional units of the kidney, are not a homogenous group of structures. Instead, those nephrons which originate closer to the cortico- medullary border (juxtamedullary nephrons), as opposed to those which originate closer to the renal surface (superficial nephrons), are the nephrons which primarily are involved in the processes giving rise to a dilute or concentrated urine. The underlying reason for this is that only juxtamedullary nephrons have long loops of Henle that protrude all the way down into the inner medulla and, furthermore, give rise to vessels (vasa recta) which perfuse the medulla. These structures produce and preserve the progressive axial osmotic concentration gradient from the cortico-medullary border to the tip of the papilla. The present paper evaluates some special features of the juxtamedullary nephrons and points out some differences in functional aspects between the two nephron sub-populations. We (8) have found that the interlobular arteries in rat kidneys behave as resistance vessels. According to calculations made by other authors (9), this is also true for dog kidneys. The blood pressure drop along an interlobular artery amounts in normotensive, normohydrated adult rats to 40-45 mm Hg. This pressure drop is a prerequisite for different haemodynamic 8-928572 107 conditions for the outermost (superficial) and innermost (juxtamedullary) nephrons within the cortex. There are indications that the glomerular capillary hydrostatic pressure is about the same in both superficial and juxtamedullary glomeruli (l), which means that the pressure-drop in afferent arterioles of superficial glomeruli is smaller than in those of juxtamedullary glomeruli. In fact, we propose (3) that the superficial afferent arterioles are nearly maximally dilated, while the juxtamedullary afferent arterioles are significantly constricted in normotensive, normohydrated rats. We have also discovered that the tubuloglomerular feedback mechanism (TGF) is normally strongly activated in juxtamedullary nephrons (3). In young Sprague-Dawley rats the papilla was exposed by excising the ureter and prepared for micropuncture procedures. The single nephron glomerular filtration rate (SNGFR) was measured by quantitative sampling of urine in a loop of Henle with an oil blockade distal to the sampling site. In this situation the urine flow to the macula densa is blocked, resulting in a stop-flow condition. In rats prepared in the same manner the SNGFR was also measured with a modified Hanssen technique (12) where a free-flow condition exists at the macula densa. During stop-flow conditions the SNGFR of juxtamedullary glomeruli amounted to 84.1 +_ 8.5 nlmin-' (n=15, mean +_ 1SEM) and during free-flow conditions it amounted to 27.7 & 2.9 nlmin-' (n=7). In other experiments (3) on the same animal model free-flow and stop-flow pressures were measured in proximal and distal tubules on the renal surface and in the loop of Henle in the papilla. From the values obtained, it was possible to estimate the net driving force for glomerular filtration at the beginning of glomerular capillaries of juxtamedullary nephrons. The results indicate a much higher net driving force during stop-flow (47 mm Hg) than during free- flow conditions (19 mm Hg). In a third series of experiments (1 1) it was possible to calculate the urine flow rate during free-flow conditions by measuring the linear velocity of small dye boli injected into the loops of Henle of juxtamedullary nephrons. The passage of the boli and the diameter of the loop lumen were recorded with a video-technique. According to the calculations the urine flow rate during free-flow conditions, 3.9 k 0.37 nlmin" (n=8), differs markedly from that during stop-flow conditions, 8.0 f 0.54 nlmin-' (n=8). All the previously given results firmly support our proposal of a strongly activated TGF in intact juxtamedullary nephrons under normal physiological conditions. Our present view on the relationship between the SNGFR and the urine flow rate at the macula densa is summarized in Fig. 1. In this figure, points A and D represent the SNGFR values obtained with the micropuncture techniques and points B and E are yielded from the modified Hanssen technique. Point C is derived from values given by Schnennann et al. (10). The value 108 of F is uncertain, as we have no experimental values available. The points B and E thus represents the operation points for the flow characteristics of normohydrated, normotensive rats. We have thus found a great influence of TGF on the glomerular filtration rate in juxtamedullary nephrons. The question then arises as to whether the blood flow to these nephrons is simultaneously influenced by the TGF. Up to date we have not been able to find a reliable method to determine the blood flow in single juxtamedullary nephrons so it has not been possible to directly study the relation between the TGF and blood flow in these nephrons. We have therefore chosen to examine the medullary blood flow which mirrors the blood flow in the juxtamedullary nephrons since only these nephrons give rise to vessels (vasa recta) which perfuse the medulla. nl . m i d . g-' K W Superficial nephron nl 0 - arb.units Flow rate at rnacula densa 0 ~I, arhunits Flow rate at rnacula densa Fig. 1. Schematic influence of urine flow rate at the macula densa on the single nephron glomerular filtration rate (SNGFR). Points A and D represent the values obtained with a blocked urine flow at the macula densa and B and E those with a normal flow. Thus B and E constitute the operation points for the two nephron populations during normal conditions. Points C and F are expected values for high urine flow rates. KW = kidney weight (from Acta Physiol Scand 14: 203-209, 1982). 109 In one series of experiments in adult Sprague-Dawley rats we estimated the plasma flow in the inner and outer medulla by the extraction of radioactively labelled rubidium (%RbCl). The method has previously been described by Karlberg et al. (7). Simultaneously, we also estimated the SNGFR with a modified Hanssen technique (described by Sjijquist et al. (12)) in three layers of the cortex, namely the outer, the middle and the inner cortex. In order to study the vasoreactivity of the vessels in the juxtamedullary vasculature, we continuously, intravenously infused the calcium entry blocker verapamil (Isoptin, Knoll AG, Germany) or the angiotensin converting enzyme inhibitor captopril (SQ14225, Squibb & Sons, NJ, USA). The results from these experiments are depicted in the left panel of Fig. 2. Neither verapamil (0.6 mg-h-'-kg-' body weight) nor captopril (3 mg.h"-kg" body weight) caused any change in the SNGFR of the different nephron populations. Thus no change in the distribution of glomerular filtration rate between different nephron populations were indicated. A redistributing effect of the drugs might, however, have been camouflaged by the drop in arterial blood pressure which for verapamil was from 127 & 3 mm Hg to 113 f 4 mm Hg and for captopril from 120 f 3 mm Hg to 102 f 5 mm Hg (4). The right panel of Fig. 2 gives the plasma flows in the outer and inner medulla. In both medullary regions verapamil and captopril markedly increased the plasma flow. Captopril- treatment caused an increase of about 90 % in the inner medullary plasma flow, which is in agreement with earlier findings by Eloy et al. (2). An estimation of the cortical plasma flow showed that captopril increased the flow by only 10-20 %. The results thus strongly indicate that the vasculature of the juxtamedullary nephrons is highly vasoreactive and has a high potential to dilate, which must be a prerequisite for the TGF to potently influence the medullary plasma flow. In another series of experiments we studied the influence of verapamil and captopril on the medullary red cell flux with a video-technique (5, 6) . In young Munich-Wistar rats the papilla was exposed and with a highly sensitive TV-camera we followed the number of fluorescently labeled (FITC) red cells passing vas rectum per minute. The same vessels were investigated before and during infusion of drugs. Figure 3 shows the results thereby obtained. Two different doses of verapamil were used and in both cases the red cell flux increased in the vasa recta; with the lower dose (0.6 mg-h-'.kg-' body weight) the increase was 30 f 6 % (p