INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CORTICOSTEROID HORMONE-INDUCED factor (CHIF) is a 6.5-kDa transmembrane protein cloned as an epithelial-specific, aldosterone-induced transcript (2, 6, 18, 23). It is a member of a new gene family termed after the invariant motif FXYD (19). The family consists of seven single-span membrane proteins, four of which have been reported to be involved in the regulation or mediation of ion transport. They are FXYD 1 (phospholemman) (15), FXYD 2 (the -subunit of Na⁺-K⁺-ATPase) (12), FXYD 3 (Mat-8) (13), and FXYD 4 (CHIF) (2).

The following observations suggest that CHIF plays a role in renal and intestinal electrolyte homeostasis. First, both its mRNA and protein are specifically expressed in kidney collecting duct [cortical collecting duct < outer medullary collecting duct < inner medullary collecting duct (IMCD)] and in distal colon surface cells (top 20% of the crypt) (6, 18, 23). They cannot be detected in many other epithelial and nonepithelial tissues, including other segments of the kidney tubule and intestine, lung, stomach, uterus, mammary gland, salivary gland, heart, brain, muscle, liver, or skin. Second, CHIF appears to be independently upregulated by low-Na⁺ intake via changes in plasma aldosterone and by high-K⁺ intake, independently of aldosterone (6, 18, 23, 24).

The -subunit of the Na⁺-K⁺-ATPase (FXYD 2) is the best-studied member of the above family. This 65-amino acid type I membrane protein associates with the complex and is specifically expressed in the kidney (4, 12, 21). Functional effects of the -subunit on pump kinetics have been demonstrated by either coexpressing it with  and  in Xenopus laevis oocytes and transfected mammalian cells or neutralizing interactions in native kidney membranes using a specific anti- antibody (1, 4, 16, 21, 22). Recently, it has been demonstrated that CHIF also interacts with the Na⁺-K⁺-ATPase. First, it was found to be localized in the basolateral membrane and it coprecipitated with the -subunit under conditions that preserve active pump conformation (7, 18). Second, coexpressing CHIF with the -subunit in X. laevis oocytes increased the Na⁺ affinity of the pump and decreased its apparent K⁺ affinity, due to an increased competition by Na⁺ at external binding sites (3). A phospholemman-like protein, too, was reported to interact with the -subunit and to mediate its regulation by protein kinase C (PKC) in shark rectal gland (10). Thus it appears that, like , CHIF and other FXYD proteins may function to modulate pump kinetics in different tissues and maintain active Na⁺ and K⁺ pumping rates under varying conditions in the cell (ATP, Na⁺, K⁺, phosphorylation, etc).

Comprehensive analysis of the role of CHIF and other FXYD proteins in body electrolyte homeostasis requires suitable animal models. The present study describes the generation and phenotypic analysis of a targeted null mutation of the CHIF gene. The mutants show two abnormalities: 1) an increased urine volume under Na⁺ deprivation and K⁺ loading and 2) lethality after a combination of high-K⁺ intake and furosemide injection. These observations are consistent with a role of CHIF in Na⁺ and K⁺ transport that is specific to the outer and inner medullary duct.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Generation of a CHIF null mutant. Mice deficient in CHIF protein were generated by a targeted gene disruption. The CHIF gene was isolated by screening a mouse S129/SvJ phage genomic library with a rat cDNA probe. A ~20-kb clone was characterized by restriction mapping and partial sequencing. A 6.6-kb fragment that contains the whole transcribed region was fully sequenced and deposited in public databases under accession number AF362729. CHIF's coding sequence was disrupted by replacing a DNA fragment containing exons 3-5 (amino acids 1-33) with a neomycin (neo)-resistant cassette. The cassette was composed of neomycin phosphotransferase (accession no. U32991) flanked by 5'- and 3'-regulatory regions of mouse phosphoglycerate kinase (accession nos. X15339 and X15340, respectively). The insertion also eliminated a DraII site at position 4752, used for genotypic analysis. A herpes simplex virus thymidine kinase gene cassette was introduced in the 5'-end of the construct, resulting in homologous DNA arms of ~7 and ~1 kb.

Northern and Western hybridizations. Mice were killed by cervical dislocation, and distal colons and kidneys were excised and rinsed in ice-cold PBS. Kidneys were dissected into cortex, medulla, and papilla, and microsomal membranes were prepared as described before (8). Distal colon surface cells (colonocytes) were isolated by a modification of the procedure described elsewhere (17). In brief, colonic tubes were flushed three times with 10 ml of ice-cold PBS+2 mM dithiothreitol (DTT) and inverted (lumen out). The inverted colons were tied at one end, filled with DMEM+10% FCS, 2 mM DTT, and 2 mM EDTA and then tied at the other end as well. The filled colons were suspended in 25 ml of the above DMEM-EDTA medium and incubated at 37°C for 40 min under shaking at 140 rpm. Colonocytes were collected from the medium by centrifugation and stored at 70°C. Colonic and kidney total RNA was extracted as described before (23). Northern hybridization was carried out using a [³²P]-labeled cDNA probe prepared from a mouse CHIF expressed sequence tag (EST) clone (accession no. aa874059). For Western hybridization, a rabbit polyclonal antibody was raised against the synthetic peptide CKATPLIIPGSANT (amino acids 75-87 of the mouse protein) coupled to keyhole limpet hemocyanin through its NH₂-terminal cysteine. Kidney microsomal membranes and colonocyte whole cell lysates were resolved on a 10% tricine gel, blotted with the above anti-sera (1:200), and overlaid with horseradish peroxidase-coupled goat-anti rabbit antibody.

Animal treatment and metabolic cage studies. Monitoring of water uptake and urine excretion was done in metabolic cages. A typical experiment consisted of two matched groups of 12-18 +/+ and / mice weighing 30-40 g with an equal number of males and females. Mice were placed in metabolic cages (3 mice/cage) and acclimated for 10 days before the experiment. They were first fed a regular diet (6.8 g K⁺/kg, 2.3 g Na⁺/kg, 1.65 g Cl/kg), and the following measurements were performed daily: body weight, water and food intake, and urine volume. They were then fed either a high-KCl diet (94.1 g K⁺/kg, 2.3 g Na⁺/kg, 81.1 g Cl/kg) or an Na⁺-deficient diet (2.5 g K⁺/kg, <0.01 g Na⁺/kg, 1.57 g Cl/kg) and monitored for an additional 7-10 days. Experiments were conducted in conformity with the Guiding Principles for Research Involving Animals and Human Beings and were supervised by the Animal Welfare Committee of the Weizmann Institute. Urine and blood sera were analyzed for K⁺ and Na⁺ content, osmolality, and creatinine concentration. Electrolyte concentrations were measured by atomic absorption spectroscopy (Spectra-AA-50) and osmolarity with a vapor pressure osmometer (Wescor). Blood and urine creatinine levels were measured with a commercial kit (Sigma Diagnostics, St. Louis, MO) and used to calculate the glomerular filtration rates (GFR). Plasma aldosterone was determined using a radioimmunoassay kit (Coat-a-Count Aldosterone; DGP, Los Angeles, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CHIF gene structure and the generation of knockout mice. CHIF has been cloned and sequenced before only from rats. The mouse mRNA and protein sequences were deduced by assembling 26 EST entries and resequencing one of them, which was found to be identical to one reported previously (19). Figure 1A depicts the predicted amino acid sequence of mouse CHIF. It shares 85% identity with the previously cloned rat protein and has the same membrane topology. The corresponding cDNA has no stop codon upstream of the first AUG. Hence, it may in principle represent a longer protein than the one depicted in Fig. 1A. However, its homology to phospholemman and the -subunit of Na⁺-K⁺- ATPase, which were sequenced at the protein level (9, 15), makes this possibility unlikely.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   A: deduced amino acid sequence of mouse corticosteroid hormone-induced factor (CHIF). B: exon organization of the CHIF gene. The nucleotide position and amino acid sequence of each exon are listed in Table 1. C: structure of the targeting construct. tk, Thymidine kinase; neo, neomycin. D: Southern blot analysis of mouse CHIF. Genomic DNA was isolated from mouse tail and digested with EcoRI, SacI, PstI, and BamHI. Samples of cut and uncut DNA were resolved on 1% agarose and blotted with a probe corresponding to nucleotides 5073-6552 (exons 6-9).

View larger version (35K): [in this window] [in a new window]   Fig. 2.   A: Southern hybridization of DraII-digested genomic DNA from CHIF (+/) × (+/) offspring. The predicted sizes of the wild-type and neo-disrupted alleles are 1.5 and 3.2 kb, respectively. B: Northern blot hybridization of distal colon total RNA from low-Na+-fed mice with CHIF cDNA. C: Western blot hybridization of colonocytes and kidney medulla membranes with anti-CHIF antibody. Matched groups of mice were fed normal (C), high-K+ (HK), and low-Na+ (LN) diets for 2 wk. For each treatment, colonocytes were collected from 3 different mice and blotted with the anti-CHIF antibody. Arrows, positions of CHIF.

Phenotypic analysis of knockout mice. CHIF / mice were viable and fertile and could not be distinguished from +/+ or +/ littermates by visual inspection. Under normal conditions, they showed slightly higher water and food intake, but nevertheless were ~15% smaller than matched +/+ mice (Table 2). Because CHIF is primarily expressed in kidney and is regulated by salt intake, we have looked for phenotypes associated with kidney function under normal and electrolyte stress conditions. Figures 3 and 4 summarize an experiment in which matched groups of 12 +/+ and 12 / mice were studied, first under normal conditions and then during K⁺ loading. The high-KCl diet increased their K⁺ intake by ~13-fold, with no change in Na⁺ intake (Table 3). Under normal conditions, either no or small differences were observed between the two groups. K⁺ loading resulted in substantial differences in two parameters. The first was the volume of urine excreted, and the second was the GFR. The high-K⁺ diet increased urine output in both wild-type and knockout mice, but urine volume in / mice was up to twofold higher than in the matched +/+ group (Fig. 3). The higher urine excretion was matched by an increased water intake so that the ratio of water intake to excretion was not significantly different in +/+ and / mice (Fig. 4A). The fractional Na⁺ and K⁺ excreted in the urine were similar as well. Because GFR changed significantly during high-K⁺ treatment, we have also calculated the rates of water reabsorption and Na⁺ and K⁺ excretion as a fraction of GFR. As above, no significant difference has been noted between +/+ and / mice (Table 4). Also, no significant differences between +/+ and / mice were observed in plasma Na⁺ and K⁺ concentrations, aldosterone levels, and osmolality (Fig. 4B, Table 3).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Matched groups of +/+ and / mice were monitored in metabolic cages. They were examined for 5 days under normal conditions and for another 10 days under K+ loading. Urine volumes (A) and glomerular filtration rates (B) (in ml · 100 g body wt1 · h1) are plotted as a function of time. Values are means ± SE of 12 mice (6 males, 6 females).

View larger version (41K): [in this window] [in a new window]   Fig. 4.   A: H2O, K+, and Na+ excreted in the urine are expressed as a fraction of their intake. Values are means ± SD of measurements over a period of 5 days. B: plasma concentrations of Na+ and K+ and osmolality in +/+ and / mice under normal (C) and high-K+ rations (HK). Values are means ± SE of 12 mice under each condition.

1

1

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Matched groups of +/+ and / mice were monitored for 5 days under normal conditions and for another 10 days under Na+ deprivation. Urine volumes and glomerular filtration rates are plotted as function of time. Values are means ± SE of 12 mice (6 males, 6 females).

View larger version (44K): [in this window] [in a new window]   Fig. 6.   A: H2O and K+ excreted in the urine are expressed as the fraction of their intake. Na+ excretion is expressed in absolute units because Na+ intake under a low-Na+ diet (LNa) is virtually zero. Values are means ± SD of measurements in 12 mice over a period of 5 days. B: plasma concentrations of Na+, K+, and osmolality in +/+ and / mice under normal (C) and low-Na+ rations. Values are means ± SE of 12 mice under each condition.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Matched groups of 18 +/+ and / mice were first monitored under normal conditions and then treated as follows: received no food and had free access to water that contained 6% sucrose (A) or received normal food and furosemide in their drinking water (1.5 mg/ml; B).

View larger version (10K): [in this window] [in a new window]   Fig. 8.   Matched groups each of 39 mice were fed high-K+ diets and received daily injections of furosemide (20 mg/kg). The no. of surviving mice is plotted as function of the treatment time.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study describes the generation and phenotypic analysis of a targeted mutation of the CHIF gene. Sequencing of the mouse CHIF gene revealed the existence of nine exons, two of which transcribe a 5'-untranslated region. This is quite unexpected for a very small mRNA (0.55 kb) and protein (88 amino acids). However, a similar multiexonic structure has been reported for two other members of the FXYD family, phospholemman and the -subunit of Na⁺-K⁺-ATPase (5, 11, 20). The obvious advantage of such genomic organization would be to enable expression of alternatively spliced products. Two -splice variants, with different extracellular sequences, and localization along the nephron have indeed been reported (9, 16, 19). Phospholemman also appears to have several splice variants (5), which differ only in the 5'-untranslated region, and their expression could be differentially regulated. So far, alternatively spliced forms of CHIF have not been detected experimentally. EST entries that deviate from the consensus mRNA sequence have been deposited in public databases, but they do not appear to represent alternatively spliced forms of the above gene.

Deletion of the CHIF gene was done by standard methods and confirmed by Southern, Northern, and Western hybridizations. The homozygous mutants appeared normal in most parameters measured. The major phenotype observed is a larger volume of urine excretion (and water intake) under conditions of K⁺ loading or Na⁺ deprivation. Such an abnormality is consistent with the high abundance of CHIF in the IMCD, a major site in regulating water homeostasis. The fact that the phenotype is apparent during electrolyte stress but not when diuresis is evoked by excessive water intake suggests a secondary response mediated by alterations in Na⁺ and K⁺ transport. It has been recently demonstrated that CHIF interacts with the Na⁺-K⁺-ATPase and increases its affinity to cell Na⁺ while decreasing its apparent K⁺ affinity (3, 7). The last effect is due to an increased competition by Na⁺ at external K⁺ binding sites. One may therefore predict that the deletion of CHIF will affect pump kinetics and partly inhibit collecting duct K⁺ excretion and/or Na⁺ absorption. This may not directly affect electrolyte homeostasis because inhibition is only partial and probably well compensated for by the increased GFR and elevated transport rates in the more proximal nephron segments. The observed increase in water excretion may therefore reflect either the elevated GFR under K⁺ loading or a secondary response of the IMCD to the inhibited Na⁺/K⁺ transport. However, a defect in K⁺ secretion could be noted when K⁺ loading was combined with the inhibition of NaCl absorption by furosemide. Under these conditions, 44% of the / mice died within 10 days, and the rest were hyperkalemic relative to the +/+ mice. While this phenomenon demonstrates an essential role of CHIF under these stress conditions, its molecular mechanism is not obvious. In principle, it may reflect either an excessive volume depletion or inhibition of K⁺ excretion, which becomes lethal in the / mice. Plasma K⁺ and osmolarity measurements summarized in Table 5 indicate that the second possibility is the more likely one.