点击显示 收起
【关键词】 mouse
1INSERM U652, IFR58, Institut des Cordeliers, Universite Rene Descartes, 2INSERM U665, Institut National de la Transfusion Sanguine, 3UMR 7134 CNRS-Universite Pierre et Marie Curie, 4Departement de Physiologie and 6Service de Biochimie, Hpital Europeen Georges Pompidou, AP-HP, 5INSERM U439, Hpital Lariboisiere
7Departement de Physiologie, Hpital Necker-Enfant Malades, AP-HP, Paris, France
ABSTRACT
NH4+ transport by the distal nephron and NH4+ detoxification by the liver are critical for achieving regulation of acid-base balance and to avoid hyperammonemic hepatic encephalopathy, respectively. Therefore, it has been proposed that rhesus type B glycoprotein (Rhbg), a member of the Mep/Amt/Rh NH3 channel superfamily, may be involved in some forms of distal tubular acidosis and congenital hyperammonemia. We have tested this hypothesis by inactivating the RHbg gene in the mouse by insertional mutagenesis. Histochemical studies analyses confirmed that RHbg knockout (KO) mice did not express Rhbg protein. Under basal conditions, the KO mice did not exhibit encephalopathy and survived well. They did not exhibit hallmarks of distal tubular acidosis because neither acid-base status, serum potassium concentration, nor bone mineral density was altered by RHbg disruption. They did not have hyperammonemia or disturbed hepatic NH3 metabolism. Moreover, the KO mice adapted to a chronic acid-loading challenge by increasing urinary NH4+ excretion as well as their wild-type controls. Finally, transepithelial NH3 diffusive permeability, or NH3 and NH4+ entry across the basolateral membrane of cortical collecting duct cells, measured by in vitro microperfusion of collecting duct from KO and wild-type mice, was identical with no apparent effect of the absence of Rhbg protein. We conclude that Rhbg is not a critical determinant of NH4+ excretion by the kidney and of NH4+ detoxification by the liver in vivo.
rhesus protein; acid-base; tubular acidosis; ammonia
RHESUS (RH) TYPE B GLYCOPROTEIN (RhBG/Rhbg) belongs to a protein superfamily (Mep/Amt/Rh) present in all groups of living things and is believed to represent a new family of NH4+/NH3 transporters highly conserved through evolution from archebacteria to mammals (13). This was initially proposed on the basis of several observations. All Rh proteins share significant sequence homologies with specific NH4+ transporters from the Mep/Amt superfamily (26). RhCG and RhAG, two RhBG orthologs, were able to restore growth when transfected into yeast made triple-deficient for their endogenous NH4+ transporters MEP1, MEP2, and MEP3 (25).
In aqueous solution, total ammonia exists in two forms: the cation NH4+ and the gas NH3. NH3 can reversibly react with a proton to form NH4+. The pKa of this buffer reaction is 9.03 at physiological pH and temperature of plasma (3). Consequently, NH4+ represents the predominant form of total ammonia at physiological pH. Before identification of Rh glycoproteins as putative NH3/NH4+ transporters, it was assumed that NH3 gas freely crosses cell membranes by nonionic diffusion, whereas the NH4+ ion is impermeant and must share membrane transport proteins with other cations, in particular K+, to cross cell membranes (11). Several groups have attempted to assess the mechanism of total ammonia (i.e., NH3 + NH4+) transport through Rh glycoproteins by heterologous expression in Xenopus laevis oocytes to determine whether these proteins mediate NH3 or NH4+ transport (2, 24, 28, 42). Although all these studies concluded that Rh glycoproteins increase total ammonia transport, they differed in their conclusions with respect to the nature of the transported species, as well as to the mechanism of transport. Some reported the electrogenic transport of NH4+ down its electrochemical gradient in Rhbg- or RhCG-expressing oocytes (2, 28), others the exchange of NH4+ against a proton in an NH4+/H+ countertransport in RhAG- or RhBG-expressing oocytes (24, 42), and Bakouh et al. (2), in addition to an electrogenic NH4+ transport pathway, also suggested that NH3 could be directly transported by RhCG. More recently, two groups reported the crystal structure of the AmtB protein, an Rh glycoprotein paralog from Escherichia coli, and proposed a model in which proteins from the Mep/Amt/Rh superfamily are NH3 channels (16, 47). Stopped-flow analyses to study the total ammonia transport in red blood cells from human and mouse genetic variants with various defects of proteins, that comprise or interact with the Rh complex, demonstrated at the same time that RhAG/Rhag facilitates NH3 transfer across the membranes (30), in agreement with the AmtB model. More recently, using the same approach with Madin-Darby canine kidney and human embryonic kidney (HEK293) cells transfected with RhBG and RhCG, Ripoche and co-workers (48) also confirmed that these proteins accelerate NH3 flux across cell membranes.
Despite compelling evidence that Rh glycoproteins can transport NH3, the physiological functions of these proteins in mammals remain largely unknown. We and others have previously reported that RhBG/Rhbg and RhCG/Rhcg are expressed in the kidney in the collecting duct (7, 29, 37), a nephron segment where total ammonia transport occurs in the NH3 form in parallel to H+ secretion (9, 10, 18). Rhbg was also found to be expressed in the liver in the plasmic membrane of the perivenous hepatocytes, which take up NH3 to be incorporated into glutamine (41). Therefore, it has been anticipated that Rhbg in the kidney may serve to accelerate NH3 transport because nonionic diffusion of NH3 through the lipid bilayer may possibly be too slow to support the high rate of NH3 transport required by this nephron segment (17). RhBG/Rhbg could also provide a molecular target for regulating NH3 transport through collecting duct epithelium and might therefore be critical for renal acid-base regulation (17, 29). In the liver, because perivenous hepatocytes represent a high-affinity, low-capacity NH4+ clearance pathway, in contrast to midzonal and periportal hepatocytes, which convert NH4+ into urea by a high-capacity, low-affinity pathway, it has been proposed that Rhbg-mediated NH3 transport may be critical to maintaining plasma NH4+ concentration in the micromolar range to avoid hepatic encephalopathy (17, 41).
Therefore, to gain insight into Rhbg physiological function, we generated mice with targeted gene disruption of RHbg and examined whether the mice displayed hallmarks of distal tubular acidosis or disturbed hepatic glutamine and nitrogen metabolism.
METHODS
Gene targeting and genotyping. The mouse RHbg gene was disrupted by insertional mutagenesis as previously described by Zheng et al. (46). An insertional vector was isolated from the mouse 129S5/SvEvBrd genomic 3'hprt library (46), using a 100-bp probe from exon 4 of the RHbg gene. This vector, RHbg 2.3, contains an 8.3-kb genomic fragment of RHbg and includes exons 2, 3, 4, and 5, as determined by restriction mapping and PCR analysis. A second EcoRI site was introduced by targeted mutagenesis (Stratagene QuikChange XL site-directed Mutagenesis kit), allowing the deletion of a 1.7-kb fragment of an intronic region adjacent to exon 2, and religated to produce the "gapped" targeting vector RHbg 2.3EcoRI.
The EcoRI linearized vector RHbg 2.3EcoRI was electroporated into 129/Ola embryonic stem (ES) cells. ES cell growth, electroporation (800 V, 0.3 μF, 80 x 106 cells for 100 μg of linearized targeting vector), and puromycin selection were performed at genOway (Lyon, France) according to standard procedures. PCR was used to screen for targeted ES cell clones. For the 3'-end 2.9-kb product, the forward primer 5'-ATG GCT TCT GAG GCG GAA AGA ACC AG-3' (from the targeting vector) and reverse primer 5'-GAG GCT GTT ATG CCT GGG GAA TTC TA-3' (from the deleted 1.7-kb fragment) were used. For the 5'-end 3.7-kb fragment, primers were forward: 5'-GCC TTC TTA CTC TCC AGA CTC CCT CCT T-3' (from the deleted 1.7-kb fragment) and reverse: 5'-GAT CGG TGC GGG CCT CTT CGC TAT TAC-3' (from the targeting vector; see blue arrows in Fig. 1B). Insertional targeting was confirmed by Southern blot analysis with BglII using a 513-bp PCR fragment from the gapped region as a probe (primers were forward: 5'-ACT GAG GAA CAC CAC CCC AC-3' and reverse: 5'-GCC AAG GCT TTA CCC ACA AA-3'). In targeted clones, an 18.3-kb BglII fragment was present in addition to the wild-type 7.7-kb BglII fragment.
Blastocyst (C57BL/6) injection and embryo transfer were performed at genOway. Male chimeras were mated to C57BL/6 females. Genotyping of progeny was performed by PCR and Southern blot analysis of tail genomic DNA as described above. Targeting was confirmed by a second Southern blot analysis using the probe to the gapped region: XhoI digestion generated a 9.7-kb fragment in the wild-type (WT) allele and two fragments, 13.7 and 9.1 kb, respectively, in the targeted allele (Fig. 1, B and C). One ES cell clone with the RHbg 2.3EcoRI insertion was transmitted through the germ line, generating heterozygous animals. Heterozygous mating pairs were crossed to obtain homozygous RHbg/ mutants and WT littermates on a hybrid 129/Ola-C57BL/6 background. Genotyping was systematically carried out by Southern blot analyses of tail DNA after XhoI digest as described above.
Immunohistochemistry. Kidneys or livers from RHbg/ mice and from their respective WT control littermates were fixed in vivo by perfusion with 4% paraformaldehyde, embedded in paraffin, and used for different staining procedures as described previously (7, 29). A rabbit anti-Rhbg (29), a rabbit anti-Rhcg (29), a rabbit anti-56-kDa 1-subunit of the H+-ATPase (a kind gift from Sylvie Breton and Dennis Brown, Harvard Medical School, Boston, MA) (5), or an antibody recognizing AE1 (a kind gift from Seth L. Alper, Harvard Medical School) (4) was used for these studies. All these antibodies have been used and characterized previously in mouse.
Physiological parameter measurements. All experiments were performed using age- and sex-matched littermates. All animals were treated in compliance with French and European Union animal care guidelines. For the determination of basal-state physiological parameters, adult RHbg+/+ (WT) and RHbg/ mice were housed in metabolic cages. Mice were allowed to adapt to these cages for 7 days, and then 24-h urine samples were collected with light mineral oil in the urine collector to determine daily urinary electrolyte excretion. The mice were given deionized water ad libitum and fed with standard laboratory chow (M20, Dietex). The same procedure was used for the acid-loading experiments, except that mice were given 0.28 M NH4Cl in place of water. Arterial pH, PCO2, and PO2 were measured with a pH/blood-gas analyzer (AVL Compact 1, AVL Instruments Medicaux, Eragny-sur-Oise, France). Serum and urine electrolytes and creatinine were measured by standard methods with a Beckman LX20 autoanalyzer (Coulter-Beckman, Villepinte, France) and Olympus AU 400 (Olympus, Rungis, France). Plasma NH4+ was measured by the standard enzymatic method. Plasma amino acid levels were determined by ion-exchange chromatography with ninhydrin detection. Urinary pH and bicarbonates were measured with a pH/blood-gas analyzer (ABL 555, Radiometer, Copenhagen, Denmark). Urinary NH4+ and titratable acid were measured by titration with a DL 55 titrator (Mettler Toledo, Viroflay, France).
Measurement of bone mineral density by dual-energy X-ray absorptiometry. Analysis of all animals by dual-energy X-ray absorptiometry (DXA) was carried out under anesthesia. Total body, whole femur, and caudal vertebral bone mineral content (BMC; mg) and bone mineral density (BMD; mg/cm2) of age-matched female mice were measured using a PIXImus instrument (version 1.44; Lunar). An ultra-high-resolution mode (resolution, x0.18 mm) was used (20). The precision and reproducibility of the instrument were previously evaluated by calculating the coefficient of variation of repeated DXA measurements. The coefficient of variation was <2% for all evaluated parameters. A phantom was scanned daily to monitor the stability of the measurements.
Isolated, perfused tubule studies with WT and RHbg/ mice. Cortical collecting ducts (CCD) were dissected from female WT and RHbg/ mice at 2332 wk of age and microperfused in vitro (10). The basic approach used to determine NH3 permeability involved construction of a transepithelial gradient of NH3 and measurement of the resulting NH3 flux as previously described by Flessner et al. (10). Mice were allowed free access to autoclaved standard chow (M20, Dietex) and distilled water until the time of the experiments. Mice were anesthetized with 50 mg/kg pentobarbital sodium. Both kidneys were cooled in situ with control bath solution for 1 min and then removed and cut into thin coronal slices for tubule dissection. To obtain CCD segments, medullary rays were dissected from the slices at 10°C under a Wild M-8 dissecting microscope in the control bath solution of the experiment (see below). The isolated tubule was transferred to the bath chamber on the stage of an inverted microscope (Axiovert 100, Carl Zeiss) and mounted on concentric glass pipettes for microperfusion at 37°C. Bath solution was delivered at a rate of 20 ml/min and warmed to 37 ± 0.5°C by a water jacket immediately upstream of the chamber. The perfusion rate was adjusted by hydrostatic pressure to 10 nl/min. For most tubules, three to four collections were made. Two reliable collections were considered a minimum. The tubules were equilibrated for 2030 min at 37°C before the beginning of collections. To construct a transepithelial NH3 gradient, the perfusion (lumen) solution contained (in mM) 140 Na+, 1 NH4+, 5 K+, 2 HCO3, and 143 Cl; the bath solution contained (in mM) 140 Na+, 1 NH4+, 5 K+, 23 HCO3, and 122 Cl; in addition, both solutions contained (in mM) 5.5 glucose, 2 Ca2+, 1.2 Mg2+, 1.2 SO42, 2.5 HPO42, and 10 HEPES. The osmolarity of the solution was 295 ± 5 mosmol/kgH2O. All solutions were equilibrated with 95% O2-5% CO2. Once the solutions were gassed and the pH checked, they were placed in a reservoir and continuously bubbled with 95% O2-5% CO2. The actual pH of the solutions was monitored several times during experiments, and the pH of solutions was checked at the end of the experiment to ensure that changes did not occur. Carbonic anhydrase (no. C2522, Sigma, Saint Quentin Falavier, France) was added to the perfusate solution (1 mg/10 ml of solution). The purpose of carbonic anhydrase was to prevent any pH disequilibrium that might arise from proton secretion or NH3 transport. Transepithelial voltage was measured with a FD-223 differential electrometer (World Precision Instruments) by the use of a Ag/AgCl electrode connected to the perfusion pipette via a 0.15 M NaCl-agar bridge; a 0.15 M NaCl-agar bridge also connected the peritubular bath to an Ag/AgCl electrode. Transepithelial voltage was measured at the tip of the perfusion pipette during each period. Total ammonia concentration was measured in 10- to 12-nl samples of peritubular, perfused, and collected fluids using an NH3 diagnostic kit (Sigma) and the flow-through microfluorometer Nanoflo apparatus (World Precision Instruments) (45). To determine the initial and end-luminal pH, measurement of luminal pH with fluorescence microspectrometry utilizing BCECF (Molecular Probes, Eugene, OR) diluted in the luminal fluid at a 10 μM final concentration. The dye was excited alternatively at 490 and 440 nm with a 100-W halogen lamp and a computer-controlled chopper assembly. Emitted light was collected through a dichroic mirror, passed through a 530-nm filter, and focused onto a CCD camera (ICCD 2525F, Videoscope International) connected to a computer. The measured light intensities were digitalized with 8-bit precision (256 grey level scale) for further analysis. For each tubule, two regions of interest were drawn (one for initial luminal pH, the other for end-luminal pH), and the mean grey level for each excitation wavelength was calculated with Starwise Fluo software (Imstar, Paris, France). Background fluorescence was subtracted from fluorescence intensity at each excitation wavelength to obtain intensities of intraluminal fluorescence. The 490- to 440-nm ratio was used as an indicator of intraluminal pH. The following procedure was performed to calibrate the dye. The dye was diluted at the same concentration as in lumen in HCO3/CO2-free, HEPES-buffered solutions titrated to 6.3, 6.5, 6.7, or 6.9. A sample of each solution was introduced in a glass tube with the same inner diameter as the tubules, and the dye was excited as explained above. A linear calibration curve was derived from these measurements, and used to determine initial and end-luminal pH values.
Calculations of transepithelial NH3 permeability. Assuming an absence of osmotic or hydrostatic pressure gradients across the epithelium and therefore an absence of net fluid transport, the passive transepithelial transport of total ammonia (Am) may be described by
where PNH3 is diffusive permeability of NH3 (cm/s), As is tubule luminal surface area (cm2), and CNH3 is the transepithelial concentration difference for NH3 (mM).
To calculate the permeability to NH3, the equation is rearranged as follows
The net rate of transport JAm is calculated as follows
where [Am]i is the concentration of total ammonia in the perfusate, [Am]o is the concentration of total ammonia in the collected fluid, V is the collection rate (nl/min), as measured in precalibrated constriction pipettes, and L is the perfused tubule length (mm).
As may be calculated as Ld, where d (mm) is the inner tubule diameter.
The total ammonia concentration ([Am]) is equal to the sum of the concentrations of the two species NH3 and NH4+ and is the quantity actually measured by the microfluorimetric assay. The equilibrium between the two species is defined by the Henderson-Hasselbalch equation
The pKa equals 9.03 at physiological pH and temperature. Knowing the values for pH and [Am], the values for [NH3] and [NH4+] may be determined simultaneously.
Intracellular pH measurement. CCD cells were loaded with the fluorescent probe BCECF, prepared as a 20 mM stock in DMSO, by exposing the cells for 20 min at room temperature to the control bath solution containing 10 μM BCECF. Loading was continued until the fluorescence intensity at 440-nm excitation wavelength was at least one order of magnitude higher than background fluorescence. The loading solution was then washed out by initiation of bath flow and the tubule was equilibrated with dye-free control bath solution for 510 min. Bath solution was delivered at a rate of 20 ml/min and warmed to 37 ± 0.5°C by water jacket immediately upstream to the chamber. Perfusion rate was adjusted by hydrostatic pressure to 20 nl/min to prevent axial changes in the composition of luminal fluid.
Intracellular dye was excited alternatively at 490 and 440 nm with a 100-W halogen lamp and a computer-controlled chopper assembly. Emitted light was collected through a dichroic mirror, passed through a 530-nm filter, and focused onto a CCD camera (ICCD 2525F, Videoscope International) connected to a computer. The measured light intensities were digitized with 8-bit precision (256 grey level scale) for further analysis. For each tubule, a region of interest was drawn and the mean grey level for each excitation wavelength was calculated with the Starwise Fluo software (Imstar). Background fluorescence was subtracted from fluorescence intensity at each excitation wavelength to obtain intensities of intracellular fluorescence. The 490- to 440-nm ratio was used as an indicator of intracellular pH (pHi).
The control solution composition was (in mM) 142 Na+, 4 K+, 1.5 Ca2+, 1.2 Mg2+, 145 Cl, 1.2 SO42, 2 HPO42, 5.5 glucose, 5 alanine, and 10 HEPES. After a 120-s recording, the peritubular solution was changed for a solution containing 4 NH4+, 138 Na+, 4 K+, 1.5 Ca2+, 1.2 Mg2+, 145 Cl, 1.2 SO42, 2 HPO42, 5.5 glucose, 5 alanine, and 10 HEPES, whereas the luminal solution was left unchanged. All solutions were adjusted to pH 7.40 with Tris and continuously bubbled with 100% O2 passed through a 3 M KOH CO2 trap.
Intracellular dye was calibrated at the end of each experiment using the high [K+]-nigericin technique. Tubules were perfused and bathed with a HEPES-buffered, 95 mM K+ solution containing 10 μM of the K+/H+ exchanger nigericin. Four different calibration solutions, titrated to 6.9, 7.2, 7.5, or 7.7, were used.
Statistics. All data are represented as means ± SE. Comparisons were assessed by unpaired or paired Student's t-test, as appropriate. Differences were considered significant at P < 0.05.
RESULTS
Targeting of RHbg gene prevents renal and hepatic Rhbg protein expression. The mouse RHbg gene has been disrupted by insertional targeting (Fig. 1) using an original method previously described by Zheng et al. (46). Breeding of heterozygous F1 mice led to WT RHbg+/+, heterozygous RHbg+/, and homozygous mutant Rhbg/ [RHbg knockout (KO)] offspring with expected Mendelian ratios (65 +/+, 157 +/, 59 /). To verify that the insertional targeting event was effective in preventing Rhbg protein expression, we performed immunohistochemistry experiments with the anti-Rhbg antibody (29) on kidney and liver sections from RHbg/ and their WT littermates. Figure 2 shows that we were able to detect a normal expression of Rhbg protein in the kidney (Fig. 2A) and liver (Fig. 2, E and G) from WT mice, whereas no Rhbg protein could be detected in the RHbg/ mice, either in the kidney (Fig. 2, B, C, and D) or in the liver (Fig. 2, F and H). This confirmed that disruption of the RHbg gene, consecutive to insertional targeting, ablated Rhbg protein. There were also no obvious changes in the morphology of the renal tubules or the glomeruli.
Physiological parameters in standard laboratory conditions. Animals grew normally and did not display any obvious difference with respect to their phenotype. There were no differences in blood acid-base status between WT and RHbg/ mice, and particularly there was no evidence of metabolic acidosis with a value of blood [HCO3] of 21.4 ± 0.4 mM in female RHbg/ vs. 21.8 ± 0.3 mM in female WT mice (n = 35 for RHbg/ and n = 25 for WT, P = 0.43), and a blood pH value of 7.33 ± 0.01 for both groups. PO2 and PCO2 values were also identical (not shown). There was also no difference between mice in urinary pH [6.14 ± 0.09 vs. 6.00 ± 0.16 in WT, not significant (NS)]; NH4+ excretion (61 ± 11 vs. 42 ± 6 μmol/day in WT, NS); titratable acid (30 ± 5 vs. 24 ± 4 μmol/day in WT, NS); and net acid excretion (91 ± 15 vs. 66 ± 8 μmol/day in WT, NS). Identical blood acid-base status, urinary pH, titratable acid, and NH4+ excretion were also observed in male mice (not shown). Therefore, because to our knowledge no sex dependence of acid-base regulation, NH3 transport, or NH3 metabolism has been reported, all the following experiments were performed only in female mice.
The other physiological parameters measured in RHbg/ and WT mice are summarized in Tables 1 and 2. There were no detectable differences between RHbg/ and WT mice. In particular, liver nitrogen metabolism was normal with the absence of hyperammonemia, and there was normal plasma (see Table 1). There was also no significant difference in plasma urea, ornithine, citrulline, and arginine concentrations, and in urea urinary excretion, indicating that it is unlikely that an increase in urea synthesis compensated for a defect in glutamine synthesis (Table 1).
There were also no differences in plasma concentrations and urinary excretion of Na+, K+, Cl, total calcium, and inorganic phosphate, between RHbg/ and WT mice (see Table 2).
Measurement of BMD of WT and RHbg/ mice. Incomplete distal tubular acidosis in human patients (i.e., impairment of distal renal acidification with no overt acidosis) is frequently discovered as isolated osteoporosis and hypercalciuria (39, 40). Therefore, we assessed BMD of 16 WT and 22 RHbg/ mice, at 4 mo of age, by DEXA. There was no evidence for bone demineralization of RHbg/ vs. WT mice, with a BMD value of 0.082 ± 0.002 mg/cm2 in RHbg/ mice vs. 0.078 ± 0.002 mg/cm2 in WT when measured at the femur (P = 0.13), and 0.064 ± 0.001 mg/cm2 in RHbg/ mice vs. 0.062 ± 0.006 mg/cm2 in WT, when measured at the caudal vertebrae (P = 0.35). Total BMD was also not different (0.056 ± 0.001 mg/cm2 in RHbg/ mice vs. 0.054 ± 0.001 mg/cm2 in WT, P = 0.16). Values of BMC were also not different (not shown). RHbg/ mice did also not exhibit hypercalciuria with respect to WT (0.78 ± 0.13 mmol/mmol creatinine in KO mice vs. 1.28 ± 0.3 mmol/mmol creatinine in controls, n = 10 for both groups, NS).
Immunolocalization of Rhcg, AE1, and the 56-kDa 1-subunit of the H+ pump in kidney sections from RHbg/ and WT mice. We next examined whether RHbg inactivation disrupted the expression of the other proteins involved in NH4+ transport in the collecting duct, i.e., the basolateral Cl/HCO3 exchanger AE1 and the apical H+ pump, or of Rhcg, the other renal Rh glycoprotein isoform, or whether regulation of these proteins could have accounted for a compensatory process masking the impairment in Rhbg-mediated NH3 transport. As shown in Fig. 3, we could find no alteration in the expression level and subcellular localization of either Rhcg (Fig. 3A compared with 3B), H+-ATPase (Fig. 3C compared with 3D), or AE1 (Fig. 3E compared with 3F) by immunohistochemistry, despite RHbg disruption.
Response to acid loading of RHbg/ and WT mice. To examine the possibility that RHbg/ mice may have an incomplete form of distal tubular acidosis, they were submitted to a chronic acid-loading challenge. Acid loading was performed as an NH4Cl load in the drinking water. All the animals tolerated the acid-loading challenge well and did not develop obvious signs of hepatic encephalopathy. The time course of the blood pH, PCO2, and HCO3 in WT and RHbg/ mice during the acid load testing is summarized in Table 3. As expected, all the mice subjected to NH4Cl loading developed a frank metabolic acidosis with a maximal decrease in blood [HCO3] and pH at day 2. There was no difference in the degree of metabolic acidosis between both groups. Within the following days, blood [HCO3] as well as blood pH progressively returned toward normal values, indicating that the mice adapted (see Table 3). Once again, the adaptation seemed not to be different because RHbg/ mice normalized their blood [HCO3] and pH values as well as DID their respective WD controls.
Figure 4 shows the renal adaptation mechanisms of the mice to the acid-loading challenge. As expected, a very rapid decrease in urinary pH was observed, from the basal value of 6.0 ± 0.2 in WT mice and of 6.1 ± 0.1 in RHbg/ mice to the value of 5.5. The plateau was reached within 48 h for both genotypes (Fig. 4A). Following this decrease in urinary pH, net acid excretion [i.e., the sum of (NH4+ + titratable acid HCO3 urinary excretions)] rose markedly in WT and RHbg/ mice, reaching a maximal value at day 3 (maximal value represented a 4-fold increase in net acid excretion in both genotypes) and accounting for the normalization of acid-base status (see Fig. 4B). Figure 4, C and D, shows that in both WT and RHbg/ mice the increase in urinary NH4+ excretion accounted exclusively for the observed increase in net acid excretion and that there was no significant change in titratable acid. Importantly, there was no difference in the different renal responses during the acid-loading challenge between WT and RHbg/ mice.
NH3 permeability of cortical collecting ducts from WT and RHbg/ mice measured by in vitro microperfusion. We next directly measured diffusive NH3 permeability across CCD epithelium from WT and RHbg/ mice. As shown in Table 4, imposing a bath-to-lumen NH3 gradient, in the nominal absence of an NH4+ gradient, generated a measurable NH3 secretion flux of 25.78 ± 4.21 pmol?min1?mm tubule1 in WT, and a similar secretion flux of 27.44 ± 4.36 pmol?min1?mm tubule1 in RHbg/ mice. This corresponded to similar NH3 permeabilities in both genotype (0.040 ± 0.009 cm/s in WT mice vs. 0.044 ± 0.012 cm/s in RHbg/ mice). NH3 permeabilities of the different portions of the mouse collecting duct have never been reported, but the values reported in our study are very close to that reported by Flessner et al. (10) in the CCD of the rat.
In the collecting duct, the basolateral membrane, where Rhbg is expressed, is believed to be the limiting step of NH4+ transepithelial transport (44). However, in the preceding experiments, even a major alteration of Rhbg-mediated NH3 uptake could be undetectable if the apical membrane actually represents the limiting step. Therefore, we next tested directly the effect of Rhbg disruption on NH3 and NH4+ uptake across the basolateral membrane of CCD cells. For that purpose, we assessed the effects of an inwardly directed NH3/NH4+ gradient on pHi of CCD isolated from RHbg/and WT mice. Figure 5A depicts the time course of the pHi changes in those experiments. As expected, after addition of 4 mM NH4+ to the bath, the cells alkalinized rapidly as the result of rapid NH3 uptake and subsequent NH3 protonation to form NH4+ in the cells. This alkalinization step was followed by a marked and prolonged intracellular acidification. The initial acidification rate is believed to reflect NH4+ entry that releases NH3 and a proton into the cells. Figure 5B shows that neither the NH3-dependent alkalinization rate nor the NH4+-dependent acidification rate was different between RHbg/ and WT mice, indicating that Rhbg disruption did not result in alteration in NH4+ or NH3 permeability across the basolateral membrane of CCD cells.
DISCUSSION
The present study demonstrates that Rhbg, an epithelial ortholog of red blood cell Rh proteins expressed in the kidney and the liver, is not necessary for renal NH4+ excretion and for hepatic clearance of NH4+ from the plasma. This conclusion is based on the observations that mice with a disrupted RHbg gene 1) do not exhibit renal metabolic acidosis in standard laboratory conditions, 2) are able to normally increase their urinary NH4+ excretion in response to metabolic acidosis, and 3) do not suffer from encephalopathy or exhibit hyperammonemia or a decrease in plasma glutamine concentration. We also demonstrate that in the CCD, NH3 diffusion permeability is not dependent on the presence of Rhbg protein and that ablation of Rhbg protein does not alter either NH3 or NH4+ entry across the basolateral membrane of CCD cells.
We used an insertional gene-targeting strategy to knock out the RHbg gene (46). Southern blot analysis of the DNA from RHbg/ mice indicated that the targeting event was effective. The insertion of the vector resulted in duplication of a part of the RHbg gene, leaving the initial five exons (of a total of 10) intact (see Fig. 1). Immunohistochemical experiments using an antibody recognizing the COOH-terminal end of the Rhbg protein (29) demonstrated that the sequence downstream of the vector was not expressed. Accordingly, only a truncated RNA species encompassing exons 15, and not the full-length Rhbg RNA, could be detected by RT-PCR analysis of the KO mice transcripts (not shown). It is very unlikely that a cell surface-expressed truncated protein could be translated from the transcript from exons 15 because this protein would lack the COOH-terminal cytoplasmic domain recently shown to be critical for membrane expression of RhBG (22). Furthermore, even if this truncated protein could be expressed, it is anticipated that the lack of transmembrane-spanning domains M8, M9, M10, and M11, including numerous highly conserved residues, such as H318, an amino acid that is involved in the stabilization of NH3 within the pore-forming channel (see Fig. 1 in Ref. 16), will preclude any transport activity.
We did not detect any impairment in renal NH4+ excretion in RHbg/ mice. However, in humans, impairment of distal tubule acidification (i.e., distal NH4+ transport function) does not always lead to overt metabolic acidosis, and patients with incomplete distal tubular acidosis present with normal urinary pH and NH4+ excretion (31). In these patients, the administration of an acid load unmasks the renal acidification disorder by demonstrating their inability to increase their net acid excretion normally, mainly as urinary NH4+ (8, 27, 32, 43). However, in the present study, RHbg/ mice exhibited a normal response to the chronic acid loading. There was also no evidence of bone demineralization, another classic feature of incomplete renal tubular acidosis.
In the collecting duct, Rhcg, the apical ortholog of Rhbg, is normally expressed at the apical pole of the same cells that express Rhbg at the basolateral membrane (29), and NH3 transport is driven by the pH gradient generated by H+ secretion (19) that occurs through the combined action of the basolateral Cl/HCO3 exchanger AE1 and the apical H+ pump (36). Rhcg was not altered by RHbg disruption and remained restricted to the apical membrane of the cells, and thus it is unlikely that Rhcg has compensated for the absence of Rhbg transport function. It is also unlikely that the other mechanisms involved in net NH4+ transport by the collecting duct could have participated in a compensatory process masking a defect in Rhbg-mediated NH3 transport. Indeed, we found no evidence for the stimulation of the main proteins involved in H+ secretion, urinary pH was identical in RHbg+/+ and RHbg/ mice under basal conditions and decreased to the same extent after acid loading, and there was no difference in the amount of titratable acid excreted by RHbg/ and WT mice despite RHbg disruption. Finally, to rule out the possibility than a compensatory mechanism has masked a defect in Rhbg-dependent NH3/NH4+ transport, we also assessed directly NH3 or NH4+ transport across the basolateral membrane of CCD cells and NH3 diffusive permeability across CCD epithelium by in vitro microperfusion. We chose to study the CCD rather than the other segments of the collecting duct because we and others have described that Rhbg is expressed in a larger amount in the cortex than in the medulla (29, 37). It is noteworthy that this higher amount of Rhbg protein parallels a higher permeability to NH3 of the CCD than that of medullary segments, as demonstrated by Flessner et al. (10). Thus we hypothesized that a defect in Rhbg-mediated NH3 transport would have a more dramatic effect that would be more easily detectable in the CCD than in medullary portions of the collecting duct. However, once again we could not find any evidence for impaired NH3 or NH4+ transport consecutive to Rhbg disruption in these studies. There were also no evidences of alteration in liver NH4+ extraction and detoxification and of glutamine metabolism. KO mice did not exhibit obvious differences in their behavior and survived normally, indicating that it was unlikely that they were suffering from hepatic encephalopathy. The alternate pathway of NH3 extraction and detoxification from the plasma by the liver is the urea cycle that occurs in midzonal and periportal hepatocytes (12). Urea synthesis and glutamine synthesis from NH3 are two independent pathways, differentially regulated (12). Thus we hypothesized that a shift from glutamine biosynthesis by perivenous hepatocytes to urea synthesis by the other hepatocytes might have limited the accumulation of NH3 in the plasma. However, there was no significant difference in plasma urea and in the concentrations of the main amino acids involved in the urea cycle, namely, ornithine, citrulline, and arginine, indicating that it is unlikely that an increase in urea synthesis compensated for a defect in glutamine synthesis.
The conclusion that Rhbg is not critical for NH4+ transport in the kidney and the liver goes counter to expectations from heterologous expression studies (2, 24, 28, 42), crystallography data (16, 47), and stopped-flow analyses of red blood cells (30). However, all these studies were performed in vitro or ex vivo and could not therefore determine the physiological impact of Rh glycoprotein-mediated transport in vivo. Moreover, they were aimed at determining the precise molecular transport mechanism of NH4+, NH3, or of its derivative, methylammonium. Because they did not test transport specificity of the different Rh glycoproteins studied, they do not rule out the possibility that, in vivo, transport of another substrate occurs preferentially to NH3 through Rh glycoproteins. CO2 is another gas that has also a high diffusion permeability through cell membranes, and it has been suggested that membrane transport proteins may account for the high CO2 permeability of some cells (6). It has also been proposed that the Mep/Amt/Rh protein family represents a family of gas channels that may also accept CO2 as a substrate (3335). However, this hypothesis remains speculative and certainly requires a more direct demonstration. Nonetheless, if Rhbg transports CO2 into the collecting duct cells, mice would probably display an acidification defect similar to that observed when the Cl/HCO3 exchanger is inactivated (1). Indeed, CO2 must enter the intercalated cells to be hydrated to yield an H+ and an HCO3, a reaction catalyzed by the enzyme carbonic anhydrase. This reaction is believed to provide the principle source of H+ required for H+-ATPase function and to provide HCO3 that is basolaterally extruded through AE1 to replenish the body's HCO3 stores (19). Accordingly, it has been demonstrated that acetazolamide (21, 38), an inhibitor of carbonic anhydrase, or the nominal absence of CO2/HCO3 (38) is able to abolish H+ secretion of isolated collecting ducts microperfused in vitro. Thus, although we did not directly measure CO2 transport across the basolateral membrane of collecting duct cells, our experiments demonstrating that RHbg/ mice do not display any urinary acidification defect render unlikely the possibility that Rhbg is a CO2 transporter.
Apart from their NH3 transport properties, several observations about the Mep/Amt/ Rh proteins are striking. It has been demonstrated that MEP2, a Rh glycoprotein paralog of yeast, and AmtB, in Escherichia coli, are NH4+-sensing receptors (14, 23). A possible role of Rh glycoproteins in cell signaling or proliferation has also been suggested. The first report of an RhCG sequence in GenBank was as human PDRC2 (accession no. AF081497), identified as a tumor-related protein. More convincing is the recent identification of RHbg and RHcg as candidate cancer-causing genes in a murine model of glioblastoma (15). The authors identified a total of 66 candidate cancer-causing loci from 108 independent tumors. Rhbg and Rhcg were both tagged, each in three independent tumors (15). Thus it is tempting to speculate that the Rh glycoproteins might not function as epithelial transporters but are rather involved in cell signaling.
In summary, we have demonstrated that RHbg inactivation does not lead to distal tubular acidosis or to hyperammonemic hepatic encephalopathy, arguing against a critical role of this protein in NH3 epithelial transport.
GRANTS
This work was supported by the Institut National de la Sante et de la Recherche Medicale (INSERM), the Institut National de la Transfusion Sanguine, and by an INSERM/AFM grant on rare diseases to M. Paillard.
ACKNOWLEDGMENTS
We thank Cyndie Dampattiah, Julien Meurot, and Cecile Rahuel for expert technical assistance; Charles Babinet for helpful discussions about the generation of the RHbg KO mice; Allan Bradley for providing us with the 3'HPRT mouse genomic library; Seth Alper for providing us with the antibody recognizing AE1; Sylvie Breton and Dennis Brown for providing us with the anti- H+-ATPase antibody; and Daniel Rabier for help in the measurements of plasma amino acid concentrations.
FOOTNOTES
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.
REFERENCES
Alper SL. Genetic diseases of acid-base transporters. Annu Rev Physiol 64: 899923, 2002.
Bakouh N, Benjelloun F, Hulin P, Brouillard F, Edelman A, Cherif-Zahar B, and Planelles G. NH3 is involved in the NH4+ transport induced by the functional expression of the human RhC glycoprotein. J Biol Chem 279: 1597515983, 2004.
Bank N and Schwartz WB. Influence of certain urinary solutes on acidic dissociation constant of ammonium at 37°C. J Appl Physiol 15: 125127, 1960.
Breton S, Alper SL, Gluck SL, Sly WS, Barker JE, and Brown D. Depletion of intercalated cells from collecting ducts of carbonic anhydrase II-deficient (CAR2 null) mice. Am J Physiol Renal Fluid Electrolyte Physiol 269: F761F774, 1995.
Brown D, Hirsch S, and Gluck S. Localization of a proton-pumping ATPase in rat kidney. J Clin Invest 82: 21142126, 1988.
Cooper GJ, Zhou Y, Bouyer P, Grichtchenko II, and Boron WF. Transport of volatile solutes through AQP1. J Physiol 542: 1729, 2002.
Eladari D, Cheval L, Quentin F, Bertrand O, Mouro I, Cherif-Zahar B, Cartron JP, Paillard M, Doucet A, and Chambrey R. Expression of RhCG, a new putative NH3/NH4+ transporter, along the rat nephron. J Am Soc Nephrol 13: 19992008, 2002.
Elkinton JR, Huth EJ, Webster GD Jr, and McCance RA. The renal excretion of hydrogen ion in renal tubular acidosis. I. Quantitative assessment of the response to ammonium chloride as an acid load. Am J Med 29: 554575, 1960.
Flessner MF, Wall SM, and Knepper MA. Ammonium and bicarbonate transport in rat outer medullary collecting ducts. Am J Physiol Renal Fluid Electrolyte Physiol 262: F1F7, 1992.
Flessner MF, Wall SM, and Knepper MA. Permeabilities of rat collecting duct segments to NH3 and NH4+. Am J Physiol Renal Fluid Electrolyte Physiol 260: F264F272, 1991.
Good DW and Knepper MA. Ammonia transport in the mammalian kidney. Am J Physiol Renal Fluid Electrolyte Physiol 248: F459F471, 1985.
Hussinger D, Meijer AJ, Gerok W, and Sies H. Hepatic nitrogen metabolism and acid-base homeostasis. In: pH Homeostasis, edited by Hussinger D. London: Academic, 1988, p. 337377.
Huang CH and Liu PZ. New insights into the Rh superfamily of genes and proteins in erythroid cells and nonerythroid tissues. Blood Cells Mol Dis 27: 90101, 2001.
Javelle A, Severi E, Thornton J, and Merrick M. Ammonium sensing in Escherichia coli. Role of the ammonium transporter AmtB and AmtB-GlnK complex formation. J Biol Chem 279: 85308538, 2004.
Johansson FK, Brodd J, Eklof C, Ferletta M, Hesselager G, Tiger CF, Uhrbom L, and Westermark B. Identification of candidate cancer-causing genes in mouse brain tumors by retroviral tagging. Proc Natl Acad Sci USA 101: 1133411337, 2004.
Khademi S, O'Connell J, Remis J III, Robles-Colmenares Y, Miercke LJ, and Stroud RM. Mechanism of ammonia transport by Amt/MEP/Rh: structure of AmtB at 1.35 A. Science 305: 15871594, 2004.
Knepper MA and Agre P. Structural biology. The atomic architecture of a gas channel. Science 305: 15731574, 2004.
Knepper MA, Good DW, and Burg MB. Ammonia and bicarbonate transport by rat cortical collecting ducts perfused in vitro. Am J Physiol Renal Fluid Electrolyte Physiol 249: F870F877, 1985.
Knepper MA, Packer R, and Good DW. Ammonium transport in the kidney. Physiol Rev 69: 179249, 1989.
Kolta S, De Vernejoul MC, Meneton P, Fechtenbaum J, and Roux C. Bone mineral measurements in mice: comparison of two devices. J Clin Densitom 6: 251258, 2003.
Lombard WE, Kokko JP, and Jacobson HR. Bicarbonate transport in cortical and outer medullary collecting tubules. Am J Physiol Renal Fluid Electrolyte Physiol 244: F289F296, 1983.
Lopez C, Metral S, Eladari D, Drevensek S, Gane P, Chambrey R, Bennett V, Cartron JP, Le Van Kim C, and Colin Y. The ammonium transporter RhBG: requirement of a tyrosine-based signal and ankyrin-G for basolateral targeting and membrane anchorage in polarized kidney epithelial cells. J Biol Chem 280: 82218228, 2005.
Lorenz MC and Heitman J. The MEP2 ammonium permease regulates pseudohyphal differentiation in Saccharomyces cerevisiae. EMBO J 17: 12361247, 1998.
Ludewig U. Electroneutral ammonium transport by basolateral rhesus B glycoprotein. J Physiol 559: 751759, 2004.
Marini AM, Matassi G, Raynal V, Andre B, Cartron JP, and Cherif-Zahar B. The human Rhesus-associated RhAG protein and a kidney homologue promote ammonium transport in yeast. Nat Genet 26: 341344, 2000.
Marini AM, Urrestarazu A, Beauwens R, and Andre B. The Rh (rhesus) blood group polypeptides are related to NH4+ transporters. Trends Biochem Sci 22: 460461, 1997.
Morris RC Jr. Renal tubular acidosis. Mechanisms, classification and implications. N Engl J Med 281: 14051413, 1969.
Nakhoul NL, Dejong H, Abdulnour-Nakhoul SM, Boulpaep EL, Hering-Smith K, and Hamm LL. Characteristics of renal Rhbg as an NH4+ transporter. Am J Physiol Renal Physiol 288: F170F181, 2005.
Quentin F, Eladari D, Cheval L, Lopez C, Goossens D, Colin Y, Cartron JP, Paillard M, and Chambrey R. RhBG and RhCG, the putative ammonia transporters, are expressed in the same cells in the distal nephron. J Am Soc Nephrol 14: 545554, 2003.
Ripoche P, Bertrand O, Gane P, Birkenmeier C, Colin Y, and Cartron JP. Human Rhesus-associated glycoprotein mediates facilitated transport of NH3 into red blood cells. Proc Natl Acad Sci USA 101: 1722217227, 2004.
Rodriguez Soriano J. Renal tubular acidosis: the clinical entity. J Am Soc Nephrol 13: 21602170, 2002.
Sartorius OW, Roemmelt JC, Pitts RF, Calhoon D, and Miner P. The renal regulation of acid-base balance in man. IV. The nature of the renal compensations in ammonium chloride acidosis. J Clin Invest 28: 423439, 1949.
Soupene E, Chu T, Corbin RW, Hunt DF, and Kustu S. Gas channels for NH3: proteins from hyperthermophiles complement an Escherichia coli mutant. J Bacteriol 184: 33963400, 2002.
Soupene E, Inwood W, and Kustu S. Lack of the Rhesus protein Rh1 impairs growth of the green alga Chlamydomonas reinhardtii at high CO2. Proc Natl Acad Sci USA 101: 77877792, 2004.
Soupene E, Ramirez RM, and Kustu S. Evidence that fungal MEP proteins mediate diffusion of the uncharged species NH3 across the cytoplasmic membrane. Mol Cell Biol 21: 57335741, 2001.
Stone DK, Seldin DW, Kokko JP, and Jacobson HR. Anion dependence of rabbit medullary collecting duct acidification. J Clin Invest 71: 15051508, 1983.
Verlander JW, Miller RT, Frank AE, Royaux IE, Kim YH, and Weiner ID. Localization of the ammonium transporter proteins RhBG and RhCG in mouse kidney. Am J Physiol Renal Physiol 284: F323F337, 2003.
Wall SM and Fischer MP. Contribution of the Na+-K+-2Cl cotransporter (NKCC1) to transepithelial transport of H+, NH4+, K+, and Na+ in rat outer medullary collecting duct. J Am Soc Nephrol 13: 827835, 2002.
Weger M, Deutschmann H, Weger W, Kotanko P, and Skrabal F. Incomplete renal tubular acidosis in "primary" osteoporosis. Osteoporos Int 10: 325329, 1999.
Weger W, Kotanko P, Weger M, Deutschmann H, and Skrabal F. Prevalence and characterization of renal tubular acidosis in patients with osteopenia and osteoporosis and in non-porotic controls. Nephrol Dial Transplant 15: 975980, 2000.
Weiner ID, Miller RT, and Verlander JW. Localization of the ammonium transporters, Rh B glycoprotein and Rh C glycoprotein, in the mouse liver. Gastroenterology 124: 14321440, 2003.
Westhoff CM, Ferreri-Jacobia M, Mak DO, and Foskett JK. Identification of the erythrocyte Rh blood group glycoprotein as a mammalian ammonium transporter. J Biol Chem 277: 1249912502, 2002.
Wrong O and Davies HE. The excretion of acid in renal disease. Q J Med 28: 259313, 1959.
Yip KP and Kurtz I. NH3 permeability of principal cells and intercalated cells measured by confocal fluorescence imaging. Am J Physiol Renal Fluid Electrolyte Physiol 269: F545F550, 1995.
Zhelyaskov VR, Liu S, and Broderick MP. Analysis of nanoliter samples of electrolytes using a flow-through microfluorometer. Kidney Int 57: 17641769, 2000.
Zheng B, Mills AA, and Bradley A. A system for rapid generation of coat color-tagged knockouts and defined chromosomal rearrangements in mice. Nucleic Acids Res 27: 23542360, 1999.
Zheng L, Kostrewa D, Berneche S, Winkler FK, and Li XD. The mechanism of ammonia transport based on the crystal structure of AmtB of Escherichia coli. Proc Natl Acad Sci USA 101: 1709017095, 2004.
Zidi-Yahiaoui N, Mouro-Chanteloup I, D'Ambrosio AM, Lopez C, Gane P, CLVK, Cartron JP, Colin Y, and Ripoche P. Human Rhesus B and Rhesus C glycoproteins: properties of facilitated ammonium transport in recombinant kidney cells. Biochem J. In press.