INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE CORTICAL COLLECTING TUBULE (CCT) of the mammalian kidney helps to regulate the K concentration in plasma by secreting or reabsorbing appropriate amounts of K from the urine (4). When dietary K is high, the CCT secretes increased amounts of K. When dietary K is low, net secretion can stop and even reverse, such that K is absorbed by this segment (4).

K secretion by the CCT is mediated by a conductive pathway in the apical membrane (13, 14). The major K channel in the apical membrane of principal cells has a low conductance with the property of inward rectification and a high open probability independent of membrane voltage (5, 15). These "SK" (small conductance or secretory K) channels are thought to be primarily responsible for K secretion by the CCT (14). Indeed, when rats are maintained on a high-K diet for 1-2 wk, such that K secretion is enhanced, there is a three- to fourfold increase in the abundance of these channels, as measured by the patch-clamp technique (10, 15).

The mechanisms by which the channel density is increased have not been elucidated. One possibility is that transcription of the gene coding for the SK channels is activated, leading to an increased abundance of mRNA coding for the channel and hence an increase in the number of K channel proteins in the membrane. The cloning of the ROMK family of K channels, which is believed to correspond to the SK channels in the CCT, as well as in the thick ascending limb of the loop of Henle (1, 6, 16), has enabled this idea to be examined. In this study, we have used the technique of in situ hybridization to directly test the hypothesis of transcriptional control of K channels through dietary K.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Biological preparation. Sprague-Dawley rats of either sex (100-150 g), raised free of viral infections (Charles River Laboratories, Kingston, NY), were fed either a normal rat chow (Na content, 5 g/kg; K content, 12 g/kg) or a matched high-K diet (K content, 100 g/kg). Both diets were obtained from Harlan-Teklad (Madison, WI). In one series of experiments, rats maintained on normal chow were first hydrated for 3 days, using 0.6 M sucrose in their drinking water. Half this group of rats was then deprived of drinking water for 30 h, and the rest continued on sucrose water. Animals were killed by cervical dislocation, and the kidneys were removed and sliced.

In situ hybridization. Single CCTs were dissected using N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid-buffered Ringer solutions prepared with diethyl pyrocarbonate-treated water to inhibit ribonuclease (RNase) activity. The tubules were transferred with a small volume of fluid to a spot of dried Cell-Tak on a glass microscope slide and attached with gentle pressure on the ends. The hybridization protocol has been described (3), except for slight modifications described below. Briefly, the tubules were then fixed for 10 min in 4% paraformaldehyde in phosphate-buffered saline (PBS) and subsequently washed three times with PBS. Next, the tubules were treated with acetic anhydride, dehydrated with an ethanol series, and dried in air. Antisense and sense RNA probes for ROMK2 or for aquaporin 3 (AQP3) were prepared with the Maxiscript kit (Ambion) using [³³P]UTP (Amersham) at a concentration of 2.5 µCi/µl and a specific activity of 2.5 × 10⁴ Ci/mol. For the ROMK probes, the sequence between nucleotides 86 and 1608 of ROMK2 (16) was recloned into the pSPORT vector. Antisense and sense probes were produced using the SP6 and T7 promoters, respectively. The AQP probes spanned the sequence from 73 to +1028 of the cDNA clone (3). The probes were purified by chromatography on a spun column of Sephadex G-50. tRNA was added as a carrier, and the mixture was precipitated with cold ethanol. The probes were hydrolyzed to give fragments of 100-150 nucleotides. The hybridization solution contained ~10⁷ disintegrations · min¹ · ml¹. After hybridization for 3-18 h at 55°C, the slides were rinsed three times with 4× standard sodium citrate (SSC), treated with RNase A, and rinsed at increasing stringency from 2× SSC to 0.1× SSC at 55°C. After dehydration with an ethanol series and drying in air, the slides were put in contact with Kodak BioMax MR film for 1-8 days for autoradiography.

Quantitation of autoradiography. Films were developed and viewed under an inverted microscope (Nikon Diaphot) equipped with a charge-coupled device video camera (Vidichip; Morell Instruments, Melville, NY), connected to a 24STV video-capture board (RasterOps, Santa Clara, CA). The digitized images were analyzed using the program NIH Image version 1.52 on a MacIntosh computer. The signal from the tubule was measured by constructing a polygonal template to match the outline of the tubule as closely as possible. The optical density was computed within this polygon. Calibration studies in liquid droplets confirmed that the the autoradiograph density was proportional to the radioactivity of the sample being imaged (see Fig. 4). Similar results were obtained by measuring the densities within several rectangular templates constructed within the boundaries of the tubule. Background levels were measured within rectangular templates constructed in the same image in areas adjacent to the tubule.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Detection of ROMK message in isolated CCT. The use of in situ hybridization to detect ROMK message in the isolated CCT was reported previously (8). We have adapted this technique to assess semiquantitatively the relative amounts of specific mRNA species in this segment. The basis for the method is shown in Fig. 1. Two tubules isolated from the same rat were fixed and treated identically for in situ hybridization. One was hybridized with an antisense probe obtained from the ROMK2 clone. The other was hybridized with a probe based on the same sequence but transcribed in the sense direction. Transmitted light images of the tubules are also shown. The tubule hybridized with the antisense probe showed strong radiolabeling. The signal was readily detectable after 24 h of exposure of the film to the sample. No labeling could be detected from the tubule hybridized with the sense probe. A faint signal from the sense probe could be detected only in some samples exposed for 8 days or longer. Thus the hybridization is quite specific.

View larger version (59K): [in this window] [in a new window]   Fig. 1.   In situ hybridization of isolated rat cortical collecting tubule (CCT) with ROMK probes. Left: light micrographs of 2 tubules from same animal. Right: autoradiograms of same tubules after hybridization with either an antisense or a sense probe, based on the ROMK sequence. Exposure time of film was 48 h.

To optimize the conditions for detecting the signal, we examined the effects of varying both the concentration of the probe and the time allotted for hybridization. Figure 2 shows the hybridization signal measured in tubules matched from the same rats at three different probe concentrations prepared from the same batch. The absolute concentration of the probe was not measured. There was some dependence on concentration over this range, but, at the higher concentrations, the signal appeared to saturate. Figure 3 shows the signal in another set of matched tubules incubated with the probe for three different time periods. A significant positive signal was obtained with as little as a 3-h incubation. The signal after 18 h of hybridization was somewhat stronger, and this overnight period was convenient to use. The tissue tended to deteriorate with significantly longer incubation times. Thus the hybridization at 18 h was presumed to be close to completion and was used for subsequent experiments.

View larger version (10K): [in this window] [in a new window]   Fig. 2.   Effect of probe concentration. Three different probe concentrations, corresponding to 4.0, 8.6, and 16.7 × 106 disintegrations · min1 · ml1, respectively, of 33P were used to label matched sets of tubules from the same rats. Specimens were processed identically in all other respects and at the same time. Autoradiographic signal was corrected for background and plotted in arbitrary units. Data are means ± SE for 9-12 tubules from 4 different animals.

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Effect of hybridization time. Matched sets of tubules from same rats were hybridized with same concentration of probe (8.6 × 106 disintegrations · min1 · ml1) for 3, 7, or 18 h. Specimens were processed identically and simultaneously after hybridization. Autoradiographic signal was corrected for background and plotted in arbitrary units. Data are means ± SE for 7-9 tubules from 3 different animals.

Two types of calibration curves were made to ensure that the method could detect differences in the amount of labeled probe present in the tubules. First, the densitometric signal was measured from filters in which known amounts of the probe were spotted. The results are shown in Fig. 4. The signal was a linear function of density over a fairly large range of probe concentration. This range includes that over which the ROMK mRNA signals were measured (see below). Ideally, we would like to have performed this calibration using a geometry for the sample similar to that of the isolated tubule. This was not practical. However, we reasoned that an increase in the time of exposure of the film to the sample should give results equivalent to those of increasing the concentration of the probe. We therefore examined the densitometric signal in films exposed for various lengths of time to tubules hybridized with antisense probe. As shown in Fig. 5, the signal was a linear function of exposure time over the range of measurement. These two calibration curves increased our confidence that we could detect differences in mRNA abundance in the isolated tubules by the in situ hybridization technique.

View larger version (77K): [in this window] [in a new window]   Fig. 4.   Calibration of autoradiographic signal vs. 33P density. A: fluid samples of equal volume (2 µl) containing dilutions of the 33P probe were spotted onto a nitrocellulose membrane, dried, and used to expose film for 4 h. Density of different areas of autoradiogram were computed and corrected for background as indicated by squares in right panel. B: corrected densities are plotted as a function of the 33P counts/min (cpm) measured with a beta-emission scintillation counter. Line represents a linear regression fit to data.

View larger version (47K): [in this window] [in a new window]   Fig. 5.   Calibration of autoradiographic signal vs. exposure time. A: CCT such as that shown in Fig. 1 was used to expose X-ray film for various time periods shown. Density of signal was measured using same template, as indicated at far right. B: signal, after correction for background, is plotted as a function of exposure time. Line represents a linear regression fit to data.

Changes in the level of AQP3 message. As a positive control for the technique, we measured the relative abundance of mRNA coding for the AQP3 protein (3) in the CCT. AQP3 is a water channel in the major intrinsic protein family, which is expressed primarily in the basolateral membrane of the CCT in the rat kidney (2). Message levels for AQP3 measured by Northern blot analysis were previously shown to increase by about twofold after thirsting the animals for 48 h (7). This appears to be an appropriate positive control for the ROMK detection experiments described below, since the messages are expressed in the same segment and the degree of regulation is similar in the two cases.

CCTs were dissected from rats thirsted by water deprivation for 30 h, as well as from well-hydrated control animals. Control and experimental tubules were placed on the same slides to ensure identical handling in terms of hybridization and washing conditions. An example of one experiment is shown in Fig. 6. The stronger autoradiographic signal is readily apparent from the autoradiographs. Quantitative analysis was carried out as described for the ROMK signal. In three experiments in which three or four tubules from control and experimental animals were analyzed, increases in the apparent abundance of AQP3 message were 1.7-, 1.8-, and 2.2-fold. Thus the in situ hybridization technique is able to detect modest changes in mRNA abundance.

View larger version (197K): [in this window] [in a new window]   Fig. 6.   Effect of water deprivation on aquaporin-3 (AQP3) mRNA. CCTs were isolated and processed as in Fig. 1, except that hybridization was to an antisense probe based on the sequence of AQP3. Top: 3 tubules from a control rat. Bottom: 3 tubules from a rat deprived of water for 30 h. Average densities (±SE) of the tubules (in arbitrary units) were 62 ± 9 for controls and 111 ± 6 for thirsted animals.

ROMK message levels are not changed by a high-K diet. To test whether modulation of ROMK expression could be observed at the mRNA level, we isolated CCTs from rats fed a high-K diet (10% KCl) or a matched normal K diet (1.1% KCl) for 10-14 days. This protocol was chosen because it led to an increase in the density of conducting SK channels in the apical membrane of the principal cells by at least threefold (10). Typical autoradiographs are shown in Fig. 7. In contrast to the case of the AQP3 message, no difference in the signal densities could be discerned by eye. This was confirmed by densitometric analysis. In seven experiments, the change in apparent mRNA abundance was 1.10 ± 0.06-fold. We conclude that increased message abundance accounts for little or none of the stimulation of SK channels by a high-K diet (Fig. 8).

View larger version (191K): [in this window] [in a new window]   Fig. 7.   Effect of high-K diet on ROMK mRNA. CCTs were isolated and processed as in Fig. 1. Top: 3 tubules from a control rat. Bottom: 3 tubules from a rat on a high-K diet for 14 days. Average densities (±SE) of the tubules (in arbitrary units) were 78 ± 8 for controls and 79 ± 4 for K-loaded animals.

View larger version (29K): [in this window] [in a new window]   Fig. 8.   Effects of high-K (hatched bars) diet on small conductance channel density and ROMK mRNA. Conducting channel density per patch is from Ref. 10. Relative mRNA abundance is from 7 experiments similar to those in Fig. 7. Values are arbitrarily scaled to match control levels for the 2 parameters. Open bars, control.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Quantitation of mRNA levels by in situ hybridization. Because the kidney is a heterogeneous organ, it is often difficult to interpret measurements of absolute levels of mRNA or of changes in these levels in extracts of whole kidney, cortex, or medulla. This is a particular problem when messages are expressed at different levels or with a different regulatory pattern in different nephron segments. One method that has been used to detect specific messages in individual nephron segments is reverse transcription-polymerase chain reaction (RT-PCR) (9). With this technique, specific mRNA species in isolated tubules are reverse transcribed to cDNA, which is then amplified by PCR. It is possible to quantitate the signal through the use of competitive inhibition with known amounts of cDNA (12).

A major advantage of the RT-PCR method is the high sensitivity, allowing detection of very small amounts of message. There are, however, at least three drawbacks to this approach. First, controls must be run to ensure that the signal from mRNA, rather than genomic DNA, is being detected. Second, the technique makes use of enzyme preparations (reverse transcriptase, Taq polymerase) that may vary in their activities. Third, quantitation of the signal, while possible, is difficult because of the large amplification through multiple PCR cycles. Although these difficulties can all be overcome by appropriate controls and competition studies, this can significantly raise the number of tubules that must be isolated for each experiment.

The in situ hybridization technique offers an alternative that avoids these difficulties. Hybridization signals are obtained directly from short tubular segments, without the use of enzyme reactions or signal amplification. Two potential disadvantages of the method need to be discussed. First, the sensitivity of the detection system is at least theoretically lower than that of RT-PCR, since the latter can, in principal, detect a single RNA molecule. We have not explored the lower detection limit of the in situ system. The messages for ROMK and for AQP3 were both detected rather easily. Much less abundant species could be observed by increasing the specific activity of the probes or increasing the length of exposure of the autoradiogram. The second problem is the quantitation of the signal, as in situ hybridization has not traditionally been used for this purpose. However, we have shown here that conditions can be employed under which the signal is linear with respect to the amount of radioactivity in the sample and the length of exposure of the film. We have also shown that the method is sufficiently sensitive to detect modest changes in the abundance of mRNA for AQP3.

The in situ technique described here offers an alternative to the RT-PCR method of detecting specific messages in single tubules. At least for messages of adequate abundance, it provides a simpler and more direct approach to the problem.

Upregulation of K channel by a high-K diet. Two laboratories have shown that the density of low-conductance K (SK) channels in the apical membrane of the rat CCT increases when the animals are fed a high-K diet for 10 days or more. Wang and colleagues (15) found that the number of apical membrane patches containing SK channels doubled. In a subsequent study, our laboratory reported a threefold increase in the mean number of conducting channels after K adaptation (10). The fact that these are long-term changes and that they are "remembered" by the tubules after isolation from the kidney into a defined environment suggests that the effect could be the result of activation of the gene encoding the channel, which would lead to higher levels of both mRNA and protein. However, in the studies reported here, we did not detect a change in the level of ROMK mRNA.

One explanation for this result might be that the SK channels are not the product of the ROMK gene. This seems highly unlikely, given the similarities in the properties of the SK channels in the native tubule and the ROMK channels expressed in Xenopus oocytes. These similarities include biophysical properties, such as single-channel conductance, kinetics of channel gating, and ion selectivity (11), and regulation by factors such as intracellular pH, ATP, and protein kinase A (13).

If the SK channels are encoded by the ROMK gene, then the lack of increase in ROMK message narrows the possible mechanisms that might underlie the increase in conducting channel density observed in animals fed a high-K diet. It is possible that the amount of ROMK protein increases, despite a lack of change of mRNA. This could be the result of a decreased rate of protein degradation or an increased efficiency in the rate of protein synthesis. Alternatively, there could be a translocation of channel proteins from an intracellular storage site to the apical plasma membrane. Finally, proteins already resident in the membrane in a nonconducting state could be activated. So far, we have no evidence to support or rule out any of these possibilities.