2022 Volume 45 Issue 10 Pages 1419-1425
Vacuolar-type H+-ATPase (V-ATPase) was first identified as an electrogenic proton pump that acidifies the lumen of intracellular organelles. Subsequently, it was observed that the proton pump also participates in the acidification of extracellular compartments. V-ATPase plays important roles in a wide range of cell biological processes and physiological functions by generating an acidic pH; therefore, it has attracted much attention not only in basic research but also in pathological and clinical aspects. Emerging evidence indicates that the luminal acidic endocytic organelles and their trafficking may function as important hubs that connect and coordinate various signaling pathways. Various pharmacological analyses have suggested that acidic endocytic organelles are important for the maintenance of cell polarity. Recently, several studies using genetic approaches have revealed the involvement of V-ATPase in the establishment and maintenance of apico-basal polarity. This review provides a brief overview of the relationship between the polarity of epithelial cells and V-ATPase as well as V-ATPase driven luminal acidification.
Epithelia are the most widespread tissues in the animal kingdom. Epithelial tissues are composed of cells that fit closely together side by side and are sometimes held together by specialized junctions. Epithelial tissues are classified according to the shape of cells and number of cell layers formed. Cell shapes can be squamous, cuboidal, or columnar. Epithelial cells have specialized apical and basolateral plasma membrane surfaces that differ in structure and function. Typically, the apical membrane of epithelial cells faces the luminal space or external milieu and is densely equipped with microvilli, wherein bundled F-actin sustains a unique structure. The basal membrane is in contact with the extracellular matrix. A shared feature between all epithelial cell types is the presence of a cell–cell junction at the apical-lateral border of the plasma membranes (Fig. 1A).
A. Epithelial polarity is initiated and maintained mainly by three cell polarity complexes, namely the Crumbs, Par, and Scribble complexes. The Crumbs complex consists of Crb, Pals1, and Patj or its paralog Mupp1; the Par complex consists of Par3, Par6, and aPKC; and the Scribble complex is composed of Scrib, Dlg, and Lgl. Furthermore, PAR-6 is regulated by the small GTPase Cdc42, which dynamically associates with the PAR complex.5–10) B. The V-ATPase complex is composed of a peripheral domain (V1) responsible for ATP hydrolysis and an integral domain (V0) involved in proton translocation across the membrane. Uppercase letters indicate the subunits of the V1 sector and lowercase letters represent the V0-sector subunits. Mammalian V-ATPase contains accessory protein Ac45 and full-length Atp6ap2 protein, both of which are essential for enzyme function.
The apical-basal polarity of epithelial cells is required for their fundamental physiological functions, including secretion, selective absorption, protection, transcellular transport, and sensing. Morphogenesis of the polarity of epithelial cells requires various cellular processes, including asymmetric organization of cytoskeletons, polarized transport of proteins, and variation in lipid composition.
Polarity arises from the asymmetric distribution of proteins in cells. The mechanisms for establishing cell polarity and regulatory circuits underlying these mechanisms have been extensively studied over the past several decades (for review see1–5)). Genetic approaches using Drosophila melanogaster and Caenorhabditis elegans have been instrumental in identifying genes that regulate cell polarity and epithelial morphogenesis. The establishment and maintenance of epithelial cell polarization is mainly controlled by the following three polarity complexes: Crumbs,6) Par/aPKC (atypical protein kinase C, PKCζ, and ι in mammals),7) and Scribble8) (Fig. 1A). These complexes localize at distinct epithelial membrane domains and function in either a cooperative or antagonistic manner to induce cellular asymmetry as well as to establish apical-basal polarity. In addition, the Rho family small guanosine triphosphate (GTP)-binding protein CDC42 is also considered to be a part of the Par/aPKC complex and plays key roles in regulating the assembly of polarity. Subsequent studies on various organisms have revealed that these polarity complexes and their roles in establishing cell polarity are evolutionarily conserved in various species.9,10)
Various studies using flies, nematodes, and mammalian epithelial cell cultures have suggested that regulated endocytic trafficking plays a fundamental role in establishing the cellular polarity.11) Transcytosis, a highly regulated coupling of endocytosis, cargo sorting, and exocytosis, is a prerequisite for establishing and maintaining apical-basal polarity.12) In the case of transcytosis from the basolateral to apical cell surfaces, endocytosed molecules pass through a series of endosomal compartments where some molecules are sorted and sent back to the basolateral surface, while others are transported to the apical cell surface (for review, see13)). One of the common features of the endocytic and secretory compartments is that their interior is kept acidic by an inward flow of protons and anions from the cytosol. Acidification is mediated by an active proton pump known as vacuolar-type ATPase (V-ATPase) and an array of transporters conferring anion conductance to the organellar membrane.14,15) V-ATPase is a multi-subunit enzyme that uses energy from ATP hydrolysis to transport protons across membranes (Fig. 1B). It consists of two major functional domains, namely, V1 and V0. The former has eight different subunits (from A to H) and contains three catalytic sites formed by the A and B subunits for ATP hydrolysis. The membrane-bound V0 domain is responsible for proton translocation across membranes. The V0 domain contains up to six subunits, including a, d, and e, and the proteolipids c, c′, and c″, as well as accessory proteins Atp6ap1/Ac45 and Atp6ap2/(pro)renin receptor.16–18) Proteolipids c and c′ are small, four-pass transmembrane proteins with both their termini in the organelle lumen. In mammals, most of the subunits are encoded by multiple genetic loci, whereas the c-subunit and Atp6ap2 are encoded by a single locus.17,19) Therefore, the genetic deletion of Atp6v0c, the locus encoding the c-subunit, results in inactivation of all V-ATPase functions in various subcellular compartments, including endosomes and the Golgi apparatus, as well as the plasma membrane.20)
The establishment and/or maintenance of cell polarity is highly dependent on proper acidification of intracellular organelles. Several strategies have been applied to intervene in intracellular acidification as follows: 1) Weak amines, such as chloroquine and ammonium chloride, passively cross the organelle membranes as uncharged molecules. Once in the acidic compartments, they are protonated, thus becoming charged molecules and are unable to pass across the membranes. This protonation-dependent trapping of weak amines consumes the protons interior and dissipates the proton gradient across organellar membranes. 2) The active transport of protons from the cytosol to the organelle lumen is a prerequisite for organelle acidification. As discussed above, V-ATPase is the major proton pump involved in this process. A class of macrolide reagents, including bafilomycin A1 and concanamycin A, is a highly potent and specific inhibitor of V-ATPase.21) These inhibitors covalently bind to the V0 sector of the intrinsic membrane region, thus inhibiting proton translocation. Another class of V-ATPase inhibitors, archazolid A and B, which are produced by Myxobacterium, are also useful for specific inhibition of V-ATPase.22) In addition to these pharmacological strategies, 3) molecular genetic manipulations of the V-ATPase subunit as well as other components have also been used. As discussed above, most V-ATPase subunits have isoforms that show specific distribution at tissue and subcellular levels.23) In contrast to pharmacological approaches, one can evaluate the effects of isoform-specific perturbations by intervening in expression through RNA interference (RNAi) and/or gene knockout technologies, because each isoform is encoded by distinct loci.
Blocking of V-ATPase functions using V-ATPase-specific inhibitors has revealed inhibition of vesicle transport, including coated vesicle formation24) and lysosomal degradation.25) Numerous studies have shown that the luminal acidification of endocytic compartments is essential for proper vesicle trafficking at various stages. Because subcellular trafficking of membrane vesicles is fundamental for distributing cell surface molecules, perturbation of this process results in the misplacement of apical and basolateral components.
4.1. Basolateral Residential ProteinsVesicular stomatitis virus G protein (VSV-G) is a transmembrane protein that comprises the viral envelope. VSV-G is targeted mainly to the basolateral plasma membrane when it is synthesized de novo in a polarized cell such as Madin–Darby canine kidney (MDCK) cells. VSV-G can be “implanted” into the apical cell surface by facilitating viral particle infection from the apical side of the epithelium.26) The VSV-G in the apical cell surface is then endocytosed and re-localized to the basolateral surface. In the presence of deacidifying reagents, such as monensin and ammonium chloride, the apical VSV-G is internalized, but fails to re-localize to the basolateral cell surface; instead, it accumulates in intracellular compartments.27) In MDCK cells, the polarized transport in the opposite direction, i.e., from basolateral-to-apical transcytosis, however, is less sensitive,28) or even stimulated by monensin and/or methylamine.29)
E-Cadherin is a calcium-dependent cell adhesion molecule that is localized to the lateral cell surface in epithelial cells, bridging each cell tightly and thus maintaining the epithelial integrity. Despite this rather static structural demand, E-cadherin is actively endocytosed and recycled back to the cell surface.30) In the presence of bafilomycin A1, the internalized E-cadherin is not recycled back to the lateral membrane but accumulates in the intracellular compartments.30) Organic anion-transporting polypeptide C (OATPC) and sodium taurocholate transporter are proteins specifically localized to the basolateral plasma membrane when expressed in MDCK cells. In the presence of the V-ATPase inhibitor, they are trapped in the intracellular compartments similar to E-cadherin.31) In a different cell line, LLC-PK1 shows an accumulation of a basal resident aquaporin-2 in the intracellular compartments upon bafilomycin A1 treatment.32) These observations demonstrate that recycling of endocytosed basolateral residents back to the correct cell surface, which is an essential process required for the maintenance of the basolateral localization of residential proteins, is highly dependent on luminal acidification of intracellular compartments.
4.2. Apical Residential ProteinsThe apical surface of cells in the highly developed epithelium is rich in microvilli, which ensures the large membrane areas required for efficient uptake of solutes via specific transporters as well as uptake of macromolecules via endocytosis. V-ATPase itself is highly enriched in the apical surface of the renal epithelium,33,34) where it participates in systemic acid/base homeostasis by secreting protons from the circulation. The newly synthesized proteins destined for the apical surface are sorted from those with basolateral or lysosomal fates in the trans-Golgi network (TGN). Basolateral or lysosomal proteins, in general, have a specific targeting motif that is recognized by cytosolic adaptor proteins. Like ligand (targeting motif) and receptor (adaptors) binding, the assembly of membrane vesicles is initiated, which bud from the TGN, travel across the cytosol, and then fuse with the basolateral plasma membranes. However, unlike the basolateral sorting of the newly synthesized molecules, which are accomplished by specific interactions between the targeting motif and cytosolic sorting molecules (adaptors), apical sorting seems to lack such adaptors, and is facilitated by a different mechanism.35) Multiple mechanisms are involved in sorting apical resident proteins along the biosynthetic pathways. For instance, oligosaccharide moieties of O-glycosylated proteins or GPI-anchored proteins have been shown to be the determinant for apical destination. TGN is equipped with V-ATPase with a2 subunit isoform,36) thus keeping its lumen acidic. Such acidic condition is known to evoke the clustering of the glycosylated polymers by nullifying negative charges on the surface. Oligosaccharide-mediated clustering is proposed to be a possible mechanism for the specific separation of apical proteins in the lumen of the TGN. In fact, de-acidification of the Golgi compartments results in defective sorting of apical membrane proteins.37)
Unlike residential proteins, cell surface receptors that are internalized along with various ligands are more sensitive to de-acidification due to their proper subcellular and cell surface distributions. They show different dependencies on endosomal acidification. MDCK cells express several receptors. The transferrin receptor (TfR) binds iron-loaded transferrin (Tf) on the basal cell surface; the Tf-TfR complex is then internalized and transported to endosomes, where the acidic environment induces dissociation of Tf and TfR. In a study using lysosomotropic agents, both ammonium chloride and monensin suppressed transcytosis and recycling as well as the degradation of epidermal growth factor (EGF), whereas both chloroquine and bafilomycin A1 reduced the degradation process with only a minimal effect on transcytosis.38)
4.3. V-ATPase and Cell PolarityThe close concomitant of organellar acidification and cell polarity suggests a possible physical association of the two machineries, i.e., the acidification complex (V-ATPase and other ion transporters) and Par-aPKC system. The first evidence implying the molecular interactions between V-ATPase and epithelial polarity regulators was shown using a conditional knockout of mouse Atp6ap2, encoding an accessory subunit of V-ATPase in the photoreceptor.39) Loss of Atp6ap2 in photoreceptor cells leads to mislocalization of cell adhesion- and polarity-related molecules in retinal pigmented epithelial cells. Atp6ap2 interacts with PAR3 and intracellular localization of the Atp6ap2–PAR3 complex in the developing retina was observed in the ER-Golgi network and apical edge of the retina.39)
Recently, a forward genetic screening using zebrafish has identified regulators of apical membrane biogenesis. One of these candidates is the atp6ap1b gene, which encodes the homologue of ATP6AP1 (Ac45), an accessory protein subunit of the V-ATPase complex that is required for luminal acidification of intracellular compartments.40) Interestingly, loss of atp6ap1b function caused defects in apical but not basolateral membrane protein delivery.37) The phenotype was recapitulated by bafilomycin A1 treatment for over 16 h. However, the mutant phenotype did not match precisely with that observed upon monensin treatment or bafilomycin A1 treatment for a shorter period. Acute inhibition of V-ATPase function affects not only the sorting of newly synthesized apical proteins in the epithelial cells of the zebrafish intestine, but also causes intracellular accumulation of basolateral membrane proteins resulting from impaired recycling. The variance in phenotypes of knockout mutants and acute inhibition of V-ATPase indicates the differential requirement of luminal acidification for sorting and trafficking of membrane proteins in epithelial cells.
A similar approach combining RNAi screen with in vivo imaging in the C. elegans intestine has proposed a unique function of V0 sector of V-ATPase in apical trafficking and epithelial polarity maintenance.41) Knockdown of the V0 sector of V-ATPase subunits induced a basolateral localization of both the PAR polarity module and brush border components, whereas no basolateral mislocalization was observed upon silencing of the V1 sector of V-ATPase subunits. The precise mechanism by which V0 sector of V-ATPase controls the maintenance of intestinal polarity in an acidification-independent manner remains to be elucidated.
In mammals, the c proteolipid subunit is encoded by a single gene, Atp6v0c. Targeting Atp6v0c causes complete loss of V-ATPase function in all cells in the entire body. Mutant embryos lacking proteolipid c are implanted in the uterine epithelium but die shortly thereafter.20) Analyses of the embryos at E5.5, a developmental stage after implantation, revealed that Atp6v0c deletion results in severe defects in the visceral endoderm (VE) of early embryos. Prior to the initiation of gastrulation, the mouse embryo is a cup-shaped structure comprising the VE and epiblast. The VE exhibits clear polarity; its apical surface faces the maternal circulation and basal side faces the embryo proper. This highly polarized epithelial tissue plays important nutritional roles and influences the development of other embryonic tissues.1,42,43) Atp6v0c deficiency results in the loss of apical-basolateral organization in the VE around E5.5. The distribution of markers of apical and basolateral membranes, including PKCζ, ezrin, β-catenin, and E-cadherin, is severely disturbed. Na+/K+-ATPase, a key epithelial transporter sorted to the basolateral membrane, exhibits apical localization in wild-type VE, similar to that reported in the retinal pigmented epithelium and choroid plexus.5) In Atp6v0c-mutant embryos, however, Na+/K+-ATPase accumulates on the entire cell surface of the VE cells located on the outer side of the embryos.
In addition, the luminal acidification of endocytic compartments is defective, and the endocytic pathway in mutant embryos is significantly affected.44) Rab4 and Rab11, which are involved in recycling pathways in polarized epithelial trafficking, exhibit smeared cytosolic distribution, suggesting that luminal acidic pH is required for the recruitment of Rab proteins to the membranes of endocytic compartments. The role of Rab proteins in polarized epithelial trafficking has been investigated in cultured cells (in vitro) or animals with genetic modification (in vivo studies).5) It is noteworthy that the phenotypes observed are frequently discrepant due to the depletion methods for the gene product. Rab8-KO mice show a defect in apical trafficking only in their intestinal epithelial cells,45) and RNAi-mediated Rab8 knockdown in MDCK cells results in abnormal apical traffic and lumen formation.46) However, knockout of rab8 gene using Cas9-mediated genome editing in MDCK cells does not recapitulate this phenotype.47) The differences in phenotypes caused by acute or long-term depletion may implicate the intricate mechanisms of gene compensation, tissue-specific expression of effectors, and undefined factors awaiting further investigation.
Phosphoinositides (PIs) and other lipids play important roles in the generation of epithelial polarity.48) PI(4,5)P2 (PIP2) is a key determinant of the apical surface, while PI(3,4,5)P3 (PIP3) is typically observed in the basolateral membrane. PIP3 may be converted into PIP2 through the action of PTEN, a phosphatase that is enriched in the apical membrane.49) Meanwhile, the distinct distribution of PIPs is commonly considered a hallmark of organelles.
Surface delivery of endosomal cargo requires loss of endosomal identity by hydrolysis of PI(3)P mediated by PI 3-phosphatase. PI(3)P turnover during endosomal exocytosis is accompanied by PI4-kinase II (PI4KII)-dependent generation of PI(4)P along with Rab conversion from Rab5 to Rab11, a GTPase switch required for endosomal recycling.50) Loss of PI(3)P is also considered a hallmark of endosome maturation. PI(3)P signal termination is controlled by the luminal acidification of endosomes. PI(3)P associated with endosomes is synthesized by Vps34, a class III phosphatidylinositol-3-kinase. Preventing acidification of the endocytic lumen with weak bases or vacuolar ATPase inhibitors causes PI(3)P to persist even longer, whereas forced acidification drives PI(3)P depletion.51) The Vps34 complex associates with the membrane in a pH-dependent manner, and the pH gradient across the membrane serves as a trigger to displace Vps34 from membranes.51) How luminal pH is sensed by the Vps34 complex leading to dissociation as well as the direct involvement of V-ATPase remain unclear.
The function of V-ATPase, on the other hand, appears to be closely related to the level of PIPs. Genetic studies in combination with biochemical analyses have revealed that PIPs in organelle membranes influence organelle pH directly as well as indirectly. In yeast, it has been observed that PIPs interact directly with the subunit of V-ATPase, and PIP levels affect the assembly and activity of V-ATPase.52) In contrast, inhibition of V-ATPase with bafilomycin A1 attenuates elevated PI(3)P levels when the phosphoinositide kinase converting PI(3)P to PI(3,5)P2, PIKfyve, is inhibited.53)
In addition to PIs, glycosphingolipids and cholesterol also play important roles in apical transport by forming lipid rafts and are involved in polarity establishment. The raft components can be utilized for segregating the apical cargos from basolateral cargos and for the generation of intracellular transport carriers.54) V-ATPase has been identified as a major component of lipid rafts.55,56) Lipid raft disruption by cholesterol depletion from cell membranes with methyl-β-cyclodextrin leads to a significant increase in the late endosomal pH.55) U18666A, a drug that leads to the accumulation of cholesterol in late endosomes, increases the association of the V1 sector with V0 of V-ATPase, while decreasing the luminal acidity of late endosomes.55) The mechanism underlying the paradoxical effects of U18666A still remains to be understood. It is possible that the excessive stabilization of V1–V0 complexes may be detrimental to the activity of the proton pump.
Various lines of studies have shown that elevating endosomal pH slows the receptor externalization rate, but does not affect receptor internalization kinetics.25,57) However, an approach using small interfering RNA (siRNA)-based screening to identify regulators of clathrin-coated vesicle formation in endocytosis has identified subunits of V-ATPase. The knockdown of V-ATPase blocks clathrin-mediated endocytosis, and this disruption in internalization can be reproduced by inhibition of V-ATPase with bafilomycin A1 for 24 h. V-ATPase depletion prevents recycling of cholesterol from endosomes back to the plasma membrane. This phenotype is reminiscent of Niemann–Pick type C disease (NPC), a lysosomal storage disease characterized by the accumulation of cholesterol, sphingomyelin, and other lipids in endosomes and lysosomes. The disease is associated with NPC1 and NPC2 genes. NPC1 encodes a polytopic membrane protein that is located in the membranes of endosomes and lysosomes, and NPC2 is a small, soluble protein present in the lumen of late endosomes and lysosomes.58) Structure-based analysis has revealed that the acidic luminal pH of lysosomes is essential for cholesterol transfer from NPC2 to NPC1.58) The retention of cholesterol in non-acidified endosomes in V-ATPase-depleted cells causes its concomitant loss from the plasma membrane. Taken together, these data highlight that endosome acidification is clearly important for the recycling of cholesterol, which is required for establishing or maintaining the cell polarity.
Cell polarity is established and maintained by highly controlled vesicular trafficking among the distinctive domains of the cell surface membrane and intracellular membrane compartments. Luminal acidification by V-ATPase plays regulatory roles in the trafficking of various elementary processes, including coated vesicle formation, targeting, and fusion. Therefore, cell polarity and luminal acidification are intimately coupled as discussed above.
Various studies have shown that the effects of dysfunction of this proton pump on newly synthesized proteins and membrane residential proteins are remarkably different. However, our understanding of these diverse sensitivities is limited. It is well known that the acidity of each compartment varies along the endocytic pathway: the Golgi is less acidified, endosomes are mildly acidic, and lysosomes are highly acidic. This diversity can be explained, at least in part, by the increase in the ratio of membrane-associated V1/V0 varies along the endocytic pathway. The relative abundance of V1 is higher in late endosomes than that in early endosomes, which provides an explanation for the higher acidity of late endosomes.55) Since intracellular organelles exhibit distinct luminal pH,52) the range of pH required for organelle functions may vary. Most recently, Maxson et al. generated a fluorescent probe derived from SidK, an effector protein from Legionella pneumophila. Using this probe, they attempted to quantify and estimate the number of V-ATPase complexes in each organelle.59) It is observed that the disturbance at the step of vesicular traffic is dependent on not only the concentration of the proton pump inhibitors but also the period of the treatment.24,25,60) However, the precise underlying mechanism still needs further investigation. Because most observations are based on the general inhibition of acidification by either pharmacological or genetic means, the acidification of specific compartments cannot be assessed. Further studies are required to explore the molecular relationship between luminal pH and regulatory proteins of vesicle trafficking. In this regard, organelle-specific components, such as lipid composition and subunit isoforms of the V-ATPase complex, would prove to be key components in deciphering the puzzle of cell polarity and acidification. We expect that these molecules would be targets of pharmacogenetics of various pathologies related to epithelial polarity.
This work was supported by Grants-in-Aid from the MEXT/JSPS Japan: Grant Nos. JP19K05950 and JP20H05323 to YW, and JP21K06553 to GHSW, and by Individual Research Grants in Doshisha Women’s College of Liberal Arts to GHSW.
GHSW prepared figures. GHSW and YW wrote and edited the review. Both authors contributed to the article and approved the submitted version.
The authors declare no conflict of interest.
