|
|
||||||||
1 Department of Physiology and Biophysics, University of Texas Medical Branch, Galveston, Texas 77551; and
2 Department of Physiology, The Johns Hopkins University School of Medicine, Baltimore, Maryland 21205
| Abstract |
|---|
| Introduction |
|---|
|
|
|---|
| Chloride channels and epithelial functions |
|---|
|
|
|---|
| A brief history of the ClC channel family |
|---|
|
|
|---|
|
The function and general structure of most ClC channels is, at present, incompletely understood. ClC protein monomers have predicted molecular masses ranging from 69 to 110 kDa. Hydropathy analyses indicate that they possess 13 hydrophobic domains (referred to as D1D13, respectively). Approximately 1012 of these domains are thought to span the cell membrane, with domain D13 and the COOH and NH2 terminals of the protein located in the cytoplasm (3, 4, 7) (see Fig. 2
). Recent analysis of two-dimensional crystals of a bacterial ClC channel homologue have confirmed that these channels are formed as dimers with two off-axis water-filled pores (9).
|
| Distribution of ClC channels in epithelia |
|---|
|
|
|---|
Epithelial cells, as well as neurons, blood cells, and skeletal and smooth muscle, can express more than one type of ClC protein, and it remains unclear how simultaneously expressed ClCs interact. Although some ClC channels such as ClC-0 have been shown to be homodimers, other ClC proteins (ClC-1 and ClC-2) form heteromeric channels with novel conductance properties when coexpressed in Xenopus oocytes (4). Whether endogenously expressed ClC channels normally form such heteromultimers remains an unanswered question.
| Functional insights from specific ClC channels |
|---|
|
|
|---|
ClC-2 has been speculated to function as a swelling-activated chloride channel in epithelial cells. In neurons, it has been hypothesized to set intracellular chloride levels, thereby modulating membrane excitability. Although ClC-2 has been the focus of considerable interest from epithelial physiologists as a possible alternative pathway for chloride transport in CF-affected epithelia, most current knowledge of its function and regulation has been gleaned from analyses using heterologous expression systems. Consequently, the necessity of considering its function in native epithelia has become increasingly clear.
To address this issue, several investigators have focused on characterizing chloride currents endogenous to epithelial cells by using molecular tools developed from knowledge of the primary structure of ClC-2. To understand whether this channel underlies swelling-activated or chloride secretory currents, these investigations have aimed to determine the distribution of ClC-2 and the regulation of its expression and conductance properties. Only a few salient findings will be considered in the following discussion.
Hyperpolarization-activated chloride currents with biophysical and pharmacological profiles similar to those of ClC-2 have been reported from whole cell recordings of numerous native and cultured epithelial cells even before its cloning. Park et al. (10) used RT-PCR and Western blot analysis to demonstrate the presence of ClC-2 transcript and protein in rat parotid gland acinar cells, a system in which they previously had measured hyperpolarization-activated chloride currents. Transient overexpression of ClC-2 in HEK cells produced currents with similar kinetics, leading these investigators to infer its involvement in the native hyperpolarization-activated chloride current of parotid acinar cells.
Kajita et al. (5), following earlier studies of hyperpolarization-activated chloride currents in choroid plexus epithelium, presented Western blot evidence for the presence of ClC-2 in these cells. Interestingly, transfection with antisense RNA specific to ClC-2 not only reduced the detected levels of ClC-2 but also significantly attenuated the current. What remains puzzling are the apparent differences in the halide selectivity and sensitivity to external pH of native chloride currents in choroid plexus compared with heterologously expressed ClC-2 (4).
One potential answer to these discrepant features of hyperpolarization-activated chloride currents in epithelial cells is the possibility that these currents result from channels formed by heteromeric interactions between ClC-2 and coexisting ClC species (4). Heterogeneity could also result from alternative splicing, as recently shown for ClC-2 clones isolated from a guinea pig intestinal epithelial library (2). Clearly, further studies are required to test the likelihood of these possibilities as potential mechanisms of functional diversity.
Few studies have addressed the question of whether ClC-2 expression in epithelial cells is polarized, i.e., restricted to either the apical or basolateral membrane. Blaisdell et al. (1) investigated the role of ClC-2 in mediating transepithelial chloride transport across primary fetal lung epithelial monolayers, a chloride secretory epithelium. Using confocal microscopy of immunofluorescently labeled ClC-2, these investigators localized this protein near the apical membrane. In addition, they were able to stimulate the short-circuit current by acidification of the apical bathing solution and to block this increase with mucosal addition of cadmium, as predicted for the functional properties of ClC-2 present in the apical membrane. Moreover, the pH-evoked increases in the short-circuit current were comparable to current levels observed following stimulation of CFTR by cAMP in the same epithelia. These findings provide indirect evidence that ClC-2 can mediate chloride transport across the apical membranes of the airway epithelia. Further work is needed to determine whether this channel can compensate for defective chloride transport function in CF.
Further insights relevant to this issue come from Tarran et al. (14). In whole cell patch-clamp studies of ciliated airway epithelial cells from wild-type and CFTR knockout mice, these investigators found evidence that CFTR might regulate a basal chloride current similar to ClC-2. This current was detected in ~3% of CFTR knockout mice compared with 27% in normal mice. Increases in intracellular chloride levels shift the voltage dependence of ClC-2 gating to more depolarized potentials closer to the physiological range (reviewed in Ref. 4). Therefore, this apparent CFTR regulation of ClC-2 could be by a direct interaction with CFTR and/or by a change in the intracellular chloride concentration concomitant with cAMP-stimulated chloride secretion.
ClC-K channels: role in renal function.
The ClC-K channels are the only group of ClC channels that are almost exclusively expressed in the kidney and urinary tract. For this reason, the cloning of the closely related channels ClC-K1 and ClC-K2 from rat, and the similar human channels ClC-Ka and ClC-Kb, spurred much speculation regarding their roles in renal physiology and pathophysiology. These channels have a high degree of sequence identity within species (91% between ClC-Ka and ClC-Kb and 82% for ClC-K1 and ClC-K2). The lack of complete correspondence between human and rat ClC-K proteins initially hindered their comparison as cross-species orthologs. In addition, early attempts to functionally express ClC-K channels by using heterologous systems resulted in controversy. Recent studies offering insights into these remaining puzzles will be discussed here.
CLC-KB.
We will first discuss ClC-Kb since a breakthrough in the understanding of ClC-K channels first came through studies of this channel. In elegant genetic investigations, Simon et al. (13) unequivocally linked mutations in the gene encoding ClC-Kb with the human disease Bartter's syndrome type III. As in the case of mutations of the Na+-K+-2Cl cotransporter (NKCC2) and the ATP-sensitive K+ channel (ROMK; Kir1.1) that result, respectively, in Bartter's syndrome types I and II, mutations of ClC-Kb result in a phenotype that includes salt wasting, hypokalemic alkalosis, and hypotension (Fig. 3
). Because Bartter's syndrome is characterized by hypotension, it follows that inhibition of ClC-Kb (as well as NKCC and ROMK) could have antihypertensive effects. Thus ClC-Kb represents a potential target for pharmacological treatment of hypertension.
|
Clearly, many gaps in our knowledge concerning ClC-Kb remain to be filled. In particular, the electrophysiological fingerprint of this molecular entity, in its entirety, must be confirmed and direct evidence for its possible role in mediating the basolateral conductance of the thick ascending limb of Henle's loop (TALH) is needed. The development of a reliable expression system is also required, not only to allow determination of the properties of the wild-type and mutant ClC-Kb channels but also to test potential inhibitors of ClC-Kb as antihypertensive agents.
CLC-KA.
In contrast to ClC-Kb, the role of ClC-Ka in renal cell epithelial physiology is less well established. Recent studies suggest that ClC-Ka is homologous to ClC-K1. A mouse ClC-K1 knockout (8) resulted in animals that were unable to concentrate urine, producing a phenotype of diabetes insipidus. In addition, the chloride conductance of the thin ascending limb was reduced. This pattern is distinct from the Bartter's phenotype. Together with the clearly divergent functional properties of heterologously expressed ClC-K1 and chimeric ClC-Kb/K1, these findings suggest that human ClC-Ka functions differently from ClC-Kb and may play a role in the urine-concentrating mechanism.
"Outwardly rectifying" ClC channels (ClC-3, -4, and -5).
ClC-3, -4, and -5 channels comprise a common branch of the ClC family and have a predicted amino acid identity of ~80% (4). These channels, like the ClC-K channels, initially proved relatively difficult to express, and perhaps for this reason their properties were disputed. Recent transient expression studies in mammalian cells have provided convincing evidence that both ClC-4 and ClC-5 currents show extreme outward rectification, conducting only at potentials exceeding +20 mV and characterized by the halide selectivity sequence of chloride > bromide > iodide (see Ref. 16). In contrast, results for ClC-3 have been controversial, with some authors reporting a more linear conductance with a selectivity of iodide > chloride. Recently, Li et al. (6) reported that transient expression of rat ClC-3 in mammalian cultured cells resulted in a current that, with respect to rectification and permselectivity, was nearly identical to those of ClC-4 or ClC-5. In the remainder of this paper, we will focus on ClC-3 and ClC-5 since presently little is known about the distribution or function of ClC-4 in epithelial tissues.
CLC-3.
ClC-3 is widely distributed and highly conserved across species. Human and rat ClC-3 differ by only two amino acids. ClC-3 also has the highest homology of any ClC channel (~60%) to scClC-1 (gef1), a yeast protein that facilitates copper incorporation into a transmembrane iron transport protein (4). Although only one ClC-3 isoform has been found in humans, a second, shorter isoform that is apparently a splice variant exists in rats. Li et al. (6) observed plasma membrane currents associated with expression of the short isoform, but not the long isoform, by transient transfection of CHO-K1 cells. Although the reasons for this difference are unclear, the presence of two isoforms certainly will complicate further functional analysis of this protein.
Several groups have reported a predominantly intracellular localization of ClC-3. In yeast studies, the closely related scClC-1 (gef1) has been localized to a post-Golgi compartment (reviewed in Ref. 4). ClC-3 has been localized to the perinuclear region in pigmented ciliary epithelial cells (18), renal A6 cells (for discussion, see Ref. 20), and retinal pigment epithelial cells (20), whereas in hepatocytes it has been localized to the canalicular membrane (see Ref. 6).
Despite its predominantly intracellular localization in many epithelia, ClC-3 has been proposed as a volume-activated membrane channel responsible for regulatory volume decrease (RVD; for review see Ref. 7). Specifically, ClC-3 has been reported to be activated by cell swelling, and mutation of a single amino acid abolished this regulation (reviewed in Ref. 7). However, subsequent evidence from several laboratories does not support this view. Although antisense inhibition of ClC-3 expression decreased swelling-activated currents in pigmented ciliary epithelium (18), these cells still exhibited significant volume-activated chloride currents and were able to volume regulate, i.e., the effects on RVD and currents were only partial, indicating that other genes are involved in RVD. In addition, transiently expressed ClC-3 currents had different properties from swelling-activated currents in CHO-K1 cells (6). For these reasons, the previously postulated role of ClC-3 in cell volume regulation presently remains unclear and controversial (see NOTE ADDED IN PROOF).
CLC-5.
The detailed linkage analysis of ClC-5 to Dent's disease qualifies this channel, from a genetic standpoint, as one of the most thoroughly studied epithelial ClC channels. Dent's disease is a renal disorder characterized by hypercalciuria, kidney stone formation, and low-molecular-weight proteinuria. Expression of ClC-5 channels with Dent's-related mutations in oocytes led to a loss of chloride conductance function. How the loss of a chloride conductance can lead to the above pathology remains a major puzzle. Two major groups have recently reported the successful engineering of ClC-5 knockout mice (11, 19). These models undoubtedly will prove useful in bridging the gap that presently exists between our understanding of the defective ClC-5 genotype and the resulting phenotype of Dent's disease.
Although ClC-5 channels in humans initially were believed to be expressed exclusively in the kidney, recent studies reported the detection of ClC-5 transcript or protein in several other human cells and tissues, including airway epithelium, retinal pigment epithelium, and aortic endothelial and smooth muscle cells (for discussion, see Refs. 12 and 20). Since there are no reports of functional alterations in organs other than the kidney in Dent's patients, other proteins or mechanisms presumably compensate for ClC-5 malfunction in these other cells.
Within the mammalian kidney, there is general agreement that ClC-5 is predominantly expressed in the proximal tubule (particularly the S3 segment) within intracellular vesicles, with some reports of expression in the apical membrane of cortical collecting duct intercalated cells. Several laboratories have colocalized ClC-5 with H+-ATPase in proximal tubule cells (for review, see Ref. 16). For this reason, it was proposed that ClC-5 mediates a chloride conductance that is necessary for endosomal acidification (16). As shown in Fig. 4A
, the negative charge carried by chloride would compensate for the accumulation of positive charges within the endosomal lumen due to the transport of protons into this space. Disruption of ClC-5 would limit proton entry and, hence, acidification, inhibiting normal endocytosis in the proximal tubule and preventing reabsorption of protein from the ultrafiltrate. The model is attractive because it explains the proteinuria of Dent's disease. However, on the basis of the present understanding of ClC-5 currents from functional expression studies, it has several shortcomings. First, ClC-5 channels do not appear to be optimally suited for the proposed shunting function. As noted above, when ClC-5 is expressed in the plasma membrane, there is a strong outward current, i.e., chloride moves from the extracellular fluid into the cytoplasm and the channel conducts relatively little current in the opposite direction. If a similar channel orientation occurs in the endosome, chloride would move into the cytoplasm, not into the endosomal lumen. Therefore the current is in the wrong direction to support efficient acidification, i.e., the movement is opposite to the direction indicated in Fig. 4A
. Second, acidic solutions inhibit ClC-5 currents, therefore making it unlikely that chloride movements through ClC-5 channels would directly facilitate acidification as shown in Fig. 4A
. Consequently, the role of ClC-5 proteins in endosomal function may be different or more complex than initially thought.
|
Because of the high urine calcium levels seen in Dent's disease, the relationship of ClC-5 to calcium homeostasis presents another unsolved riddle. Recently, decreased ClC-5 mRNA and protein levels were observed in the renal cortices of vitamin D-deficient, thyroparathyroidectomized rats (12). Injection of these animals with parathyroid hormone (PTH) restored expression levels to normal. It is difficult to discern whether PTH directly upregulates ClC-5 channels or whether the effects are indirect. PTH is known to stimulate the trafficking of sodium phosphate transporters in renal cells by targeting these proteins for lysosomal degradation. Further studies using channels tagged as fluorescent fusion proteins to monitor trafficking of ClC-5 channels may prove helpful in this regard.
Recently, two separate groups have independently engineered knockout mice by using a similar strategy of disrupting regions of the ClC-5 gene encoding functionally important channel structures and inserting a neomycin resistance cassette in place of these areas (11, 19). The phenotype generated by one of these models is astonishingly close to that manifested in Dent's-affected individuals (19) and includes not only the low-molecular-weight proteinuria that also was observed in the other mouse (11) but also, importantly, the development of hypercalciuria, kidney calcium deposits, and bone abnormalities. Together, these two mouse models provide exceptional promise for future physiological studies that will be useful in solving the mysteries of how ClC-5 dysfunction leads to impaired endocytosis and the pathology of Dent's disease.
| Summary and remaining issues |
|---|
|
|
|---|
NOTE ADDED IN PROOF
This point (see p. 165) has recently been clarified by studies of a ClC-3 knockout mouse that show no change in volume-regulated Cl currents (see Strowbawa SM et al., Neuron 29: 185196, 2001).
| Acknowledgments |
|---|
Our work is supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-53352 (N. K. Wills) and Cystic Fibrosis Research, Inc. (P. Fong).
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
C. A. Wagner, K. E. Finberg, S. Breton, V. Marshansky, D. Brown, and J. P. Geibel Renal Vacuolar H+-ATPase Physiol Rev, October 1, 2004; 84(4): 1263 - 1314. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yamamoto-Mizuma, G.-X. Wang, L. L. Liu, K. Schegg, W. J. Hatton, D. Duan, T. L. B. Horowitz, F. S. Lamb, and J. R. Hume Altered properties of volume-sensitive osmolyte and anion channels (VSOACs) and membrane protein expression in cardiac and smooth muscle myocytes from Clcn3-/- mice J. Physiol., June 1, 2004; 557(2): 439 - 456. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Mo, W. Xiong, T. Qian, H. Sun, and N. K. Wills Coexpression of complementary fragments of ClC-5 and restoration of chloride channel function in a Dent's disease mutation Am J Physiol Cell Physiol, January 1, 2004; 286(1): C79 - C89. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Fong, B. E. Argent, W. B. Guggino, and M. A. Gray Characterization of vectorial chloride transport pathways in the human pancreatic duct adenocarcinoma cell line HPAF Am J Physiol Cell Physiol, August 1, 2003; 285(2): C433 - C445. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Li, S. T. Halm, and D. R. Halm Secretory activation of basolateral membrane Cl- channels in guinea pig distal colonic crypts Am J Physiol Cell Physiol, April 1, 2003; 284(4): C918 - C933. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Isnard-Bagnis, N. Da Silva, V. Beaulieu, A. S. L. Yu, D. Brown, and S. Breton Detection of ClC-3 and ClC-5 in epididymal epithelium: immunofluorescence and RT-PCR after LCM Am J Physiol Cell Physiol, January 1, 2003; 284(1): C220 - C232. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. X. Weng, B. F. Godley, G. F. Jin, N. J. Mangini, B. G. Kennedy, A. S. L. Yu, and N. K. Wills Oxidant and antioxidant modulation of chloride channels expressed in human retinal pigment epithelium Am J Physiol Cell Physiol, September 1, 2002; 283(3): C839 - C849. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. A. Gilbertson Hypoosmotic Stimuli Activate a Chloride Conductance in Rat Taste Cells Chem Senses, May 1, 2002; 27(4): 383 - 394. [Abstract] [Full Text] [PDF] |
||||
![]() |
X. Li, T. Wang, Z. Zhao, and S. A. Weinman The ClC-3 chloride channel promotes acidification of lysosomes in CHO-K1 and Huh-7 cells Am J Physiol Cell Physiol, June 1, 2002; 282(6): C1483 - C1491. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |