Epithelial Na channels were investigated using patch-clamp techniques in connecting tubule (CNT) segments isolated from rat kidney. Cell-attached patches with Li+ in the patch pipette contained channels with conductances for inward currents of 13–16 pS and slow opening and closing kinetics, similar to properties of Na channels in the cortical collecting tubule (CCT). Macroscopic amiloride-sensitive currents (INa) were also observed under whole cell clamp conditions. These currents were undetectable in cells from control rats but were large when the animals were infused with aldosterone (1,380 ± 340 pA/cell at a holding potential of -100 mV) or fed a high-K diet (670 ± 260 pA/cell) for 1 wk. Under both of these conditions, currents in cells of the CNT were two- to fourfold larger than currents in cells of the CCT of the same animals. In aldosterone-treated animals, currents in cells of the initial collecting tubule (iCT) were intermediate, such that the relative magnitude of INa was as follows: CNT > iCT > CCT. Quantitative analysis of the results suggests that the maximal capacity of the aggregate population of CNTs to reabsorb Na could be as high as 18 μmol/min, or ∼10% of the filtered load of Na. This capacity is ∼10 times higher than that of the CCT.
- epithelial sodium channel
- dietary potassium
in the mammalian kidney, Na channel activity and channel-mediated Na reabsorption have been extensively studied using the cortical collecting tubule (CCT). This segment, particularly in the rabbit and in rodents, can be isolated and perfused in vitro. It can also be split open, making the apical membrane of the cells accessible to patch-clamp pipettes (16). There are several lines of evidence for Na channel-mediated transport in earlier parts of the distal nephron, particularly the connecting tubule (CNT). A variety of micropuncture data suggest that Na is reabsorbed in this segment by an amiloride-sensitive, electrogenic process (for review see Ref. 18). Early studies of isolated perfused tubules reported net Na reabsorption rates in rabbit CNTs that were larger (per millimeter of tubule length) than those in the CCT (1, 22). More recently, antibodies raised against epithelial Na channel (ENaC) subunits recognize protein in the late distal convoluted tubule (DCT) and the CNT in addition to the CCT (4, 9, 10). Finally, in a recent study of a conditional α-ENaC knockout mouse line, Rubera et al. (19) showed that animals in which ENaC activity was deleted in the collecting duct system, but preserved in earlier segments, could retain Na normally in response to a low-Na diet. Because it was also shown that amiloride treatment prevented this natriferic response, at least in rats (6), this suggested that channel activity in the CNT and late DCT was sufficient for this response and was possibly more important quantitatively than the activity in the CCT. In this study, we report measurements of channel activity in the rat CNT using cell-attached patch-clamp and whole cell clamp approaches. We confirm that channel activity in this segment is qualitatively similar to, but quantitatively greater than, that in the CCT.
Animals. Sprague-Dawley rats of either gender (100–150 g), raised free of viral infections (Charles River Laboratories, Kingston, NY), were fed standard rat chow or a high-K (10% KCl) diet (Harlan-Teklad, Madison, WI). A third group of animals was fed normal chow and implanted subcutaneously with osmotic minipumps (model 2002, Alza, Palo Alto, CA) to increase levels of circulating aldosterone. Aldosterone was dissolved in polyethylene glycol 300 at 2 mg/ml to give a calculated infusion rate of 24 μg/day.
Isolation of the CNT. The technique used to dissect and open segments of CNTs was similar to that previously described to obtain CCTs (14) with the following modifications. The rat was anesthetized with thiobutabarbital (Inactin, 150 mg/kg ip). To avoid clotting in the renal vessels and, thus, facilitate dissection, the kidneys were perfused in situ through the abdominal aorta with ∼10 ml of cold dissection solution containing heparin. From thin coronal slices of the kidneys, wedges of cortex containing mostly cortical labyrinth around a radial artery were teased apart with fine forceps. In these wedges, we searched for CNTs by separating the proximal tubules with fine needles. The CNT had a clearly different appearance and smaller diameter than the surrounding proximal tubules and could sometimes be further identified by the presence of branching (Fig. 1). Not all the tubules were branched, however, and we cannot rule out the possibility that some of the isolated tubules may have originated from the DCT. Usually, the isolated segments were short (<0.2 mm) compared with those that could be obtained from the CCT. Initial collecting tubule (iCT) segments were isolated by dissecting CCTs from medullary rays and following the tubules upstream beyond the first junction encountered.
Electrophysiology. After dissection, tubules were opened manually with a very fine needle and forceps to expose the luminal surface. The split tubules were attached to a small plastic rectangle coated with Cell-Tak (Collaborative Research, Bedford, MA) and placed in a perfusion chamber mounted on an inverted microscope. The perfusate was prewarmed to 37°C. For cell-attached recordings, the perfusing solution consisted of (in mM) 140 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 2 glucose, and 10 HEPES, adjusted to pH 7.4 with NaOH. Pipettes were filled with solution containing (in mM) 140 LiCl, 1 MgCl2, and 10 HEPES, adjusted to pH 7.4 with LiOH. Pipettes were made from hematocrit tubing, pulled in a three-step process, coated with Sylgard, and fire polished with a microforge. Pipette resistances ranged from 2 to 5 MΩ.
For whole cell clamp measurements, tubules were superfused with solutions containing (in mM) 135 sodium methanesulfonate, 5 KCl, 2 CaCl2, 1 MgCl2, 2 glucose, 5 mM BaCl2, and 10 HEPES, adjusted to pH 7.4 with NaOH. In some cases, a reduced-Cl- solution, in which KCl, CaCl2, and BaCl2 were replaced with methanesulfonate, nitrate, and acetate salts, respectively, was used. The patch-clamp pipettes were filled with solutions containing (in mM) 7 KCl, 123 aspartic acid, 20 CsOH, 20 tetraethylammonium hydroxide, 5 EGTA, 10 HEPES, 3 MgATP, and 0.3 NaGDPβS, with pH adjusted to 7.4 with KOH. In some measurements, a low-Cl- solution, in which 5 mM KCl was replaced with potassium aspartate, was used. Amiloride-sensitive currents were measured as the difference in current with and without 10 μM amiloride in the bath solution.
Counting of cells. To estimate the number of cells per millimeter of tubule, we dissected intact segments of CCT and CNT from rats fed a high-K diet for 1 wk. The tubules were attached to a coverslip coated with dried Cell-Tak and then fixed with 2% paraformaldehyde solution in PBS for 5 min. After the excess fixative was quenched and washed with PBS, the coverslip was mounted onto a glass slide with Vectashield containing 4′,6-diamidino-2-phenylindole (Vector Laboratories, Burlingame, CA) to stain the nuclei. These were then counted in a fluoresence microscope equipped with UV lighting at 400 augments, and the tubule length was measured.
Figure 1 shows the preparation of a CNT fragment from the rat renal cortex. The segments that could be isolated tended to be shorter than those that can be obtained from the CCT. A junction of two tubules, indicative of the highly branched nature of the CNTs within the cortical labyrinth, can be seen. The section of the tubule above the branch point was opened, revealing the apical membranes of the cells. Our ability to identify different cell types in these tubules was limited. However, we avoided putative intercalated cells, recognized as with the CCT by a surface that appeared to be raised above the plane of the flattened tubule. The data presented here presumably reflect a combination of principal and CNT cells, which together comprise about two-thirds of the total in this segment (3).
Single-channel openings and closures from a cell-attached patch with 140 mM LiCl in the pipette are illustrated in Fig. 2. These currents have characteristics typical of Na channels previously described in the rat CCT (14, 15, 23). Single-channel conductance was 15.9 pS in this case, and the mean value for similar channels was 14.9 ± 1.3 (SE) pS (n = 8), similar to that found in the CCT under the same conditions (23). The reversal potential was positive to the resting membrane potential, and only inward currents could be measured. In many patches, such as that shown in Fig. 2, currents at pipette potentials ≤0 were partially obscured by the presence of additional, presumably low-conductance K, channels, which mediated outward currents with faster opening and closing kinetics. The kinetics of the inward Li+ currents were slow, again characteristic of Na channels. The open and closed times were not quantitated because of the low incidence of patches containing single channels with stable activity.
The overall activity of Na channels in these cells was assessed using whole cell currents. A typical example from a cell of a rat treated with aldosterone is shown in Fig. 3. Steady-state currents were measured at test potentials of -120 to +60 mV from a holding potential of 0 mV in the presence and absence of 10 μM amiloride. Amiloride blocked about half of the current at large negative cell potentials. Currents through Na channels, measured as the amiloride-sensitive component of the whole cell currents, were inwardly rectifying and reversed at >50 mV, consistent with a high Na-to-K permeability ratio. Again, these characteristics are in qualitative agreement with results reported previously for the rat CCT (5, 7).
One significant difference between the whole cell currents in the CNT and those in the CCT was the relatively high amiloride-insensitive currents in the CNT, which in the case of the cell in Fig. 2 was ∼50% of the total current at large negative voltages. These currents were carried in large part by Cl-. With low-Cl- solution in the pipette, reduction in the concentration of Cl- in the bath from 21 to 2 mM greatly diminished outward and inward amiloride-insensitive currents (Fig. 4). The fall in outward current was presumably the direct result of decreased extracellular Cl-. The reduction in inward currents may have been caused by a decrease in the submembrane Cl- concentration. We did not attempt to further identify the channels responsible for these currents in this study. The function of this presumably basolateral Cl- conductance is unknown. Cells of the rat CNT are not thought to express the apical thiazide-sensitive NaCl cotransporter, and the existence of a significant transcellular Cl- transport pathway is uncertain (18).
The presence of amiloride-sensitive currents in the cells of the CNT depended on pretreatment of the animals with protocols designed to stimulate Na reabsorption in the distal nephron (Fig. 5A). No INa could be detected in control animals fed normal rat chow. However, large currents were observed when the rats were infused with aldosterone via osmotic minipumps or when they were fed a diet with elevated KCl content (Fig. 5). Figure 5 also illustrates the quantitative comparison of INa values in the CNT and in the CCTs from the same animals. Values for control animals were similarly low in the two segments. However, with aldosterone treatment or with the high-K diet, INa values per cell were two- to fourfold higher in the CNT.
In the animals treated with aldosterone, we also compared currents in iCTs. Here the values of INa were intermediate between those measured in the CCT and in the CNT (Fig. 5B). Thus there is an axial gradient of Na channel activity in the cortical portion of the late distal nephron, with higher levels in the more proximal segments.
There was a considerable variation in INa levels among different cells and different tubules. This variability is shown in Fig. 5C for the CNT and for the CCTs of rats treated with aldosterone. A few cells of the CNT had extremely high INa values of 5–10 nA at a membrane voltage of -100 mV. We do not know the basis for this variability. However, even if these cells are discounted, the mean currents are significantly higher in the CNT than in the CCT.
To estimate transport per unit length of tubule from whole cell currents, we counted the number of cell nuclei in tubules stained with 4′,6-diamidino-2-phenylindole. In the CNT, the mean value in 12 tubules from 3 rats was 516 ± 21 cells/mm. In CCTs, the mean value in 11 tubules from 4 rats was 509 ± 14 cells/mm.
Expression of Na channels in the CNT. This is, to our knowledge, the first direct demonstration of epithelial amiloride-sensitive INa in the CNT. The presence of these currents is, however, not surprising. Micropuncture studies indicated the presence of amiloride-sensitive Na reabsorption in the “late distal tubule,” a region of the nephron defined operationally as the most distal micropuncture-accessible site (2). This probably includes CNT segments. Furthermore, the development of antibodies that recognize ENaC, the heterotrimer comprising the epithelial channels, has enabled the expression of channel protein along the nephron to be mapped. These studies have shown ENaC protein in the late DCT, CNT, and collecting ducts of the rat and mouse (4, 9, 10).
The Na channels in the CNT have single-channel conductances (for Li+) that are similar to those in the CCT. We have not analyzed the kinetics of the channels in the CNT. However, they clearly have the same general characteristics as those in the CCT, with slow gating and long open and closed times, as well as variable open probabilities. We have no reason to believe that the biophysical properties of the channels in the earlier segments are different from those in the CCT.
Regulation. CNT cells from “control” animals fed normal rat chow had no detectable channel activity. This again is similar to the situation in the CCT (5, 12). Using in vivo micropuncture, Costanzo (2) found amiloride-sensitive Na reabsorption in the late distal tubule of control rats, leading us to suspect that the CNT might have a more substantial constitutive channel activity. The discrepancy in these results might reflect a down-regulation or endocytosis of channels that could occur during the isolation of the tubules. Such a process could result from mechanical perturbation or from the removal of hormones or other factors in vivo. However, the very low channel activity seen in vitro under these conditions is consistent with immunocytochemical studies that show that ENaC subunits are found predominantly in cytoplasmic stores in control animals (9–11).
We have used two prior conditioning treatments to increase the activity of Na channels in the CNT, both of which are also effective in the CCT. One is the infusion of aldosterone by osmotic minipumps. This leads directly to an increase in circulating aldosterone, which presumably exerts direct effects on the epithelial cells. A low-Na diet is a more physiological maneuver, which increases endogenous aldosterone secretion and activates channels in the CCT to the same extent. We have not investigated the effects of Na depletion on the CNT, because under these conditions the tubules become thinner and more difficult to dissect and open. We have also used a high-K diet to upregulate the channels. In the CCT, this treatment increases the number of conducting channels through an aldosterone-independent mechanism (13). We have not investigated the role of mineralocorticoids in mediating the effect of K loading on the CNT, but we presume that the same pathways that control the channels in the CCT are probably important.
Quantitative comparison of CNT and CCT. The amiloride-sensitive current per cell was consistently higher in the CNT than in the CCT in aldosterone-treated and K-loaded rats. This finding is in agreement with early studies of isolated perfused rabbit tubules. Here the rates of net Na reabsorption per tubule length were higher in CNTs than in CCTs studied under comparable conditions (1, 22). Because the number of cells per millimeter of tubule was not very different in the two segments, we predict that the higher conductance per cell would correspond to a greater rate of channel-mediated transport in the CNT of the rat (Table 1). In general, the overall Na conductance, at least under conditions of elevated mineralocorticoid status, has a gradient from more proximal to more distal portions of the distal nephron. The decreasing maximum transport capacity along the nephron could match Na reabsorption rates to the decrease in the delivery of Na to the more distal segments.
Physiological implications. The maximal rate of Na reabsorption through Na channels in the CNTs and CCTs can be estimated from the measured electrical properties (Table 1). For the CCT, we assume 37,500 nephrons per kidney and a convergence of 5 nephrons per CCT, giving a value of ∼7,500 tubules per kidney (8), with an average length of 1.5 mm/tubule (17). Using the measured value of 509 cells/mm tubule and assuming that two-thirds of these are principal cells (3), we estimate a total of 7.6 × 106 principal cells per two kidneys. The Na transported through the channels per cell will depend on the apical membrane voltage. We use the published value of -80 mV (20, 21), which, on the basis of the average current-voltage relations for aldosterone-treated animals, corresponds to a current of 230 pA/cell. This gives an overall transport rate of 1.8 mA or 1.1 μmol/min. For less negative apical membrane potentials, the rate would be correspondingly lower.
For the CNT, we use the same method of calculation using one tubule for each nephron. The length of the CNT is more difficult to estimate. For superficial nephrons, the segment is ∼0.5 mm long, but it is probably considerably longer in the deeper nephrons, which enter the cortical labyrinth (3). We therefore used an average value of 1 mm. We again use measured values of 516 cells/mm and assume that two-thirds are principal or CNT cells (3). Using the same value of -80 mV for the apical membrane potential, each cell has a higher transport rate of 1,180 pA/cell for an overall transport rate of 30.4 mA, or 19 μmol/min. Obviously, the estimate is a rough one, but it is clear that the transport capacity of the aggregate system of CNTs is much larger than that of the CCTs. It is also apparent that the channel-mediated maximal transport capacity of the CNTs is an appreciable fraction (∼10%) of the filtered Na load in these animals, which in a previous study was 150–200 μmol/min (5). An intermediate transport capacity was estimated for the iCTs (Table 1).
The measurements on the CNT cells were motivated in part by a study of Rubera et al. (19) in mice, in which α-ENaC gene expression and Na channel activity were selectively suppressed in the collecting ducts but not in the CNTs or late DCT. These animals could conserve Na normally in response to a low-Na diet. This was somewhat surprising, because we previously showed that treatment of rats with amiloride at doses that were designed to be specific for distal nephron Na channels prevented a full Na-conserving response (6). These observations may be reconciled by the finding that the number of conducting channels in the CNTs greatly exceeds that in the CCTs. This is based on a combination of a greater total length of tubule and a larger current per cell. Thus the CNT, perhaps together with the late DCT cells, probably plays a quantitatively more important role than does the CCT in reducing urinary Na under conditions of salt deprivation.
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-59659.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society