H⁺V-ATPase-Dependent Luminal Acidification in the Kidney Collecting Duct and the Epididymis/Vas Deferens: Vesicle Recycling and Transcytotic Pathways

Several vertebrate transporting epithelia contain a population of specialized cells that transport protons or bicarbonate across the epithelial barrier (Al-Awqati, 1996; Brown and Breton, 1996; Madsen and Tisher, 1986; Schuster, 1993). These cells are often referred to as ‘mitochondria-rich’ (MR) cells (Brown and Breton, 1996), because they contain large numbers of mitochondria. They also share several characteristic features, including high levels of cytosolic carbonic anhydrase and membrane-associated H⁺V-ATPase. In the kidney, these cells are located in the collecting duct and connecting segment of the urinary tubule and are known as intercalated cells. The epididymis and vas deferens also contain similar, but not identical, epithelial cells that express high levels of H⁺V-ATPase on their luminal plasma membranes as well as in intracellular vesicles (Brown et al., 1997). This review will examine selected aspects of the cell biology of proton-transporting cell types that are related to the regulation of transepithelial acid–base transport in these two distinct organ systems.

At least two subtypes of intercalated cell are present in the kidney cortex, so-called A (or alpha)-cells and B (or beta)-cells. These subtypes are involved in proton secretion and bicarbonate secretion, respectively, in the distal nephron and collecting duct (Al-Awqati, 1996; Brown and Breton, 1996; Madsen and Tisher, 1986; Schuster, 1993). A-cells have an H⁺V-ATPase on their apical plasma membrane and on intracellular vesicles, and they express the kidney homolog of the band 3 Cl⁻/HCO₃⁻ exchanger (AE1) on their basolateral plasma membrane (Alper et al., 1989; Drenckhahn et al., 1985; Schuster et al., 1991) (Fig. 1). B-cells contain no detectable AE1, but the cellular location of the H⁺V-ATPase can be apical, basolateral, bipolar or diffuse (Alper et al., 1989; Brown et al., 1988a,b; Drenckhahn et al., 1985; Schuster et al., 1991). This complex array of phenotypes is probably at least partially responsible for the different functional categories of intercalated cells (including so-called gamma-cells) that have been identified using intracellular fluorescence pH measurements and ion-substitution experiments (Emmons and Kurtz, 1994; Weiner and Hamm, 1990). In the kidney medulla, only A-type intercalated cells are found, and they account for approximately 40 % of the epithelial cell population in the outer medulla (Brown et al., 1988b). This number falls to approximately 10 % in the initial portion of the inner medulla, and intercalated cells are completely absent from the terminal two-thirds of the papillary collecting duct. A- and B-cells together also make up approximately 40 % of the total number of cells in the cortical collecting duct, with approximately equal numbers of each subtype being present in rat kidney. This differentiated cellular profile in various regions of the kidney corresponds to functional observations; cortical collecting ducts are capable of either net proton or bicarbonate secretion (McKinney and Burg, 1977), whereas medullary collecting ducts can secrete only net acid (Lombard et al., 1983). All intercalated cells contain a large amount of cytoplasmic carbonic anhydrase (CAII) (Brown et al., 1983; Kim et al., 1990), which is also a key protein in supporting the proton-transporting capacity of these cells by providing membrane-associated pumps and transporters with a supply of protons and bicarbonate (Fig. 2).

The luminal fluid along much of the male reproductive tract is maintained at an acidic pH (Carr et al., 1985; Levine and Marsh, 1971; Rodriguez et al., 1990). This low pH is involved in sperm maturation and in maintaining sperm in an immotile state during their passage through the epididymis and vas deferens, probably in conjunction with factors including specific proteins, weak acids and other ions (Hinton and Palladino, 1995). The rat epididymis and vas deferens contain a distinct population of epithelial cells that express high levels of H⁺V-ATPase on their luminal plasma membranes as well as on intracellular vesicles (Brown et al., 1992). We have shown that H⁺V-ATPases and the cytoplasmic enzyme CAII are co-localized in the same cell type (Breton et al., 1996, 1999). These cells are the so-called ‘apical’ cells or ‘narrow’ cells in the head (caput) portion of the epididymis and the ‘light’ cells or ‘clear’ cells in the body (corpus) and tail (cauda) (Reid and Cleland, 1957; Sun and Flickinger, 1979).

In collaboration with Dr Peter Smith at the BioCurrents Research Center, Marine Biology Laboratory, Woods Hole, MA, USA, a proton-selective, non-invasive microelectrode was used to detect apical proton secretion in the isolated vas deferens. We showed that the proton flux is inhibited rapidly by the H⁺V-ATPase inhibitor bafilomycin (Breton et al., 1996) and more slowly by the carbonic anhydrase inhibitor acetazolamide (Breton et al., 1998a) (Fig. 3). These results indicate that H⁺V-ATPase-driven proton transport is a major source of luminal acidification in the vas deferens, although other acid–base transporters, including an Na⁺/H⁺ exchanger (Au and Wong, 1980) and a Na⁺/HCO₃⁻ cotransporter, may play an important role in more proximal segments of the epididymis (Jensen et al., 1999a). In rat epididymis, the proton-secreting cells do not stain with anti-AE1 antibodies, in contrast to kidney A-type intercalated cells. However, all epithelial cells in the epididymis contain AE2, a different isoform of the anion exchanger, on their basolateral plasma membranes (Jensen et al., 1999b). Its contribution to transepithelial proton secretion remains to be determined.

In the kidney and the reproductive tract, cells rich in proton pumps display a high rate of apical membrane recycling. This recycling can be readily detected by following the internalization of fluid-phase markers, such as fluorescent dextrans, introduced into the tubule lumen (Brown and Breton, 1996; Brown and Sabolic, 1993; Schwartz et al., 1985). The recycling process involves the transport of H⁺V-ATPase molecules between the cell surface and the cytoplasm in highly specialized intracellular vesicles that have a well-defined ‘coat’ structure formed of the peripheral, cytoplasmic V₁ subunits of the H⁺V-ATPase (Brown et al., 1987a). In the kidney, systemic acidosis or basolateral perfusion of isolated collecting tubules with high-[CO₂] solutions causes vesicle exocytosis and results in the delivery of more pumps to the apical plasma membrane and an increase in H⁺ secretion (Schwartz et al., 1985). Upon reversal of the stimulus, pumps are removed from the membrane by endocytosis. A similar recycling process also occurs in the epididymis (Breton et al., 1998c), but the physiological stimulus that controls this process has not yet been identified. In contrast to the kidney, systemic acidosis does not cause an increase in apical plasma membrane H⁺V-ATPase density in the epididymis (S. Breton, L. J. Jensen and D. Brown, unpublished results).

Traditionally, endocytosis was thought to be mediated either by clathrin-coated vesicles or by ‘smooth’ pinocytotic vesicles. Clathrin is involved in the endocytosis of many receptors as well as some other membrane proteins including the facilitated glucose transporter GLUT4 (Holman and Cushman, 1994) and aquaporin 2 (AQP2) (Brown, 1989; Van Hoek et al., 1998). Endocytosis can also occur via a clathrin-independent pathway: these smooth vesicles or caveolae are decorated with a protein called VIP21 or caveolin, which is a substrate for the v-src tyrosine kinase in rous-sarcoma-virus-transformed fibroblasts (Glenney and Soppet, 1992). We have shown that endocytosis of the H⁺V-ATPase occurs in intercalated cells via a clathrin-independent mechanism (Brown and Orci, 1986; Brown et al., 1987b), yet caveolin cannot be detected in these endocytic vesicles (Breton et al., 1998b). Finally, β-COP, another vesicle coat protein associated with vesicle trafficking between the Golgi and the rough endoplasmic reticulum, could not be detected on proton-pump transporting vesicles in the kidney or the epididymis (Breton et al., 1997). It is possible, therefore, that some subunits of the H⁺V-ATPase are, in some as yet undefined way, involved in the endocytotic mechanism in these cells.

Despite considerable effort by many investigators, interconversion of A- and B-type intercalated cells in vivo has not been demonstrated directly. The pattern of proton-pump staining clearly changes in the cortical intercalated cell population after acute and chronic acid–base disturbances, with more cells showing tight apical staining in acid conditions, and more showing basolateral staining during alkalosis (Bastani et al., 1991; Sabolic et al., 1997). However, the data can be equally well explained by rearrangements of proton pump polarity in distinct populations of A- and B-cells, without the need to invoke interconversion between these two cell types. This conclusion stands in contrast to data from in vitro studies, where an extracellular matrix protein named hensin has been proposed to be a regulator of the intercalated cell phenotype on the basis of its ability to reverse the functional polarity of intercalated cells in culture (Al-Awqati et al., 1998; Vijayakumar et al., 1999).

In vivo, as mentioned above, cells with basolateral AE1 always have apical or subapical H⁺V-ATPases, but AE1-negative cells, classified as B-cells, can also have apical H⁺V-ATPase (Alper et al., 1989). Basolaterally located H⁺V-ATPases have never been found in collecting ducts in the inner stripe of the outer medulla, where only A-cells are present. This distribution may reflect the inability of A-cells in this region to insert proton pumps into the basolateral membrane domain, suggesting that they do indeed represent a distinct cell type. However, the tubular and extracellular environment of the medulla might be such that the A-cell phenotype is favored in this kidney region. Future work addressed at modulating this environment to render it more similar to that of the cortex might provide some answers to the issue of whether A-cells can indeed convert to B-cells under appropriate conditions. These issues have been discussed in previous reviews, to which the reader is referred (Al-Awqati, 1996; Al-Awqati et al., 1998; Brown and Breton, 1996; Schuster, 1993).

Type B intercalated cells can insert H⁺V-ATPases either apically or basolaterally, and some even have a bipolar distribution of H⁺V-ATPase (Alper et al., 1989; Bastani et al., 1991; Brown et al., 1988a,b). Our working hypothesis is that B-cells have an active transcytotic pathway analogous to that present in intestinal epithelial cells, e.g. where the polymeric immunoglobulin (Ig) receptor is first inserted into the basolateral plasma membrane and subsequently re-internalized and transported to the apical plasma membrane after ligand binding. A 14 amino acid domain at the C terminus of the receptor determines initial basolateral targeting, and phosphorylation of Ser664 stimulates endocytosis and transcytosis (Casanova et al., 1990). The transcytotic vesicles fuse with an apical recycling endosomal compartment before the Ig receptor is resorted and finally inserted into the apical membrane (Apodaca et al., 1994). Hepatocytes also have an especially active transcytotic pathway (Bartles and Hubbard, 1990). Microtubules (Hunziker et al., 1990) and GTP-binding proteins (Bomsel and Mostov, 1993) have been implicated in the process of transcytosis, and brefeldin A also inhibits transcytosis (Hunziker et al., 1991).

Some data from our laboratory suggest that A-cells and B-cells have different intracellular trafficking pathways for the lysosomal glycoprotein LGP120 that may be relevant to the ability of B-cells (but not A-cells) to shuttle H⁺V-ATPases between apical and basolateral membrane domains by transcytosis. Using antibodies against LGP120 (provided by Dr Ira Mellman, Yale University), B-intercalated cells showed variable levels of this lysosomal membrane protein on their basolateral plasma membranes, as well as in lysosomes (Fig. 4). The distribution of LGP120 on the plasma membrane was similar to that of the H⁺V-ATPase in each individual B-cell. In A-cells, however, LGP120 was detectable only in lysosomes and not at the cell surface. This observation suggests that B-cells may insert newly synthesized LGP120 first into their basolateral plasma membrane, before re-targeting the protein to lysosomes. This indirect pathway of lysosomal membrane protein trafficking to lysosomes via the cell surface has been previously reported in other cell types (Gough and Fambrough, 1997; Nabi et al., 1991). However, other studies have shown that this transient cell surface expression is not necessary for lysosomal targeting (Harter and Mellman, 1992). Indeed, LGP120 may be directly targeted to lysosomes in A-cells, with no detectable accumulation at the cell surface, at least by immunocytochemical analysis.

Because re-internalization and subsequent transcytosis of some proteins from the basolateral membrane of epithelial cells depends on an intact microtubular network, the distribution of LGP120 in cells from colchicine-treated rats was examined. Rats were injected with colchicine for 8–12 h as previously described (Brown et al., 1991; Gutmann et al., 1989), and fixed kidneys were examined by immunofluorescence microscopy. In B-cells, significantly more LGP120 was located on the basolateral plasma membrane, as assessed by quantitative fluorescence imaging (Fig. 5) indicating that, in the absence of microtubules, LGP120 was delivered to the basolateral cell surface but was not re-internalized and directed towards lysosomes. Consistent with this idea, considerably less intracellular lysosomal staining for LGP120 was seen in B-cells from colchicine-treated rats (Fig. 6).

A schematic representation of the proposed trafficking pathways for LGP120 in A- and B-type intercalated cells is shown in Fig. 7. Although this cell-specific difference in protein trafficking in these two cell types remains to be confirmed for acid–base transporters, it is supported by observations from Gandhi et al. (1990), showing that another ‘apical’ marker, ecto 5′ nucleotidase, is also detectable on the basolateral plasma membrane of B-cells. Thus, we propose that the polarity reversal of the H⁺V-ATPase, at least in B-type intercalated cells, is mediated by a physiologically regulated transcytotic pathway that may be analogous to that existing in other cell types.

Physiologically modulated recycling of proton pumps in acid-transporting cells is important for the normal function of many tissues and organ systems in mammals and many other vertebrate and non-vertebrate species (Brown and Breton, 1996; Wieczorek et al., 1999). It will be important to determine the protein composition of the specialized transporting vesicles that are involved in this process to understand the mechanisms involved in exo- and endocytosis of these novel ‘coated’ vesicles, which contain none of the other so-called vesicle ‘coat’ proteins, clathrin, caveolin or β-COP (Breton et al., 1997, 1998b). In contrast, cellubrevin (McMahon et al., 1993) and related vesicle fusion proteins that form a protein complex necessary for vesicle fusion to target membranes seem to be involved in proton pump exocytosis in the epididymis (Breton et al., 1998c) and in cultured renal collecting duct epithelial cells (Alexander et al., 1997). Thus, the potential role of synaptic-vesicle-like fusion proteins in exocytosis of proton pumps should be explored in more detail.

The opposite polarity of membrane insertion of proton pumps in subpopulations of intercalated cells remains a fascinating biological observation that has not so far been explained in terms of traditional protein sorting paradigms. However, the 56 kDa, or B1, subunit of the H⁺V-ATPase that is preferentially expressed in renal intercalated cells and in epididymal proton-secreting cells has a C-terminal DTAL motif (Nelson et al., 1992), which suggests that it is a PDZ domain binding protein. PDZ domains are approximately 90-residue repeats found in a number of proteins implicated in the clustering of ion channels, receptors and other proteins in the plasma membrane. They are named after the first letters of the three proteins that were initially found to contain this amino acid domain – a postsynaptic density protein PSD-95, the Drosophila tumor-suppressor protein discs large, and the tight-junction-associated protein ZO-1 (Hall et al., 1998; Wang et al., 1998). Interaction between PDZ-binding proteins and PDZ proteins occurs via specific amino acid sequences on their C terminus including, but not restricted to, the consensus sequence, DS/TXL, as for the B1 H⁺V-ATPase subunit. The possibility that one of the V-H⁺ATPase subunits might associate with these anchoring proteins raises the exciting possibility that proton pumps may be selectively concentrated in some membrane domains via this newly discovered intracellular scaffolding system.

This work was supported by NIH grants DK 42956 (D.B.) and DK 38452 (S.B. and D.B.).