INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

ROMK IS A RECENTLY CLONED gene encoding inwardly rectifying, ATP-regulated K⁺ channels (11, 31). The mRNA encoding these channels is widely expressed in distal cortical and outer medullary nephron segments (2, 13).

Alternative splicing of ROMK exons yields several different transcripts ROMK1-3 that are differentially expressed along the nephron (2, 10, 13). ROMK2 is the most widely distributed isoform. ROMK1 is specifically expressed in collecting ducts. ROMK3 is expressed in the earlier nephron segments [medullary thick ascending limb (MTAL), cortical TAL (CTAL), and distal convolute tubule (DCT)]. ROMK2 is expressed in all the above-mentioned segments except the outer medullary collecting duct (OMCD). The distribution of ROMK channel isoforms is consistent with the possibility that ROMK may represent a major subunit of the low-conductance ATP-regulated secretory K⁺ channel in TAL and principal cells of the collecting duct (10).

Recent studies have suggested a predominant apical location using an ROMK-specific polyclonal antibody, consistent with a role for this channel in K⁺ secretion (30). In the kidney, the apical ATP-regulated secretory channels serve several important roles (10). In the thick ascending limb of Henle (both MTAL and CTAL) K⁺ channels provide a K⁺ efflux pathway for K⁺ entering via the basolateral Na⁺ pump or the apical Na-K-2Cl cotransporter (10), enabling apical K⁺ recycling. In the principal cells of the cortical collecting duct (CCD), K⁺ channels mediate K⁺ secretion into the urine and serve as the major mechanism for maintaining K⁺ balance.

K⁺ homeostasis is controlled mainly by aldosterone and K⁺ load (17, 19, 25, 26). Part of the response to K⁺ load may be mediated by changes in plasma aldosterone (21, 23). However, several studies suggest mineralocorticoid independent effects of K⁺ on K⁺ secretion (5, 12, 15, 22, 23). The present study was designated to explore the role of dietary K⁺ and aldosterone in the regulation of ROMK-mRNA in the rat kidney.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animal treatment and RNA isolation. Experiments were carried out using male 8- to 10-wk-old Wistar rats. The following six groups of rats were studied: 1) NK, control rats fed a normal chow (10 g K⁺/kg, 2.7 g Na⁺/kg); 2) LK, rats fed a K⁺-deficient diet (<0.01 g K⁺/kg 1.0 g Na⁺/kg; ICN, Cleveland, OH); 3) HK, rats fed a high-K⁺ diet prepared by adding KCl to normal chow (80 g K⁺/kg, 2.3 g Na⁺/kg); 4) A + NK, adrenalectomized (Adx) rats fed a normal chow like in group 1; 5) A + LK, Adx rats fed a K⁺-deficient diet like in group 2; 6) A + HK, Adx rats fed a high-K⁺ diet like in group 3.

Rats were kept on the different diets 12 days with free access to water.

Animals were killed by cervical dislocation, and kidneys were excised and dissected into cortex and medulla.

Total RNA was prepared from kidney slices using a Tri-Reagent kit (Molecular Research Center, Cincinnati, OH). Plasma electrolytes and aldosterone levels were determined in arterial blood samples drawn immediately after killing the animals. Serum concentrations of Na⁺ and K⁺ were determined by flame photometry, and aldosterone levels were measured using a radioimmunoassay kit (Coat-a-Count aldosterone; DPC, Los Angeles, CA).

Northern hybridization. Aliquots of 10-20 µg total RNA were resolved electrophoretically on 1% agarose gels under denaturing conditions (formamide/formaldehyde). Nucleic acids were transferred to nylon membrane (GeneScreen; New England Nuclear Research Products, Boston, MA) and cross linked by ultraviolet irradiation. Membrane strips were hybridized for 16-20 h with ³²P-labeled cDNA fragments corresponding to ROMK, ROMK2, and ₁-Na-K-ATPase and ₁-Na-K-ATPase under stringent conditions. The radioactive probe was prepared with a Rediprime DNA labeling kit (Amersham). Two ROMK probes were prepared by digesting a full-length clone of ROMK2 (accession number S69385) in pSport 1 (BRL Life Technologies). The first, a ~1.5-kb fragment obtained by digestion with BamH I/Pst I (nucleotide 1-1496), should hybridize with all ROMK isoforms. The second, a ~300-bp fragment obtained by digestion with Hind III (nucleotide 1577 to end), is specific to ROMK2 and ROMK2b. In addition, hybridizations were performed with the Pst I/EcoR I fragment of the ₁-subunit of Na-K-ATPase (nucleotide 3060-3636) and EcoR I fragment of ₁-subunit of Na-K-ATPase (nucleotide 343-1600). Membranes were washed and autoradiographed by standard procedures. Bound cDNA probes were removed by 1 to 2 min of boiling in 1× standard sodium citrate + 0.1% sodium dodecyl sulfate, and the same membranes were hybridized with a control probe synthesized from a cloned fragment of 18S ribosomal RNA. The abundance of this RNA species was independent of any of the treatments described in this study. Bindings were quantified by phosphorimaging (Fujix, BHS 1000) and expressed as the ratio of intensities obtained by hybridizing the same stripe with the cDNA studied and 18S cDNA, respectively. Each result was confirmed by repeating the Northern hybridization with at least two different RNA preparations and more than four animals. Data are expressed as means ± SE, and statistical significance was calculated using a two-tailed t-test.

Materials. [-³²P]dCTP (10 mCi/ml, 3,000 Ci/mmol) was purchased from Amersham Radiochemicals (Buckinghamshire, UK). Ultrapure agarose (molecular biology grade) was from IBI (New Haven, CT). All conventional chemicals were from Sigma Chemical (St. Louis, MO).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Effect of Adx and modulation of K⁺ intake on ROMK expression in kidney cortex. The effect of Adx and modulation of K⁺ intake on the abundance of ROMK mRNA in kidney cortex is depicted in Fig. 1 and Table 1. Figure 1A shows a representative Northern hybridization, and Fig. 1B shows the mean results pooled from four to eight rats. LK diet reduced ROMK mRNA expression to 47.1 ± 5.1% of normal (P < 0.001), whereas HK diet did not change it compared with normal. Adx reduced the message to 30.7 ± 6.8% of control (P < 0.001). LK diet fed to Adx rats further reduced ROMK expression to 16.0 ± 9.2%, whereas HK diet given to Adx rats increased it to 55.5 ± 2.1%, above the value observed in Adx rats on NK intake (P < 0.02).

View larger version (59K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Fig. 1.   Regulation of ROMK-mRNA by K+ intake and aldosterone. A: Northern hybridization of ROMK (top) and 18S control probe (bottom) with RNA from kidney cortex and medulla in intact rats and adrenalectomized (ADX) rats fed normal K+ (NK), low-K+ (LK), and high-K+ (HK) rations for 12 days. B: diet- and ADX-induced changes in ROMK mRNA in kidney cortex and medulla. Data are normalized to the amount of 18S cDNA and expressed as percentage of the abundance in intact rats fed normal chow (100%); n = 8 for intact rats, and n = 4 for ADX rats.

View this table: [in this window] [in a new window]   Table 1.   Relative mRNA expression of ROMK, ROMK2, and 1- and 1-Na-K-ATPase in cortex and medulla in the groups studied

Effect of Adx and modulation of K⁺ intake on ROMK expression in kidney medulla. The effect of Adx and modulation of K⁺ intake on the abundance of ROMK mRNA in kidney medulla is depicted in Fig. 1 and Table 1. Figure 1A shows a representative Northern hybridization, and Fig. 1B shows the mean results pooled from four to eight rats.

Effect of Adx and modulation of K⁺ intake on ₁- and ₁-subunits of Na-K-ATPase in kidney cortex and medulla. Since intracellular potassium ([K⁺]_i) is extruded via the ROMK channel and [K⁺]_i is determined primarily by the Na⁺ pump allowing transcellular K⁺ extrusion, we also studied the effect of Adx and modulation of K⁺ intake on the expression of the ₁- and ₁-subunits of Na-K-ATPase. Figure 2 and Table 1 illustrate the effect of Adx and modulation of K⁺ intake on the abundance of ₁- and ₁-Na-K-ATPase mRNA. Figure 2A shows a representative Northern hybridization with ₁-Na-K-ATPase, and Fig. 2B shows the measurements pooled from four rats in cortex and medulla for both ₁- and ₁-Na-K-ATPase.

View larger version (49K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Fig. 2.   Regulation of 1- and 1-Na-K-ATPase mRNA by K+ intake and aldosterone. A: Northern hybridization of -Na-K-ATPase (top) and 18S control probe (bottom) with RNA from kidney cortex and medulla in intact rats and ADX rats fed NK, LK, and HK rations for 12 days. B: diet- and ADX-induced changes and 1-Na-K-ATPase mRNA (solid bars) and 1-Na-K-ATPase mRNA (open bars) in kidney cortex and medulla. Data are normalized to the amount of 18S cDNA and expressed as percentage of the abundance in intact rats fed normal chow (100%); n = 4.

View larger version (15K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Fig. 3.   A: correlation between ROMK mRNA expression and serum K+ concentration in kidney medulla (R = 0.9406; n = 6, P < 0.001). B: correlation between ROMK mRNA and 1-Na-K-ATPase mRNA in kidney medulla (R = 0.9756; n = 6, P < 0.001).

Effect of Adx and modulation of K⁺ intake on ROMK2 expression in kidney cortex and medulla. For the Northern hybridizations mentioned above, ROMK was detected using a cDNA fragment common to the different ROMK isoforms. Thus the regulation of the specific isoforms by Adx and K⁺ intake modulation could not be defined. Therefore, we used also an ROMK2-specific cDNA fragment.

View larger version (67K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Fig. 4.   Regulation of ROMK2-mRNA by K+ intake and aldosterone. A: Northern hybridization of ROMK2 (top) and 18S control probe (bottom) with RNA from kidney medulla in intact rats and ADX rats fed NK, LK, and HK rations for 12 days. B: diet- and ADX-induced changes in ROMK2-mRNA in kidney cortex and medulla. Data are normalized to the amount of 18S cDNA and expressed as percentage of the abundance in intact rats fed normal chow (100%); n = 4.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The results of this study show that in the cortex ROMK expression is regulated in concert by aldosterone and potassium, whereas in the medulla ROMK expression is regulated by serum K⁺ levels irrespective of aldosterone. In the cortex, in intact rats LK decreased ROMK expression, whereas HK returned it to normal. Adx decreased ROMK expression to less than one-third of normal. LK in Adx rats further reduced ROMK expression, whereas HK in Adx rats increased it. In the cortex, ROMK is expressed in the distal segments CTAL, DCT, connecting tubule (CNT), and CCD but not in the proximal nephron. Among these, DCT, CNT, and mainly CCD are aldosterone-responsive segments (3, 8). The dramatic decrease in ROMK expression in Adx rats may therefore be attributed to these segments and mainly to the CCD. The residual expression of ROMK in the cortex of Adx rats may represent the CTAL ROMK, and presumably this portion is decreased by LK intake in Adx rats. Since the downregulation in cortical ROMK expression due to Adx occurred in face of severe hyperkalemia, it may be suggested that the decrease in cortical ROMK expression evident in adrenal-intact rats on LK diet stemmed mainly from decreased aldosterone.

In the medulla, both in intact and Adx rats, LK diet decreased ROMK expression, whereas HK increased it above normal. Adx induced a significant increase in ROMK mRNA expression. A major finding is the existence of a close correlation between the abundance of ROMK mRNA and plasma [K⁺] independent of adrenal function. It is suggested that the increase in ROMK in Adx rats accrued directly from the severe hyperkalemia. In the medulla, ROMK is expressed in the MTAL and in the OMCD. MTAL is relatively an aldosterone-nonresponsive segment (3, 8), whereas OMCD is responsive to aldosterone. Therefore it is assumed that the changes observed in ROMK expression with K⁺ intake modulation independently of adrenal function may be attributed mainly to the MTAL. This is further supported by our results with ROMK2. In the medulla, the ROMK2 isoform is located exclusively in the MTAL, and its expression changed similarly to the common core ROMK cDNA fragment in response to Adx and modulation of K⁺ intake. These observations emphasize the central role of serum [K⁺] in the regulation of ROMK expression in the MTAL. With regard to the responsiveness of the MTAL to aldosterone, it should be noted that conflicting results are reported in the literature. Studies by Stanton (20) and by Work and Jamison (29) show inhibition of MTAL NaCl transport in Adx rats and correction to normal by aldosterone. A study by Grossman and Hebert (9) shows increase of MTAL Na-K-ATPase by deoxycorticosterone acetate (DOCA) treatment; however, in the same study, Na⁺ deprivation that is associated with increased endogenous aldosterone failed to increase MTAL Na-K-ATPase, whereas it increased by 60% CCD Na-K-ATPase. In studies by Garg et al. (8) and Mujais et al. (14), mineralocorticoids administration did not induce any change in MTAL Na-K-ATPase and markedly increased CCD Na-K-ATPase. In the present study, Adx did not reduce medullary Na-K-ATPase expression but rather tended to increase it. Comparison of our expression results with experiments measuring Na-K-ATPase activity in medulla or MTAL of Adx animals does not support our results. However, these apparent discrepancies can be reconciled. Two studies show reduction in MTAL or medullary Na-K-ATPase in Adx animals, but this reduction could not be reversed by mineralocorticoids or glucocorticoid administration (6, 24) as in the cortex or CCD, but rather by NaCl (24). Taken together, these studies suggest that the CCD is the main target of aldosterone, whereas the MTAL Na-K-ATPase is much less dependent on mineralocorticoids. The effect of mineralocorticoids in the MTAL may be direct or indirect in nature, through its effect on sodium balance.

The ROMK channels mediate the downhill movement of K⁺ from the intracellular compartment. [K⁺]_i is determined mainly by the Na⁺ pump. Therefore we studied the effect of aldosterone and K⁺ intake modulation on the expression of the ₁- and ₁-subunits of Na-K-ATPase. The changes observed in the expression of ₁- and ₁-Na-K-ATPase were similar to those of ROMK. The expression of cortical ₁- and ₁-Na-K-ATPase was controlled synergistically by aldosterone and K⁺, whereas the expression of medullary ₁- and ₁-Na-K-ATPase correlated highly with serum potassium concentrations. In addition, ₁- and ₁-Na-K-ATPase expression correlated highly with that of ROMK expression both in cortex and medulla. Parallel observations on the protein level were recently reported by Anzai et al. (1) for Kir 6.1 in rat CCD. It was shown by Western blotting that CCD Kir 6.1 and ₁-Na-K-ATPase were 50% decreased by Adx and three- to fourfold increased by DOCA treatment. Taken together, these results show that the modulation of ROMK mRNA and protein by aldosterone and K⁺ reflects their effect on the Na⁺ pump. These observations suggest that the regulation of ROMK by aldosterone and K⁺ may be mediated by Na-K-ATPase and probably depends on [K⁺]_i.

As mentioned in the introduction, the apical ATP-regulated secretory channels serve different roles in the CCD and MTAL. In the CCD, apical K⁺ channels mediate K⁺ secretion into the urine, whereas in the MTAL K⁺ channels provide a K⁺ efflux pathway for K⁺ recycling via the apical Na-K-2Cl cotransporter (10). Our results support the idea that K⁺ secretion in CCD is regulated mutually by aldosterone and K⁺, whereas K⁺ recycling via the apical K⁺ channel in the MTAL depends on plasma K⁺ irrespective of aldosterone. Mineralocorticoids are well known to increase the apical K⁺ conductance of the CCD (18, 19, 27). These observations support our finding that Adx markedly decreased the expression of ROMK in the cortex. In addition, low-K⁺ diet that is associated with decreased aldosterone levels also decreased by 50% ROMK expression. Several studies demonstrated stimulation of CCD apical K⁺ conductance or potassium secretion by high plasma K⁺ independently of mineralocorticoids (15, 16, 22, 28). In some cases, the effect was not complete (22), or the presence of an intact adrenal gland exerted a permissive effect (16). These observations are in line with our finding that low K⁺ intake decreased ROMK expression even in Adx rats. However, high-K⁺ diet did not increase ROMK expression above normal in adrenal-intact rats despite increased aldosterone levels. These observations are in variance with those mentioned above (15, 16, 22, 28). One possibility to reconcile this discrepancy is that the increases observed in K⁺ conductance or K⁺ secretion due to high K⁺ intake stemmed from mobilization of channels from intracellular compartments to the cell membrane rather than increased transcription of new channels.

Inhibition of the cortical or CCD Na-K-ATPase activity in Adx animals is a well-documented phenomenon (3, 6, 24). Treatment of Adx animals with mineralocorticoids restored this activity (6, 24). Administration of DOCA to adrenal-intact rabbits increased Na-K-ATPase activity mainly in the CCD (8). Farman et al. (7), using in situ hybridization, demonstrated reduction in the ₁-subunit of Na-K-ATPase in the CCD and not in the proximal convoluted tubule in Adx rats. These results are in line with our present observations demonstrating more than 50% reduction in the expression of cortical ₁-Na-K-ATPase mRNA in Adx rats. The results of Farman et al. (7) also support our claim that the reduction in cortical ₁-Na-K-ATPase expression represents mainly the CCD ₁-Na-K-ATPase. The same observations on the protein level were reported by Anzai et al. (1) both for ₁-Na-K-ATPase and Kir 6.1 (1).

The Adx rats in this study were not supplemented with glucocorticoids or given NaCl in their drinking water. The reason for this is that previously we have shown (24) that the extreme changes in serum electrolytes observed in Adx rats, the effect of which we intended to study, were greatly corrected toward normal by these treatments (see table 2 in Ref. 24). Normalization in serum K⁺ in Adx rats was obtained by the LK diet (Table 2). This treatment decreased ROMK and Na-K-ATPase expression in most cases studied.

In summary, our present findings show that in the cortex ROMK expression is regulated mutually by aldosterone and K⁺, whereas in the medulla ROMK expression is regulated by serum K⁺ levels irrespective of aldosterone. It may be suggested that different mechanisms regulate K⁺ secretion in the collecting duct and K⁺ recycling in the MTAL. Further studies are required to substantiate this speculation, mainly by looking at the expression of these channels on the protein level by immunoblotting and measuring their activity under the different conditions studied here.