|Home | About | Journals | Submit | Contact Us | Français|
The kidney plays key roles in extracellular fluid pH homeostasis by reclaiming bicarbonate (HCO3−) filtered at the glomerulus and generating the consumed HCO3− by secreting protons (H+) into the urine (renal acidification). Sodium-proton exchangers (NHEs) are ubiquitous transmembrane proteins mediating the countertransport of Na+ and H+ across lipid bilayers. In mammals, NHEs participate in the regulation of cell pH, volume, and intracellular sodium concentration, as well as in transepithelial ion transport. Five of the 10 isoforms (NHE1-4 and NHE8) are expressed at the plasma membrane of renal epithelial cells. The best-studied isoform for acid-base homeostasis is NHE3, which mediates both HCO3− absorption and H+ excretion in the renal tubule. This article reviews some important aspects of NHEs in the kidney, with special emphasis on the role of renal NHE3 in the maintenance of acid-base balance.
The free hydrogen ion (H+) concentration in body fluids is regulated exquisitely around 40 nmol/L (pH 7.40) whereas H+ flux through the body greatly exceeds this magnitude. In a 70-kg human being at a basal state, normal metabolic and dietary acid production rate is about 50 to 70 mmoles/d and respiratory volatile acid production at the basal state is around 15,000 mmoles/d, with peak production at maximal exercise reaching 200 mmoles/min. The flux of H+ through the organism over 24 hours is 8 orders of magnitude greater than the total pool of free H+ in total body water (<2 μmoles). This remarkable homeostatic feat is accomplished by concerted efforts of extracellular and intracellular buffers, highly efficient ventilatory responses, metabolic functions of the liver, and renal ammoniagenic and solute transport mechanisms. For excretion of nonvolatile acid and base loads, the kidney assumes the pivotal role.
A filtration-reabsorption nephron bears an exorbitant burden of having to reclaim a vast amount of valuable solutes indiscriminately dispensed in the filtrate; one of which of course is the approximately 4,000 mmoles of bicarbonate (HCO3−) per day. The luminal acid disequilibrium pH implies that the predominant mode of HCO3− absorption involves H+ secretion, although a small concomitant degree of direct HCO3− reabsorption cannot be excluded.1 Two points are noteworthy. First, complete reclamation of the approximately 4,000 mmoles of filtered HCO3− forestalls a physiologic disaster but does not lead to net acid secretion. Further elaboration of H+ into the urine is necessary. Second, whether the organism is catering to the need of excreting a physiologic amount of acid (eg, 50 mmoles of H+ added to the body/d) or base (eg, 50 mmoles OH− added to the body/d), the kidney is always engaged in luminal H+ extrusion because HCO3− reclamation far exceeds physiologic H+ or OH− excretion. Luminal H+ secretion is a quintessential part of renal homeostatic function. Extrusion of H+ into the urinary lumen against its electrochemical gradient is an energetically costly process. Luminal H+ secretion can be coupled directly to adenosine triphosphate (ATP) hydrolysis by the multisubunit V-type ATPase or to the inwardly directed (urine lumen to cell) Na+ gradient because of the low cell [Na+] generated by the Na+/K+-ATPase (Fig 1).
In 1949, Pitts et al2 made the observation of an inverse relationship between urine pH and [Na+] (lowest urine pH with highest urinary Na+) and postulated a Na+/H+ exchange process between the renal epithelium and urine. Although the observation is correct, we are now cognizant that the analysis of bladder urine does not possess the resolution to permit conclusion about such transport mechanisms. Pitts et al2 in fact were witnessing the effect of distal Na+ delivery to enhance luminal H+ secretion by the collecting duct, which can be interpreted as a form of Na+/H+ exchange. In 1976, Murer et al3 first showed Na+-driven H+ movement and H+-driven Na+ movement in isolated cortical brush-border membrane vesicles, thereby definitely showing Na+/H+ exchange activity in the kidney. A subsequent report by Kinsella and Aronson4 further characterized this process in more detail and precision. Both of these reports have been revisited as milestone reports by the investigators.5,6 A cornucopia of data on renal Na+/H+ exchange using membrane vesicles and some cultured cells emerged in the 1980s that set the stage for the next level of research.
The era of phenomenologic analysis was supplemented by gene- and protein-specific reagents when the first mammalian NHE was cloned by Sardet et al.7 The relationship between mammalian NHE genetic sequence and those from a multitude of organisms is shown in Fig 2. Na+/H+ exchange across lipid bilayers and proteins that sustain this function are universal in prokaryotic, animal, and plant biology. Genes coding for Na+/H+ exchangers have been cloned from the simplest prokaryote to the most advanced multicellular eukaryotes. A remarkably high degree of conservation exists between NHE gene sequences from different organisms (Fig 2). Although the identification of paralogs and orthologs has been successfully accomplished by nucleotide homology–based cross-hybridization (across genera and species but not across orders and phyla), one common theme in breaking ground in the cloning of complementary DNAs (cDNAs) for Na+/H+ exchangers is the elegant use of functional complementation. One observes remarkable similarity in the approach in cloning NhaA from Escherichia coli by Goldberg et al8 and human (NHE1) by Sardet et al.7 NHE null and NHE over-expressing mutant cells were generated genetically followed by functional selection. Null mutants then were rescued by genomic sequences derived from overexpressing mutants, and functional complementation was used as an index to track down sequences coding for Na+/H+ exchange activity. These pioneering efforts then allowed subsequent cDNA identification by homology cloning of a multitude of cDNAs in vitro and in silico.
Mammalian NHEs that have been classified functionally to date are all electroneutral transporters with a 1Na+:1H+ stoichiometry. The putative structure of the mammalian protein with a 10- to 12-transmembrane N-terminus and a largely cytoplasmic C-terminus has been discussed in recent reviews.9,10 Currently, the structure of the mammalian NHEs is really unknown. Although a low-resolution version of the prokaryotic protein is available,11 the degree of sequence similarity does not permit homology modeling of the mammalian protein at present. Brett et al12 proposed an elegant categorization of mammalian NHEs into 3 classes based on their sequence and cell biology: (1) primarily plasma membrane-residing, (2) recycling between plasma membrane and endosomes, and (3) intracellular organellar.
A number of mammalian NHEs have been identified in the kidney. The role of intracellular NHEs in the kidney is not known at the moment. Data in model systems of proximal tubule epithelia suggest possible mediation of albumin endocytosis and degradation by endosomal NHE3.13-15 Intracellular [Na+] in renal epithelia is less than 20 mmol/L, whereas basolateral [Na+] ranges from 140 mmol/L in the cortex to as high as 300 mmol/L in the deep medulla. Basolateral NHEs undoubtedly will eject H+ from the cell to the interstitium and hence are unlikely to contribute to luminal acidification. Base-excreting cells (eg, β-intercalated collecting duct cells) use H+-ATPase rather than Na+/H+ exchange for H+ addition into the plasma. Apical NHEs on the other hand are poised strategically to add H+ into the lumen using the inward Na+ chemical gradient. Luminal [Na+] decreases while luminal [H+] increases axially toward the distal nephron to the point at which the ion gradients will no longer support luminal H+ extrusion so the task of luminal H+ extrusion is relegated to the V-ATPase where H+ pumping is energized directly by ATP hydrolysis (Fig 1). Because NH4+ can substitute for H+ as a substrate, NHEs also are important NH4+ transporters. NHEs also perform multiple other transport functions via parallel coupling with other transporters, but these functions are not discussed here. Five Na+/H+ exchanger isoforms (NHE1-4 and NHE8) are expressed at the plasma membrane of renal epithelial cells, with specific distribution within cells and along the nephron (Fig 3).
NHE1 is expressed at the plasma membrane of most mammalian cells where it plays multiple roles including in cell pH, sodium and volume homeostasis, cell motility, and provision of a platform for signaling complexes. In the kidney, NHE1 is expressed at the basolateral membrane of all nephron segments (Fig 3), with the exception of the macula densa and cortical collecting duct intercalated cells.16,17 Isolated, microperfused, thick ascending limbs (TALs) from NHE1 knockout mice have decreased HCO3− absorption compared with TALs from wild-type mice.18 This was postulated as regulatory cross-talk between basolateral NHE1 and apical NHE3 because pharmacologic inhibition of basolateral NHE1 downregulates apical NHE3 activity.19,20 The mechanism of this postulated cross-talk is unclear. However, 2 different mouse models lacking NHE1 have no overt disturbance of whole-body acid-base homeostasis.21,22 In addition, NHE1 potentially could mediate basolateral ammonium (NH4+) extrusion in the TAL (Fig 4).
NHE2 is expressed at the apical membrane in the TAL along with NHE3. It is the only luminal Na+/H+ exchanger from the macula densa to the distal nephron (Fig 3). In conjunction with NHE3, NHE2 contributes to acidification and bicarbonate absorption in the TAL23 (Fig 4). However, deletion of NHE2 in mice results in no overt acid-base disturbance,24 and double knock-out of NHE2 and 3 does not lead to further worsening of the metabolic acidosis observed in NHE3−/−mice.25 NHE2 thus may have a relatively minor role in the maintenance of acid-base balance compared with NHE3.
NHE3 is expressed at the apical (luminal) membrane of the proximal tubule, some long thin descending limbs, and the TAL of the loop of Henle (Fig 3), where it plays important roles in bicarbonate absorption (Fig 4), salt and volume homeostasis, and in the absorption of other solutes by functional coupling to a variety of other transporters. Mice with targeted disruption of NHE3 have decreased renal absorption of Na+, fluid and HCO3−, metabolic acidosis, hypovolemia, hypotension, and increased mortality.26 This phenotype is ameliorated partially but not abolished by rescue of the intestinal NHE3 defect.27,28 The moderate phenotype of NHE3 knockout mice can be attributed to a number of compensatory mechanisms, including decreased filtered HCO3− load and increased HCO3− absorption in the collecting duct, mediated by the H+-ATPase and H+-K+-ATPase.29 Compensation by the Na+/H+ exchange activity of NHE2 in the TAL and NHE8 in the proximal tubule theoretically is possible, but has not yet been supported by experimental findings.
A large body of literature exists addressing the regulation of NHE3 studied at the level of the intact microperfused tubule, renal cortical slices, enriched proximal tubules in suspension, brush-border membrane vesicles, cultured cells expressing native NHE3 (opposum kidney OK cells and porcine kidney LLC-PK1 cells), various eukaryotic hosts transfected with heterologous NHE3, and as purified recombinant NHE3-derived polypeptides. Table 1 lists some of the acute and chronic regulatory agents.
NHE4 is expressed ubiquitously at the basolateral membrane of epithelial cells along the nephron, together with NHE1. NHE4 is the only basolateral NHE isoform in the macula densa and intercalated cells of the cortical collecting duct9,30 (Fig 3). Similar to NHE1, NHE4 may mediate NH4+ reabsorption in the TAL (Fig 4). NHE4 knockout mice have no documented overt disease phenotype to date.31
NHE8, the most recently identified renal NHE isoform, is expressed at the luminal membrane of the proximal tubule32,33 (Fig 3). In the neonate, proximal tubule NHE activity is relatively high despite very low NHE3 antigen levels.34 NHE8 expression is higher in young animals and lower in adults, suggesting a potential role for NHE8 during early development.33 The clear presence of apical NHE activity in the double NHE2 and NHE3 knock-out mice may be a reflection of NHE8.35 NHE8 function may be similar to NHE3 (Fig 4), but its precise role in the maintenance of acid-base balance remains to be explored. NHE8 protein expression is not increased in NHE3 knockout mice,33 suggesting that it may not be regulated by acid-base status the way NHE3 is.36
Transcripts of the more ubiquitous intracellular NHE isoforms are present in the kidney but their antigenic localization and functional characterization have not been performed.
One remarkable feature of the kidney is to escalate its capacity to excrete net acid in the urine when the organism is confronted with a sustained increment in acid load (or base loss). This adaptation takes the coordinated tripartite form of increased H+ pumping (decreasing urine pH) to trap buffers, increased absolute amount of urinary buffer, and reduction (or elimination) of urinary base, which is primarily bicarbonate and citrate. The immediate response can be a result of kinetic effects such as titration of luminal citrate from the trivalent to divalent form to increase citrate absorption,37 and titration of divalent to monovalent phosphate to decrease phosphate absorption38 in the proximal tubule as a result of a decrease in plasma and filtrate pH. This triggers instantaneous reduction of base equivalents in the urine (hypocitraturia)38 and increased urinary buffer (hyperphosphaturia)39 to carry H+. In addition, acute acid loads in whole animals and in model epithelia can alter H+-ATPase and Cl−/HCO3− exchanger distribution to increase H+ pumping and decrease HCO3− secretion into the distal tubular lumen.40,41 These kinetic and protein trafficking events are quick in onset and rapidly reversible. More sustained acid loads elicit a more permanent adaptation of the tubular epithelium usually involving gene transcription. For example, the suppression of proximal phosphate and enhanced citrate absorption occur by different mechanisms with chronic metabolic acidosis compared with acute decreases of luminal pH.42,43 There are a multitude of genes and gene products that are regulated by chronic systemic metabolic acidosis in the kidney. Table 2 summarizes some of the studies of adaptation of renal metabolism and transport in chronic metabolic acidosis and aims to highlight rather than provide an exhaustive catalogue. The integrated response of increased H+-secretion results in the following: (1) increased absorption and metabolism of potential base in urine such as citrate; (2) increased titration of luminal HCO3− to enhance lumen-to-blood HCO3− flux; (3) increased excretion of low-pK, low-capacity buffers such as phosphate; and (4) increased synthesis and secretion of the high-pK, high-capacity buffer NH3.
A significant amount of work has been devoted to studying how chronic low-ambient pH regulates renal Na+/H+ exchange. NHE1 has been shown to be stimulated by acid in both animals and cell culture but this is unlikely to be related to transepithelial transport of H+ equivalents.44-46 Chronic adaptation of NHE3 has been described in both the proximal tubule (increased NHE3 activity and protein) and TAL (increased NHE3 protein and transcript) in animals given a sufficient acid load to chronically suppress serum [HCO3−].36,47-53 Presently it is unclear whether the signals and mechanisms responsible for increased NHE3 are the same for the proximal tubule and TAL. The subsequent discussion focuses only on proximal tubule NHE3.
Chronic metabolic acidosis in animals decreases proximal HCO3− transport when measured by free-flow micropuncture,54 but increases maximal proximal tubule HCO3− transport capacity measured by in vivo microperfusion when identical luminal HCO3− loads are presented to the tubule from control or acidotic rats.55 This is an unusual example of adaptive increase in reabsorptive capacity in response to a reduced filtered HCO3− load caused mainly by reduced ultrafilterable [HCO3−] and possibly some reduction in glomerular filtration rate. This finding has been touted by some as a paradox. The teleology of this increased tubular transport capacity and increased NHE3 Vmax and protein in the face of reduced HCO3− load is unclear because logic predicts that capacity should change to accommodate needs. It is difficult to fathom why the kidney in its wisdom would react to a diminished load by escalating capacity. It is conceivable that increased proximal tubule HCO3− reabsorption is to compensate for conditions other than metabolic acidosis. In respiratory acidosis, increased proximal HCO3− absorption serves to counter the high plasma CO2 tension. In potassium deficiency, increased proximal HCO3− absorption is needed to minimize distal delivery of HCO3−, which is kaliuretic. Because both hypercapnia and potassium deficiency may signal via proximal intracellular acidification, increased NHE3 Vmax in reaction to decreased intracellular pH may be hardwired into the proximal tubule cell. Could the response of NHE3 to metabolic acidosis be a mere happenstance because cell pH is decreased?
Having indulged in the earlier-described teleologic reverie, an alternative, and in all likelihood more probable (and desirable) option, is that increased NHE3 Vmax actually is adaptive for metabolic acidosis in addition to respiratory acidosis and potassium deficiency. The potential benefits to acid-base balance are summarized in Fig 5. The purpose of increased NHE3 Vmax can be primarily to accommodate the high reserve of ammonia synthesis56 because a significant portion of the NH3 synthesized reacts with H+ inside the cell and is excreted into the lumen by NHE3 as NH4+.57 Although the filtered HCO3− load is reduced in metabolic acidosis, the increased HCO3− transport capacity may serve to counteract the interstitium-to-lumen HCO3− backleak. In addition, the increased H+ transport by NHE3 does not have to engage HCO3− in the lumen. It can contribute to the increase in citrate absorption and decrease in phosphate absorption as part of the integrated response to acid load (Fig 5). The net result is increased excretion and titration of urinary buffers. Because NHE3 is tethered to numerous scaffold proteins,58 one can speculate that there may be special microlocales where NHE3 may be associated physically with various other transporters to facilitate titration of substrates.
The mechanism by which a low ambient pH and HCO3− concentration increases NHE3 has been studied in detail by Preisig and Alpern48,59-69 using both cell culture and animal models. Figure 6 summarizes the cascade based on the current body of data. It is important to note that this is an evolving model. Either an acid load or frank systemic metabolic acidosis (hypobicarbonatemia) leads to a decrease in proximal tubule cell pH, which activates proline-rich tyrosine kinase 2 (Pyk2, a pH sensor).59 This is a direct specific effect because acidic solutions can activate Pyk2, but not the closely related focal adhesion kinase, in a cell-free system.59 Pyk2 then activates c-Src. Both Pyk2 and c-Src activation are required for acid to activate NHE3 activity.59-62 Downstream from c-Src is endothelin-1 (ET-1) production and secretion by the proximal tubule.53,63 The mechanism by which c-Src stimulates ET-1 secretion is not yet known. ET-1 activates the endothelin receptor type B (ETB), which then activates NHE3 exocytosis into the apical membrane.53,64,65
ET-1/ETB mediates acid stimulation of NHE3 involving exocytic insertion of NHE3 into the apical membrane. This process requires an intact cytoskeleton,65 and is associated with NHE3 phosphorylation.64 The functional significance of NHE3 phosphorylation for exocytosis is not known at present. The effect of ET1/ETB on NHE3 required both an increase in cell calcium level and intact tyrosine kinases.66,67 ET-1 stimulation of NHE3 requires an intact C-terminal tail and the consensus sequence KXXXVPKXXXV in the second intracellular loop of the ETB receptor.68 ETA is not involved in ET-1 stimulation of NHE3.64 Acid incubation has been shown to increase NHE3 transcript and total cellular protein in cultured cells. The role of increased NHE3 transcript in the intact kidney is less clear.
In parallel with the Pyk2/c-Src pathway is the acid stimulation of extracellular signal related kinase (ERK) and increased c-Fos/c-Jun expression, which contributes to the increase in ET-1 synthesis via an AP-1 site in the ET-1 promoter.61,69 Although the presence of a second pathway is clear, the identity of this second pH sensor is still elusive.
Renal Na+/H+ exchangers, in particular the NHE3 isoform, are paramount for the maintenance of whole-organism acidbase homeostasis. However, despite the exponentially growing body of data on NHE function and regulation, fundamental questions remain unanswered in areas ranging from molecular structure and detailed mechanism of ion translocation to the integration of different NHE isoforms in the big picture of renal and acid-base physiology. Although the study of NHEs at the cellular and molecular level is undoubtedly of crucial importance, one should not lose sight of the fact that understanding the precise roles of NHEs at the whole-organ and whole-organism level is the ultimate goal of any research in the field.
The authors are grateful to Dr. Daniel Fuster for his help in constructing the phylogenetic tree and endless scintillating discussions over the years.
Supported by grants from the National Institutes of Health (DK48482, DK20543 to O.W.M.), the National Kidney Foundation (to I.A.B.), the Charles and Jane Pak Research Fellowship (to I.A.B.), and the Simmons Family Foundation (to O.W.M.).