Enigmatic Roles of the Epithelial Sodium Channel (ENaC) in Articular Chondrocytes and Osteoblasts: Mechanotransduction, Sodium Transport or Extracellular Sodium Sensing?

Mobasheri A, Barrett-Jolley R, Shakibaei M, et al.

Publication Details

The mammalian MEC/DEG/ENaC gene superfamily encodes membrane proteins which are involved in diverse functions including acid sensing, maintenance of sodium homeostasis and transduction of mechanical stimuli and nociceptive pain. In the principal cells of the distal nephron amiloride sensitive ENaC activity represents the rate-limiting step for sodium reabsorption. Epithelial sodium channel (ENaC) subunits have also been found in articular chondrocytes and osteoblasts. In this article we discuss the enigmatic roles of ENaC in skeletal cells including articular chondrocytes and osteoblasts and review recent papers in which ENaC has been proposed to participate in skeletal mechanotransduction, sodium transport and extracellular sodium sensing.

Introduction

The amiloride sensitive epithelial sodium channel (ENaC) is a multimeric protein system consisting of three subunits; α, β and γ involved in apical Na+ uptake in a variety of epithelia [4, 5]. ENaC belongs to the MEC/DEG/ENaC superfamily whose members are involved in many diverse functions including acid sensing, maintenance of sodium homeostasis and transduction of mechanical stimuli and nociceptive pain [13]. In the epithelia of the distal renal tubule (Fig. 1) amiloride sensitive epithelial sodium channels represent the rate-limiting step for sodium reabsorption [29]. Under the endocrine influence of the renin-angiotensin-aldosterone system, ENaC is a major participant in the fine regulation of systemic sodium balance, blood volume and blood pressure and also plays a key role in lung fluid clearance in neonates [16]. However, ENaC is also found in cartilage, skin and bone. In these tissues the roles of ENaC are not well understood and so the main focus of this article is a brief discussion and update of ENaC's putative roles in these non-epithelial tissues.

Figure 1. Immunohistochemical localization of the β subunit of ENaC in tissues represented on human tissue microarrays (TMAs) obtained from the Cooperative Human Tissue Network (CHTN) of the National Cancer Institute (NCI), the National Institutes of Health, Bethesda, MD (http://faculty.

Figure 1

Immunohistochemical localization of the β subunit of ENaC in tissues represented on human tissue microarrays (TMAs) obtained from the Cooperative Human Tissue Network (CHTN) of the National Cancer Institute (NCI), the National Institutes of Health, (more...)

ENaC homologues as mechanotransducers in C. elegans

Long before the identification and cloning of the ENaC superfamily gene members in mammalian species a number of genes required for mechanosensitivity had already been identified in the nematode worm genetic model Caenorhabditis elegans. Deletions in the mechanosensitive genes MEC-4 and MEC-10 and the degenerin gene DEG-1 rendered the invertebrates insensitive to touch and prone to neurodegeneration. The remarkable structural similarity between mammalian ENaC members and the C. elegans MEC/DEG genes hinted that MEC and DEG encoded similar cation channels and gave strong support to the hypothesis that mammalian ENaC genes may also be involved in the process of mechanotransduction [34,35,36]. For many years the evidence for this role remained tenuous and indirect. However, recent in vivo imaging studies of C. elegans mechanosensory neurons have demonstrated a specific role for the MEC-4 channel in the process of gentle touch sensation [33]. Furthermore, elegant electrophysiological studies of the DEG/ENaC channels in C. elegans touch receptor neurons have shown that these proteins are capable of transducing mechanical signals [26]. By recording from C. elegans touch receptor neurons in vivo, it has been found that external force evokes rapidly activating mechanoreceptor currents carried mostly by Na+ and blocked by amiloride. Null mutations in the DEG/ENaC gene mec-4 and in the accessory ion channel subunit genes mec-2 and mec-6 eliminate mechanoreceptor currents. These recent findings link the application of external force to the activation of a molecularly defined metazoan sensory transduction channel.

Other degenerin-epithelial sodium channel (DEG-ENaC) family members involved in mechanosensation

Other ion channels in the (DEG-ENaC) family may also be involved in mechanosensation. Recent studies have explored the role of the acid-sensing ion channel 2 (ASIC2), in auditory transduction [27]. Although ASIC2 null mice showed no significant hearing loss, ASIC2 expression was found in spiral ganglion neurons in the adult cochlea and externally applied protons induced amiloride-sensitive sodium currents and action potentials in spiral ganglion neurons in vitro. This new data suggests that ASIC2 may contribute to sensory functions of the cochlea but exactly how is not clear. It is possible that the presence of ASIC2 in spiral ganglion neurons may provide sensors to directly convert local acidosis to excitatory responses.

Candidate mechanotransducers in cartilage and bone cells

Osteoblasts and chondrocytes have been shown to be mechano-sensitive cells. The processes of anabolism and catabolism (regulated primarily by mechanical loading) are a second-to-second, minute-to-minute, and hour-to-hour processes that function together with local and systemic hormones to ensure that the tissue can meet the demands of the mechanical environment. Mechanical strain or impact loading act as mechanostimulants. Osteoblasts for example respond to intermittent mechanical strain by increasing their mitotic rate and production of type I collagen extracellular matrix [15] and stretch-activated cation channel activation [10]. Similarly chondrocytes respond to mechanical loading by decreasing production of extracellular matrix macromolecules such as aggrecan [2]. The observation of a metabolic response to mechanical strain has led a number of groups to hypothesize that following physiological mechanostimulation the processes of mechanoreception and mechanotransduction may involve the cellular cytoskeleton [17, 40] and plasma membrane ion channels [6]. The current published information on mechanosensitive ion channels in skeletal cells is relatively limited. However, it has been established that skeletal cells such as osteoblasts and chondrocytes express a number of ion channels with different biophysical and pharmacological properties (for recent reviews see [11, 22]). In osteoblastic cells Ca2+ channels have been shown to play fundamental roles in cellular responses to external stimuli including both mechanical forces and hormonal signals. Ca2+ channels are also proposed to modulate paracrine signalling between bone forming osteoblasts and bone-resorbing osteoclasts at local sites of bone remodelling. Calcium signals are characterized by transient increases in intracellular Ca2+ levels that are associated with activation of intracellular signalling pathways that control cell behaviour and phenotype, including patterns of gene expression. Development of Ca2+ signals is a tightly regulated cellular process that involves the concerted actions of plasma membrane and intracellular Ca2+ channels, along with Ca2+ pumps and exchangers.

Evidence for ENaC in articular chondrocytes and osteoblasts

Recent studies performed by Killick and Richardson [18] have established that avian chondrocytes express αENaC; Northern blot experiments showed that a 2.5-kb transcript of ENaC is present in avian cartilage. Elegant studies by Kizer and colleagues [19] have also demonstrated expression of αENaC at the transcriptional and translational levels in an established rat osteoblast-like cell line (UMR-106) and primary cultures of human bone marrow stromal cells. Using electrophysiological techniques they also showed that osteoblasts stably transfected with mammalian expression vectors encoding αENaC express a stretch activated, non-selective cation channel that is only active when negative pressure is applied to cell-attached patches. The above studies prompted us to investigate the expression of ENaC subunits in human chondrocytes, osteoblasts and anterior cruciate ligament in vitro and in vivo. Expression of βENaC in human cartilage is shown in panel D of Figure 1. Expression of the α and β subunits of ENaC in human chondrocyte-like cells is shown in figure 2. Using polyclonal antibodies raised against the α, β and γ subunits of ENaC [9] our own collaborative group has obtained evidence for the presence of α and β ENaC subunits in situ in human chondrocytes [38,39] and in a human chondrocyte cell line ([24]; see Fig. 2). Using a molecular approach we have identified α and β ENaC in human femoral bone derived osteoblasts ([23]; Fig. 3).

Figure 2. Expression of the α and β subunits of ENaC in C20/A4 human chondrocyte-like cells.

Figure 2

Expression of the α and β subunits of ENaC in C20/A4 human chondrocyte-like cells. In the cell line the α and β subunits appear to be localized in different subcellular compartments: immunostaining for the α subunit (more...)

Figure 3. RT-PCR evidence for αENaC expression in human osteoblasts and human ACL cells.

Figure 3

RT-PCR evidence for αENaC expression in human osteoblasts and human ACL cells. The 600 base pair PCR products found in human osteoblasts and human anterior cruciate ligament (ACL) cells correspond to αENaC, which was found to be expressed (more...)

In a recent study, we investigated the possible co-localization of ß1-integrins with the α subunit of ENaC in organoid cultures derived from limb-buds of 12-day-old mice [31]. We used indirect immunofluorescence to demonstrate that ß1-integrin and αENaC co-localize in limb bud chondroblasts. Co-imunoprecipitation experiments also revealed that ß1-integrins associate with the α subunit of ENaC (Fig. 4). In light of these observations we have suggested that chondrocyte responses and signalling cascades initiated by the influx of sodium through mechanosensitive ENaC channels may be regulated more effectively if such channels were organized around ß1-integrins with receptors, kinases and cytoskeletal complexes clustered about them [31].

Figure 4. Co-immunoprecipitation assays and immunoblotting of ß1-integrin and the α subunit of ENaC.

Figure 4

Co-immunoprecipitation assays and immunoblotting of ß1-integrin and the α subunit of ENaC. Immunoprecipitation with anti ß1-integrin antibody was followed by immunoblotting with anti-αENaC antibody. The western blot lane (more...)

ENaC expression is therefore not strictly limited to epithelia. Indeed, ENaC expression has also been demonstrated in epidermal keratinocytes which surround touch sensitive hair follicles [30]. All this evidence points to the presence of functional epithelial sodium channels in the plasma membranes of mechanosensitive cells and leads us to hypothesize that ENaC or its homologues may mediate mechanotransduction in skeletal cells. Further support for this hypothesis comes from cloning studies of unc105, another C. elegans homologue that is believed to interact with type IV collagen, providing evidence that mechanostimulation of the extracellular matrix may be transduced into electrochemical signals within the cell. The extracellular domain of the protein encoded by unc-105 probably interacts with collagen on the extracellular face and mechanical forces exerted on the collagen network could open the channel [20].

ENaC expression is altered in osteoarthritis

Remarkably, the expression and distribution of epithelial sodium channels appears to vary in pathologies of human articular cartilage. Although chondrocytes from normal cartilage express α and β ENaC, in osteoarthritis, ENaC seems to be absent altogether. In contrast, in rheumatoid cartilage ENaC expression is upregulated [39]. The significance of these findings is that the altered extracellular matrix and/or the inflammatory response in degenerate cartilage appear to have a direct influence on the expression and abundance of stretch-activated ENaC. Thus, in terms of cellular pathophysiology our results suggest that the expression and abundance of epithelial sodium channels is altered in pathologies of articular cartilage.

Conclusions and perspectives

The α subunit of ENaC shows sequence homology with a group of actin-binding proteins known as dystrophins. The α and β dystrophins are implicated in the ability of muscle cells to contract. Interestingly dystrophins are also members of an actin-binding family including spectrin and fodrin [12]. Studies on renal epithelial cells have demonstrated that the epithelial sodium channel is linked to the cytoskeleton [32]. Thus, if ENaC was also demonstrated to be physically and functionally linked to the cellular cytoskeleton in skeletal cells it would perfectly fulfil the criteria for a mechanosensitive ion channel. This hypothesis of cytoskeletal regulation is partially supported by at least three other ion transport proteins. The catalytic α subunit of the Na+, K+-ATPase is connected to the cytoskeleton via attachments to ankyrin and fodrin [25] and the enzyme itself has been shown to bind actin in a way that its ATP hydrolytic activity is enhanced [7]. Stimulation of the Na+/H+ exchanger by serum is regulated by F-actin [41] and the band-3 anion exchanger AE1 is attached to the cytoskeleton via ankyrin and spectrin in the rat kidney and cultured epithelial cells [8].

In terms of ion channel cell biology these findings raise a number of important and fundamental questions:

  1. Does ENaC function as a stretch-activated ion channel in vivo as it does in vitro? Does ENaC contribute to the expression of a mechanosensitive channel complex in skeletal cells?
  2. Given that Na+ is the most abundant cation in the extracellular matrix of articular cartilage is ENaC specific for Na+ or is it unable to discriminate cations?
  3. Is ENaC attached to the cytoskeleton in skeletal cells? Does the cytoskeleton or the extracellular matrix regulate ENaC?
  4. Is it possible to interfere with ENaC function by using compounds that disassemble the cellular cytoskeleton?
  5. Are there any other members of the ENaC superfamily present in skeletal cells? If so, do they play a role in normal growth and skeletal development?

Whether ENaC functions as a non-selective stretch activated channel or contributes to the expression of a mechanosensitive complex in skeletal cells is a challenging question. However, recent electrophysiological studies in rabbit oesophageal epithelia have provided evidence for non-selective cation transport via apical ENaC channels (non-selective for the following monovalent cations: Li+, Na+, and K+) [1].

In summary, we feel that the discovery of epithelial sodium channels in osteoblasts and chondrocytes has reinvigorated interest in skeletal mechanotransduction. Although the direct involvement of ENaC in transducing mechanical signals in the Xenopus laevis oocyte or planar lipid bilayers remains controversial [28] studies in mechanically responsive cells such as osteoblasts and chondrocytes should provide a more suitable physiological model for studying mechanotransduction. Furthermore, we regard skeletal cells as highly appropriate for studies of mechanostimulation, mechanotransduction and for elucidating the detailed molecular anatomy of interactions between ion channels, the cytoskeleton and extracellular matrix macromolecules.

Finally, and most interestingly, ENaC is also expressed in skin [3, 21]. ENaC expression in keratinocytes is not only related to cell differentiation but is also required for normal epidermal growth. Since the physiological role of ENaC in articular cartilage is still not clear, one may speculate that ENaC could play a role as a mechanotransducer and an extracellular Na+ sensing mechanism that regulates osmotically inactive Na+ storage in the extracellular matrix [37]. In cartilage, chondrocytes are embedded in an extracellular matrix with an unusually high polyanionic glycosaminoglycan content. The extracellular ionic environment is thus different from that of most other cells: extracellular [Na+] in cartilage is variable, i.e., 250–350 mmol/l and cartilage glycosaminoglycan content directly correlates with extracellular matrix [Na+]. Consequently, the osmolality of cartilage is 350–450 mOsmol/kg. This high osmotic strength is essential for maintaining a high water content in the cartilage matrix, providing it with unique weight-bearing biomechanical properties. However, the Na+ in cartilage is not free – much of it is associated with glycosaminoglycans and ENaC may be responsible for sensing this matrix [Na+]. Therefore, in addition to sodium homeostasis, mechanosensation, and mechanoperception, even more novel roles for ENaC may include sodium sensing in cartilage and regulation of epidermal differentiation in skin [14].

Acknowledgements

We wish to thank Dr. Martin Francis (Nuffield Orthopaedic Centre, Oxford University) for supplying us with cells derived from human skeletal tissues.

References

1.
Awayda MS, Bengrine A, Tobey NA, Stockand JD, Orlando RC. Nonselective cation transport in native esophageal epithelia. Am J Physiol Cell Physiol. (2004);287:C395–C402. [PubMed: 15197006]
2.
Buschmann MD, Hunziker EB, Kim YJ, Grodzinsky AJ. Altered aggrecan synthesis correlates with cell and nucleus structure in statically compressed cartilage. J Cell Sci. (1996);109:499–508. [PubMed: 8838673]
3.
Brouard M, Casado M, Djelidi S, Barrandon Y, Farman N. Epithelial sodium channel in human epidermal keratinocytes: expression of its subunits and relation to sodium transport and differentiation. J Cell Sci. (1999);112:3343–3352. [PubMed: 10504339]
4.
Canessa CM, Horisberger JD, Rossier BC. Epithelial sodium channel related to proteins involved in neurodegeneration. Nature. (1993);361:467–470. [PubMed: 8381523]
5.
Canessa, CM, Schild L, Buell G, Thorens B, Gautschi I, Horisberger J-D, Rossier BC. Amiloride-sensitive epithelial Na+ channel is made of three homologous subunits. Nature. (1994);367:463–467. [PubMed: 8107805]
6.
Cantiello HF. Role of the actin cytoskeleton on epithelial Na+ channel regulation. Kidney Int. (1995);48:970–984. [PubMed: 8569107]
7.
Cantiello HF. Actin filaments stimulate the Na+, K+-ATPase. Am J Physiol. (1995);269:F637–F643. [PubMed: 7503229]
8.
Drenckhahn D, Schulter K, Allen DP, Bennett V. Colocalization of Band 3 with ankyrin and spectrin at the basal membrane of intercalated cells in the rat kidney. Science. (1985);230:1287–1290. [PubMed: 2933809]
9.
Duc C, Farman N, Canessa CM, Bonvalet JP, Rossier BC. Cell-specific expression of epithelial sodium channel alpha, beta, and gamma subunits in aldosterone-responsive epithelia from the rat: localization by in situ hybridization and immunocytochemistry. J Cell Biol. (1994);127:1907–1921. [PMC free article: PMC2120291] [PubMed: 7806569]
10.
Duncan R, Hruska K. Chronic, intermittent loading alters mechanosensitive channel characteristics in osteoblast-like cells. Am J Physiol. (1994);267:F909–F916. [PubMed: 7528987]
11.
Duncan RL, Akanbi KA, Farach-Carson MC. Calcium signals and calcium channels in osteoblastic cells. Semin Nephrol. (1998);18:178–190. [PubMed: 9541272]
12.
Ervasti JM, Campbell KP. Dystrophin and the membrane cytoskeleton. Curr Opin Cell Biol. (1993);5:82–87. [PubMed: 8448034]
13.
Fyfe GK, Quinn A, Canessa CM. Structure and function of the Mec-ENaC family of ion channels. Semin Nephrol. (1998);18:138–151. [PubMed: 9541269]
14.
Guitard M, Leyvraz C, Hummler E. A nonconventional look at ionic fluxes in the skin: lessons from genetically modified mice. News Physiol Sci. (2004);19:75–79. [PubMed: 15016907]
15.
Harter L, Hruska K, Duncan R. Human osteoblast-like cells respond to mechanical strain with increased bone matrix protein production independent of hormonal regulation. Endocrinology. (1994);136:1–7. [PubMed: 7530647]
16.
Hummler E, Barker P, Gatzy J, Beermann F, Verdumo C, Schmidt A, Boucher R, Rossier BC. Early death due to defective neonatal lung liquid clearance in alpha-ENaC deficient mice. Nat Genet. (1996);12:325–328. [PubMed: 8589728]
17.
Ingber DE. Tensegrity: the architectural basis of cellular mechanotransduction. Annu Rev Physiol. (1997);59:575–599. [PubMed: 9074778]
18.
Killick R, Richardson G. Isolation of chicken alpha ENaC splice variants from a cochlear cDNA library. Biochim Biophys Acta. (1997);1350:33–37. [PubMed: 9003454]
19.
Kizer N, Guo X-L, Hruska K. Reconstitution of stretch-activated cation channels by expression of the α-subunit of the epithelial sodium channel cloned from osteoblasts. Proc Natl Acad Sci USA. (1997);94:1013–1018. [PMC free article: PMC19631] [PubMed: 9023374]
20.
Liu J, Schrank B, Waterston RH. Interaction between a putative mechanosensory membrane channel and a collagen. Science. (1996);273:361–364. [PubMed: 8662524]
21.
Mauro T, Guitard M, Behne M, Oda Y, Crumrine D, Komuves L, Rassner U, Elias PM, Hummler E. The ENaC channel is required for normal epidermal differentiation. J Invest Dermatol. (2002);118:589–594. [PubMed: 11918703]
22.
Mobasheri A, Mobasheri R, Francis MJO, Trujillo E, Alvarez De La Rosa D, Martín-Vasallo P. Ion transport in chondrocytes: Membrane transporters involved in intracellular ion homeostasis and the regulation of cell volume, free [Ca2+] and pH. Histol Histopathol. (1998);13:893–910. [PubMed: 9690144]
23.
Mobasheri A, Pocock AE, Trujillo E, Ferraz I, Martin-Vasallo P, Francis MJO. Detection of mRNA and protein of the α subunit of the epithelial sodium channel (ENaC) in human osteoblasts and human ACL cells. J Physiol (London) (2001);535:4P–5P.
24.
Mobasheri A, Killick R, Trujillo E, Alvarez De La Rosa D, Marín R, Canessa CM, Martín-Vasallo P. Subunits of the epithelial sodium channel in avian and human chondrocytes. J Physiol (London) 1999;518P:117P.
25.
Morrow JS, Cianci CD, Aidito T, Mann AS, Kashgarian M. Ankyrin links fodrin to the α subunit of the Na+, K+-ATPase in MDCK kidney cells and in intact renal tubule cells. J Cell Biol. (1989);108:455–465. [PMC free article: PMC2115445] [PubMed: 2537316]
26.
O'hagan R, Chalfie M, Goodman MB. The MEC-4 DEG/ENaC channel of Caenorhabditis elegans touch receptor neurons transduces mechanical signals. Nat Neurosci. (2005);8:43–50. [PubMed: 15580270]
27.
Peng BG, Ahmad S, Chen S, Chen P, Price MP, Lin X. Acid-sensing ion channel 2 contributes a major component to acid-evoked excitatory responses in spiral ganglion neurons and plays a role in noise susceptibility of mice. J Neurosci. (2004);24:10167–10175. [PubMed: 15537887]
28.
Rossier BC. Mechanosensitivity of the epithelial sodium channel (ENaC): controversy or pseudocontroversy? J Gen Physiol. (1998);112:95–96. [PMC free article: PMC2525749] [PubMed: 9689020]
29.
Rossier BC, Canessa CM, Schild L, Horisberger JD. Epithelial sodium channels. Curr Opin Nephrol Hypertens. (1994);3:487–496. [PubMed: 7804746]
30.
Roudier-Pujol C, Rochat A, Escoubet B, Eugene E, Barrandon Y, Bonvalet JP, Farman N. Differential expression of epithelial sodium channel subunit mRNAs in rat skin. J Cell Sci. (1996);109:379–385. [PubMed: 8838661]
31.
Shakibaei M, Mobasheri A. Beta1-integrins co-localize with Na, K-ATPase, epithelial sodium channels (ENaC) and voltage activated calcium channels (VACC) in mechanoreceptor complexes of mouse limb-bud chondrocytes. Histol Histopathol. (2003);18:343–351. [PubMed: 12647783]
32.
Smith PR, Saccomani G, Joe E, Angelides KJ, Benos DJ. Amiloride-sensitive sodium channel is linked to the cytoskeleton in renal epithelial cells. Proc Natl Acad Sci USA. (1991);88:6971–6975. [PMC free article: PMC52215] [PubMed: 1651488]
33.
Suzuki H, Kerr R, Bianchi L, Frokjaer-Jensen C, Slone D, Xue J, Gerstbrein B, Driscoll M, Schafer WR. In vivo imaging of C. elegans mechanosensory neurons demonstrates a specific role for the MEC-4 channel in the process of gentle touch sensation. Neuron. (2003);39:1005–1017. [PubMed: 12971899]
34.
Syntichaki P, Tavernarakis N. Genetic models of mechanotransduction: the nematode Caenorhabditis elegans. Physiol Rev. (2004);84:1097–1153. [PubMed: 15383649]
35.
Tavernarakis N, Driscoll M. Molecular modeling of mechanotransduction in the nematode Caenorhabditis elegans. Ann Rev Physiol. (1997);59:659–689. [PubMed: 9074782]
36.
Tavernarakis N, Driscoll M. Mechanotransduction in Caenorhabditis elegans: the role of DEG/ENaC ion channels. Cell Biochem Biophys. (2001);35:1–18. [PubMed: 11898851]
37.
Titze J, Shakibaei M, Schafflhuber M, Schulze-Tanzil G, Porst M, Schwind KH, Dietsch P, Hilgers KF. Glycosaminoglycan polymerization may enable osmotically inactive Na+ storage in the skin. Am J Physiol Heart Circ Physiol. (2004);287:H203–H208. [PubMed: 14975935]
38.
Trujillo E, Alvarez De La Rosa D, Mobasheri A, Gonzalez T, Martín-Vasallo P. Na+ transport in cartilage. Changes in osteoarthritis. Arthritis Rheum. (1998);41:1610. (Abstract).
39.
Trujillo E, Alvarez De La Rosa D, Mobasheri A, González T, Canessa CM, Martín-Vasallo P. Sodium transport systems in human chondrocytes II. Expression of ENaC, Na+/K+/2Cl- cotransporter and Na+/H+ exchangers in healthy and arthritic chondrocytes. Histol Histopathol. (1999);14:1023–1031. [PubMed: 10506918]
40.
Wang N, Butler JP, Ingber DE. Mechanotransduction across the cell surface and through the cytoskeleton. Science. (1993);260:1124–1127. [PubMed: 7684161]
41.
Watson AJM, Levine S, Donowitz M, Montrose MH. Serum regulates Na+/H+ exchange in Caco-2 cells by a mechanism which is dependent on F-actin. J Biol Chem. (1992);267:956–962. [PubMed: 1309780]