| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Division of Nephrology, Université Catholique de Louvain Medical School, B-1200 Brussels, Belgium; and 2 Department of Physiology, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
THE CLC FAMILY OF...
CFTR
CONCLUSIONS AND PERSPECTIVES
REFERENCES
|
|---|
Cl
channels are
involved in a range of functions, including regulation of cell volume
and/or intracellular pH, acidification of intracellular vesicles, and
vectorial transport of NaCl across many epithelia. Numerous
Cl channels have been identified in the kidney, based on
single-channel properties such as conductance, anion selectivity,
gating, and response to inhibitors. The molecular counterpart of many
of these Cl channels is still not known. This review will
focus on gene-targeted mouse models disrupting two structural classes
of Cl channels that are relevant for the kidney: the CLC
family of voltage-gated Cl channels and the CFTR.
Disruption of several members of the CLC family in the mouse provided
useful models for various inherited diseases of the kidney, including
Dent's disease and diabetes insipidus. Mice with disrupted CFTR are
valuable models for cystic fibrosis (CF), the most common autosomal
recessive, lethal disease in Caucasians. Although CFTR is expressed in
various nephron segments, there is no overt renal phenotype in CF.
Analysis of CF mice has been useful to identify the role and potential
interactions of CFTR in the kidney. Furthermore, observations made in
CF mice are potentially relevant to all other models of
Cl channel knockouts because they emphasize the
importance of alternative Cl pathways in such models.
Bartter's syndrome; bicarbonate transport; chloride channel; cystic fibrosis transmembrane conductance regulator; Dent's disease; endocytosis; nephrogenic diabetes insipidus; vesicular acidification
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
THE CLC FAMILY OF...
CFTR
CONCLUSIONS AND PERSPECTIVES
REFERENCES
|
|---|
IN PLANTS AND ANIMAL TISSUES,
Cl is the most abundant anion and Cl
transporters are involved in a range of physiological processes, including regulation of cell volume, regulation of intracellular pH,
acidification of intracellular vesicles, and transepithelial transport
(79). Cl channels are involved in the
vectorial transport of NaCl and fluid across many epithelia. In the
kidney, most of the Cl filtered is reabsorbed, and this
process involves different mechanisms operating in the apical and
basolateral membranes of tubular epithelial cells. Several
Cl transporters use the energy stored in transmembrane
gradients of other ions to move Cl across the apical
plasma membrane often against its electrochemical gradient.
Cl channels located in the basolateral membrane mediate
net Cl efflux from the cell and participate in NaCl
reabsorption by the nephron (19). In addition to the
specialized, NaCl-reabsorbing function, passive Cl
diffusion through Cl channels is involved in cell volume
regulation, for which the parallel movement of Cl and
K+ through swelling-activated channels results in a
reduction of intracellular osmolytes when cells are exposed to external
hypotonicity (96). Transport of Cl through
Cl/HCO
/HCO
conductance in endosomal vesicles is indeed necessary both for endosomal acidification (113) and normal progression and
recycling along the endocytic apparatus (182).
Numerous Cl channels have been identified in all segments
of the nephron, based on single-channel properties such as conductance, anion selectivity, gating, and response to inhibitors
(19). The molecular counterpart of many Cl
channels in the kidney is still not known, and this review will focus on two structural classes of Cl channels that have
been cloned and sequenced: the CLC channels family and the CFTR.
Mutations of two members of the CLC Cl channels family
(ClC-5 and ClC-Kb) in humans lead to two distinct inherited diseases of
the kidney: Dent's disease (102), and Bartter's syndrome
(155), respectively. These two human diseases are
completed by the diabetes insipidus symptoms in the ClC-K1 knockout
mouse (111). Loss-of-function mutations affecting CFTR are
responsible for cystic fibrosis (CF), the most common autosomal
recessive, lethal disease in Caucasians (186). We will
detail how phenotypical analyses in mouse models and human diseases
have provided new insights into the diverse roles played by
Cl channels in the kidney (Table
1).
|
| |
THE CLC FAMILY OF VOLTAGE-GATED CL
CHANNELS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
THE CLC FAMILY OF...
CFTR
CONCLUSIONS AND PERSPECTIVES
REFERENCES
|
|---|
Several members of the CLC family of voltage-gated
Cl channels have been identified in the mammalian kidney
(for review, see Refs. 79 and 183). The first member of
this family, ClC-0, was isolated by Jentsch and colleagues
(80) from the electric organ of the Torpedo. To
date, at least nine different CLC genes, highly conserved during
evolution, have been identified in mammals (79).
Biochemical studies performed chiefly on ClC-0 suggested that CLC
channels have a double-barelled, homodimeric configuration, with each
CLC subunit containing its own pore (105). An alternative model for ClC-1 was proposed, based on a single pore composed of two
subunits (49). Different numbers of transmembrane domains of CLC channels have also been suggested (50, 144). The
two issues have been clarified by the recent crystallization of two bacterial CLC proteins at 3.0-Å resolution (46). The CLC
channels are homodimers with a double-barreled configuration. Each
subunit has its own pore and contains 18 
-helices, 17 of which are
inserted into the membrane. Each subunit is composed of two halves that are structurally related but have opposite orientations within the
membrane. This antiparallel orientation brings together residues that
will form the Cl ion selectivity filter
(46).
With the exception of ClC-1, which is predominantly expressed in skeletal muscle, all other CLC channels have been detected in the kidney. Three members of the CLC family (ClC-K1, ClC-Kb, and ClC-5) were found to be involved in kidney disease in either humans or mice. The broadly expressed ClC-2 channel was postulated to have a role in early nephrogenesis (73), but its disruption in mice did not confirm this hypothesis (21).
ClC-K1 and Nephrogenic Diabetes Insipidus
ClC-K1 and its human homologue ClC-Ka. The ability to concentrate urine is an essential function of the mammalian kidney, which is achieved by water reabsorption in the collecting duct (CD) according to the osmolar gradient across the tubule. In response to increased plasma osmolality, the neurohypophysal antidiuretic hormone arginine vasopressin (AVP) is released and binds to its specific V2 receptors (V2R) located in the principal cells of the CD. The binding of AVP to V2R activates the stimulatory GTP-binding protein Gs, which in turn stimulates adenylate cyclase. The subsequent increase in the cytosolic levels of cAMP leads to protein kinase A activation, phosphorylation of the water channel aquaporin-2 (AQP2), and, eventually, its insertion into the apical membrane (90). On the other hand, urine concentration necessitates the generation of an osmotic gradient extending from the corticomedullary junction to the inner medulla. In the outer medulla, the gradient is generated by the countercurrent multiplication of the transepithelial reabsorption of NaCl in the thick ascending limb (TAL) of the loop of Henle (56). Both NaCl reabsorption and urea recycling play a crucial role in maintaining high interstitial osmolality in the inner medulla (110). Any disruption of the complex mechanism of urinary concentrating ability will potentially lead to diabetes insipidus, a clinical condition qualified as "central," if the urinary concentrating ability is corrected by the exogenous administration of AVP or an analog, or "nephrogenic," if the kidney remains unresponsive to AVP (118).
The inner medulla is characterized by a variable density of long-looped nephrons containing the thin ascending limb (tAL) of Henle's loop. The tAL has long been considered to play an important role by diluting urine and maintaining the hypertonicity of the interstitium through reabsorption of NaCl. In vitro perfusion studies showed that the tAL has the highest Cl permeability of the nephron segments
(77). Following the descending thin limb, where urine has
been concentrated through aquaporin-1 (AQP1)-mediated water
permeability, the tAL constitutes the first diluting segment of the
Henle's loop, and this is achieved through the passive reabsorption of
NaCl (78). Insights into the molecular basis of NaCl
reabsorption in the tAL were provided by Kondo et al. (93)
and Yoshitomi et al. (193), whose work showed evidence for
the presence of stilbene-sensitive Cl channels in both
luminal and basolateral membranes of the tAL. Using a PCR cloning
strategy, Uchida and colleagues (173) isolated ClC-K1, a
member of the CLC family that is almost exclusively expressed in the
kidney. The homologue in humans was named ClC-Ka (88). The
rat ClC-K1 mRNA is mainly expressed in the tAL, and its expression is
significantly increased by dehydration (173). These data
were confirmed at the protein level, with the ultrastructural localization of ClC-K1 in both the apical and basolateral membranes of
tAL epithelial cells (175, 176). When ClC-K1 is expressed in Xenopus laevis oocytes, Cl currents are
induced and regulated by extracellular pH and Ca2+
concentration. In contrast to other functionally analyzed CLC channels,
the gating of ClC-K1 is not voltage dependent. These currents are
inhibited by furosemide [inhibition constant
(Ki) 100 µM] and stimulated by
N-ethyl-maleimide (175). Because rapid NaCl
exit from the tAL was thought to be a component of the countercurrent multiplication system in the inner medulla (78, 92), it
was suggested that ClC-K1 plays a role in urinary concentration
(175).
Nephrogenic diabetes insipidus in the Clcnk1 knockout mouse.
The hypothesis that ClC-K1 is a major Cl channel
mediating transepithelial Cl transport in tAL and thereby
participating in urinary concentration was confirmed by a mouse model
in which ClC-K1 had been deleted by homologous recombination
(111) (Fig. 1). There were
no apparent differences among Clcnk1/,
Clcnk1+/, and
Clcnk1+/+ mice in survival, gross physical
appearance, and organ morphology. The plasma values of creatinine,
Na+, K+, and HCO mice,
Clcnk1/ mice showed a fivefold increase in
urine volume, coupled to a significant decrease in urine osmolality.
Water deprivation induced a 27% weight loss in
Clcnk1/ mice compared with 12% weight loss
in Clcnk1+/+ and
Clcnk1+/ mice. These
manifestations were associated with a minimal increase in urine
osmolality after administration of the V2R agonist dDAVP, confirming
the clinical picture of nephrogenic diabetes insipidus. In vitro
microperfusion studies showed that establishment of a lumen-to-bath
Cl gradient in the tAL did not yield a Cl
conductance in Clcnk1/ mice, in contrast to
that observed in Clcnk1+/+ and
Clcnk1+/ mice (111).
|
,
and urea were not significantly different between
Clcnk1/ and Clcnk1+/+
mice, clearly demonstrating that polyuria in
Clcnk1/ mice is not osmotic diuresis but
water diuresis. The Clcnk1/ mice were
characterized by increased urinary AVP excretion, in both control and
dehydrated conditions, thus demonstrating the reactivity of the
hypothalamus to high plasma osmolality. In these conditions, the water
diuresis observed in Clcnk1/ mice could be
due to either a loss of water permeability in the CD or impairment in
the generation of hypertonicity in the inner medulla. The first
possibility was ruled out by showing increased expression of AQP2 by
Northern analysis in Clcnk1/ mice, with a
clear immunolocalization of AQP2 in the apical area of CD principal
cells. Because the transcription of AQP2 is regulated by the V2R and a
cAMP-responsive element in its promoter (43, 174),
increased expression of AQP2 was interpreted as the result of increased
secretion of AVP in Clcnk1/ mice. The second
hypothesis was substantiated by demonstrating that papillary osmolality
was significantly lower in Clcnk1/ mice in
both control and dehydrated conditions, an effect that was attributed
to a significant decrease in the accumulation of Na+,
Cl, and urea in the inner medulla (3).
Furthermore, Liu et al. (100) characterized the nature of
NaCl transport in the tAL using in vitro microperfusion. Neither
Cl transport inhibitor
5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB) nor acidification of
the bathing fluid, which typically inhibit the major Cl
transport pathway in tAL of Clcnk1+/+ mice,
changed the transepithelial diffusion voltage
(Vd) values in
Clcnk1/ mice. In contrast, both
Clcnk1+/+ and Clcnk1/
mice showed the same protamine-sensitive Na+ conductance of
the paracellular shunt pathway.
These data obtained in Clcnk1 mice provided the first
demonstration that a molecular deletion in the tAL affects the
countercurrent mechanism in vivo. They also showed that generation of
hypertonicity in the inner medulla probably results from multiple
interactions between NaCl and urea permeability in the tAL
(3). Finally, these studies indicated that ClC-K1 mediates
Cl transport sensitive to NPPB or pH in the tAL and that
the paracellular shunt pathway of this segment has a selective
Na+ permeability (100).
ClC-K2/b and Bartter's Syndrome
ClC-K2 and its human homologue ClC-Kb.
ClC-K2 was isolated from rat kidney by using a PCR-based cloning
strategy (2). This Cl channel shares an
~80% amino acid identity with ClC-K1 and is also exclusively
expressed in the kidney (2). The human isoform ClC-Kb
(which is 95% identical to ClC-Ka) has also been identified in the
kidney (88). One homologue has been identified in the rabbit (rbClC-Ka) (196). The intrarenal distribution of
ClC-K2 is clearly distinct from that of ClC-K1. In situ hybridization with specific riboprobes directed against the nonhomologous
3'-untranslated regions of ClC-K1 and ClC-K2 demonstrated the absence
of colocalization between the two transcripts. It also showed abundant
expression of ClC-K2 in the distal convoluted tubule (DCT), connecting
tubule (CT), and cortical CD and moderate expression in the medullary TAL (192). The homology between ClC-K1 and ClC-K2 hampered
the generation of specific antibodies against ClC-K2 (172,
183), but the difficulty has been overcome by staining
Clcnk1/ kidneys with an anti-ClC-K antibody
that recognizes both ClC-K1 and ClC-K2 (91). This study
confirmed that ClC-K2 is broadly expressed in the distal nephron,
including TAL, DCT, CT, and intercalated cells of the CD. In all these
locations, ClC-K2 is located in the basolateral membrane
(91).
Bartter's syndrome and mutations in ClC-Kb.
Although it was initially not possible to functionally express
ClC-Kb/K2 in X. laevis oocytes (88), its high
degree of sequence homology with other members of the CLC family
predicted that it likely plays a role in transepithelial transport of
Cl in the distal nephron. Even before the generation of a
ClC-Kb/K2 knockout mouse, Simon et al. (155) verified that
hypothesis by reporting mutations of the CLCNKB gene in
patients with Bartter's syndrome. Bartter's syndrome is an autosomal
recessive renal tubular disorder characterized by salt wasting,
hypokalemic alkalosis, and hypercalciuria (142). Because
the syndrome is similar to the action of the loop diuretic furosemide,
it was suggested that Bartter's syndrome results from impaired NaCl
reabsorption in the distal nephron/diluting segment of the loop of
Henle. Simon et al. (155-157) actually proved that
assertion by documenting three types of Bartter's syndromes: Bartter
type I (due to mutations in the apical
Na+-K+-2Cl cotransporter NKCC2);
Bartter type II (due to mutations in the apical K+ channel
ROMK); or Bartter type III (due to mutations in the basolateral Cl channel ClC-Kb). A fourth variant, that associates
Bartter's syndrome with sensorineural deafness, has recently been
linked to mutations in BSND, a gene that encodes barttin, an
integral membrane protein that acts as an essential 
-subunit for
ClC-Ka and ClC-Kb Cl channels (18, 48). This
remarkable collection of genetic and functional evidence puts ClC-Kb
and its crucial -subunit barttin into a functional relationship with
the apical Na+-K+-2Cl
cotransporter and the ROMK K+ channel, suggesting a
mechanism byn which Cl is taken up apically by the
cotransporter and exits basolaterally via ClC-Kb (Fig. 1).
/ pups died
within 2 wk, characterized by low weight, hydronephrosis, dehydration,
renal failure, and impaired ability to concentrate urine. Treatment of
Slc12a1/ pups with indomethacin enabled some
of them to survive into adulthood. The
Slc12a1/ adult mice presented nephrogenic
diabetes insipidus, characterized by polyuria and failure to
concentrate urine after administration of a vasopressin analog. In
contrast to Clcnk1 knockout mice, in which the primary
deficit is located in the tAL, the nephrogenic diabetes insipidus in
Slc12a1/ mice is explained by the lack of
apical NKCC2 in the TAL, resulting in the abolition of transcellular
and paracellular NaCl reabsorption, which eventually alters the
generation of the osmotic gradient necessary for urine concentration
(166).
ClC-5 and Dent's Disease
Identification and characterization of ClC-5.
The CLCN5 gene that codes for the ClC-5 Cl
channel was identified by positional cloning in families with Dent's
disease, an X-linked renal tubular disorder that includes
low-molecular-weight proteinuria, generalized PT dysfunction (renal
Fanconi syndrome), hypercalciuria, nephrocalcinosis, kidney stones, and
renal failure (51). ClC-5 encodes a 746-amino acid protein
and shares ~80% homology with ClC-3 (83) and ClC-4
(181). These three channels form a branch of the CLC
family with only ~30% amino acid identity with other branches of the
family (79). When ClC-5 is expressed in X. laevis oocytes (52, 102, 160) or in Chinese hamster ovary cells (52), it induces strong outwardly rectifying
Cl currents that are inhibited by acidic extracellular
pH. The Cl currents elicited by injection of ClC-5 in
X. laevis oocytes require voltages of +20 mV, a condition
rarely reached in most cells, except for the urinary bladder of the
amphibian Necturus (67). Because some CLC
channels require -subunits to form heteromers (48), it
is possible that acting in combination with -subunits could modify
this voltage dependence.
-type intercalated cells of the cortical CD,
where it again colocalizes with apical H+-ATPase (42,
62, 107, 138) (Fig. 1). A diffuse, intracellular pattern of
ClC-5 expression has also been reported in the medullary TAL of
Henle's loop, at least in rat and human kidney (42, 107). Subcellular fractionation analyses have shown that ClC-5 colocalizes with H+-ATPase and the small GTPases rab4 and rab5a in
endosomes (42, 177). The intracellular distribution of
ClC-5 and its colocalization with endocytosed proteins have been
confirmed in opposum kidney cells that endogenously express ClC-5
(42), as well as in ClC-5-transfected MDCK and COS-7 cells
(62). Schwake et al. (145) have identified a
proline-rich motif in the COOH terminus of ClC-5 resembling the PY
motif, which is critical for the internalization of the endothelial
sodium channel ENaC by clathrin-mediated endocytosis (153), and shown that this motif is important for the
endocytosis of ClC-5 itself.
Mutations of CLCN5 and Dent's disease.
Mutations of CLCN5 were found in Dent's disease and three
other phenotypic variants referred to as X-linked recessive
nephrolithasis, X-linked recessive hypophosphataemic rickets, and the
idiopathic low-molecular-weight proteinuria of Japanese children
(70, 102, 103). Mutations of ClC-5 that are associated
with Dent's disease abolish or markedly reduce the Cl
currents elicited in heterologous expression systems (75,
101-103). Low-molecular-weight proteinuria represents the
most consistent manifestation of Dent's disease and is almost always
detected in female carriers (103, 135). In contrast, there
is considerable interfamilial and intrafamilial variability in other
manifestations of the disease, including hypercalciuria, PT solute
wasting, hypercalciuria, distal tubule disorders, and rickets
(141, 188, 194). There are no clear correlations between
the clinical phenotype of Dent's disease and the >35 mutations of
CLCN5 identified thus far (141, 194).
ClC-5 and endocytosis in the PT.
The expression of ClC-5 in PT cells and its colocalization with
H+- ATPase to subapical endosomes suggested that ClC-5
may have a role in the counterion transport mechanism that facilitates endosomal acidification (42, 62, 107), despite the fact
that its biophysical properties (outward rectification and inhibition by acidic pH) are apparently not ideal for such a role (52, 102). The endosomes form part of the megalin receptor-mediated endocytic pathway that reabsorbs proteins such as albumin and low-molecular-mass (<70 kDa) proteins [e.g.,
1-microglobulin, 2-microglobulin, vitamin
D binding protein (DBP), Clara cell protein (CC16)] that are freely
filtered and almost totally reabsorbed by PT cells (29).
It is well established that a Cl conductance is necessary
for acidification of PT endosomes (9, 134) (Fig.
2). Luminal acidification of the
endosomes is required for the distribution and degradation of
internalized ligands in the endocytic pathway (182) and
for the progression of early endosomes to late endosomes/lysosomes
(7). Endosomal acidification is generated primarily by the
vacuolar H+-ATPase, with acidification increasing
progressively from endocytic vesicles and early endosomes to late
endosomes and, ultimately, lysosomes. The varying degree of endosomal
acidification depends on whether the electrogenic H+-ATPase
is antagonized by other transporters (e.g., by the electrogenic Na+ pump) or facilitated by a coupled Cl
conductance (113, 152). Although the exact nature of the
Cl channels operating in endosomes remains undefined
(22, 143), it has been hypothesized that an impaired
acidification of the endosomal apparatus due to loss of function of
ClC-5 could affect endocytosis, thus causing the systematic occurrence
of low-molecular-weight proteinuria in Dent's disease (42, 62,
102). However, it is still unknown whether luminal acidification
is required for the formation of primary endocytic vesicles or whether
poor entry in the endocytic apparatus is a secondary defect, e.g., due
to inefficient recycling of a rate-limiting partner of the endocytic machinery.
|
ClC-5 in distal nephron segments.
In addition to PT cells, ClC-5 is also expressed in the intercalated
cells of the CD (42, 62, 122, 138) and, at least in human
and rat kidney, in the TAL (42, 107). In the
acid-secreting -type intercalated cells, ClC-5 colocalizes with
H+-ATPase in intracellular vesicles and apical plasma
membrane (42, 62, 138), and it could play a role in distal
urinary acidification (Fig. 1). However, this question remains
debatable, because urinary acidification is normal in the majority of
patients with Dent's disease (141, 188). The TAL is an
important site for regulated Ca2+ reabsorption
(142), and Silva et al. (154) recently showed that ClC-5 mRNA expression in the kidney is regulated by parathyroid hormone (PTH). Thus ClC-5 may play a role in Ca2+
homeostasis, although it remains unclear how an apparently
intracellular Cl channel may participate in the positive
luminal potential that drives paracellular Ca2+
reabsorption in that nephron segment (142). Furthermore,
at variance with Bartter's syndrome (cf. Bartter's syndrome and
mutations in ClC-Kb), Dent's disease is not consistently
associated with hypercalciuria and is usually not associated with
clinical dehydration (141, 188).
Mouse models of ClC-5 disruption. The potential roles of ClC-5 in endocytosis and Ca2+ homeostasis have been substantiated by three distinct mouse models of disrupted ClC-5 expression (Table 1). Luyckx et al. (108) reported the first mouse model of reduced ClC-5 expression by a ribozyme approach, resulting in ~80% reduction in renal ClC-5 protein expression. The transgenic ribozyme-expressing mice showed no obvious phenotype except for a borderline hypercalciuria (that was found only in males), without PT dysfunction and low-molecular-weight proteinuria (108, 194). Of note, the hypercalciuria was prevented by a low-Ca2+ diet, suggesting a role for increased intestinal Ca2+ absorption (108).
Two groups have independently generated and characterized knockout mouse models for Dent's disease by targeted disruption of part of the exon 5 and/or exon 6 of Clcn5 (129, 184). The loss of ClC-5 mRNA and protein in the kidney of both Clcn5/Y mice was demonstrated by Northern
blotting and immunoblotting with NH2-terminal antibodies
(129, 184). The Clcn5/Y mice
reported by Wang et al. (184) were found to have
remarkable similarities to Dent's disease, including
low-molecular-weight proteinuria, generalized aminoaciduria,
glycosuria, phosphaturia, and hypercalciuria. Renal failure, which is
generally only observed in adults with Dent's disease
(188), was not observed in the Clcn5/Y mice, and this may partly be due to
the young age (5-9 wk) of the mice at the time of the initial
assessment. In contrast, some young Clcn5/Y
mice already showed renal calcifications (184) and
subsequently developed nephrocalcinosis and parenchymal lesions
including mild fibrosis and renal cysts (Devuyst O, Guggino S, and
Guggino WB, unpublished observations). The
Clcn5/Y mouse reported by Piwon et al.
(129) was also characterized by low-molecular-weight
proteinuria and generalized PT dysfunction, but it lacked
hypercalciuria. As discussed elsewhere, the heterogeneity in terms of
urinary Ca2+ excretion may reflect genetic background,
differences in residual ClC-5 expression, dietary phosphorus intake, or
other environmental conditions (26, 194).
The two knockout mouse models provided new insights into the role of
ClC-5 in the PT and the pathophysiology of PT dysfunction in Dent's
disease. Both Clcn5/Y mice showed
low-molecular-weight proteinuria in parallel with defective apical
endocytosis, suggesting that ClC-5 is a key mediator of the
Cl conductance that is necessary for early endosomal
acidification and receptor-mediated reabsorption in PT cells
(129, 184) (Fig. 2). Any failure in this process would not
only lead to low-molecular-weight proteinuria but also a loss of amino
acids and glucose, presumably by a failure to recycle their specific
transporters to the apical cell membrane (89). Recent
studies showed that other members of the CLC family of Cl
channels fulfill similar functions in other cells and organelles. Kornak et al. (94) have generated a Clcn7
knockout mouse, demonstrating that ClC-7 provides the Cl
conductance required for acidifying the extracellular resorption lacuna
in osteoclasts. Mutations in CLCN7 have been identified in
patients with familial (autosomal dominant or recessive) osteopetrosis, a condition characterized by impaired bone and cartilage resorption (33, 94). In another example, the broadly expressed ClC-3 channel is involved in granular acidification in the insulin-secreting pancreatic cell (13), and its disruption in mice
impairs acidification of synaptic vesicles and induces severe postnatal
degeneration of the retina and hippocampus (162).
The phenotypical analysis of ClC-5 knockout mice also provided
interesting hypotheses to explain the phosphaturia and hypercalciuria inconsistently observed in Dent's disease. Piwon et al.
(129) showed that mice lacking ClC-5 had reduced renal
expression of megalin, together with a slight, 1.7-fold increase in
urinary PTH. Because megalin mediates the endocytosis and degradation of PTH in PT cells (68), they hypothesized that the
reduced endocytosis of filtered PTH in Clcn5/Y
mice may lead to the subsequent stimulation of functional apical PTH
receptors in PT cells (170). The PTH-mediated regulation of renal phosphate reabsorption involves internalization and subsequent lysosomal degradation of the type II Na+-phosphate
cotransporter (NaPi-2) (169), which mediates most of the
phosphate transport in the PT (14). Accordingly, Piwon et
al. (129) showed reduced expression and enhanced
internalization of NaPi-2 in the PT of Clcn5/Y
mice, which could explain the increased phosphaturia in these mice. The
apical stimulation of PTH receptors may also promote the
internalization of the Na+-H+ exchanger NHE3 in
PT cells, which may contribute to the reduced NaCl reabsorption in
Clcn5/Y mice (129).
The 1-hydroxylase enzyme expressed in PT cells converts the
25-hydroxyvitamin D substrate in its active metabolite,
1,25-dihydroxyvitamin D. A marked increase in the urinary excretion of
the 50-kDa DBP, suggesting a lower delivery of 25-hydroxyvitamin D
substrate to the 1-hydroxylase, was observed in the
Clcn5/Y mice (129, 184). It is
thus conceivable that endocytic defect observed in
Clcn5/Y mice is reflected in terms of vitamin
D and Ca2+ homeostasis. Piwon et al. (129)
observed lower levels of 1,25-dihydroxyvitamin D, partially compensated
for by slightly increased PTH levels, in their
Clcn5/Y mice. In contrast, increased levels of
1,25-dihydroxyvitamin D were found in most patients with Dent's
disease (135, 188) and also in the
Clcn5/Y mouse as reported by Wang et al. (184;
Guggino S, unpublished observations). The complexity of
hormonal regulation and environmental factors may explain differences
in vitamin D metabolism and urinary Ca2+ excretion between
the two strains of Clcn5/Y mice. Further
studies will also be necessary to distinguish between the potential
contributions of intestinal absorption and bone resorption in
maintaining a normal plasma Ca2+ despite hypercalciuria in
patients with Dent's disease (188) and
Clcn5/Y mice (184). A certain
degree of Ca2+ reabsorption in the PT occurring via the
megalin receptor-mediated endocytosis (30) may account for
part of the observed hypercalciuria in patients with Dent's disease
and Clcn5/Y mice (184). However,
it is important to note that regulated Ca2+ reabsorption
occurs in the TAL and DCT, segments in which the role of ClC-5 remains
to be substantiated. Alternatively, the expression of ClC-5 in
enterocytes, where it colocalizes with H+-ATPase in
endosomes (177), may also suggest a primary defect in
intestinal regulation by vesicular recycling of the epithelial Ca2+ channel ECaC (69).
ClC-2 and the Kidney
ClC-2 was cloned by homology to ClC-1 (168). The ClC-2 Cl channel can be activated by hyperpolarization,
acidic extracellular pH, and osmotic cell swelling (79).
Northern analysis demonstrated that ClC-2 is ubiquitously expressed,
with relatively high levels detected in the brain, intestine, and
kidney (168). It has been suggested that ClC-2 is involved
in lung development (120) and may provide an alternative
pathway for Cl secretion in CF airway epithelial cells
(147; see Mouse Models of CF and the Concept of Alternative
Pathways for Cl Conduction). The identification of
alternatively spliced ClC-2 isoforms together with Cl
currents slowly activated by hyperpolarization in embryonic epithelia at the tip of the ureteric bud suggested that ClC-2 exerts a role in
early nephrogenesis (73). Further studies in rat kidney
suggested that ClC-2 expression may be modulated both by AVP
(117) and aldosterone (124). The above
hypotheses were not supported by the disruption of ClC-2 in mice
(21). Indeed, the Clcn2/ mice
did not show abnormalities in lung or kidney development but revealed a
severe degeneration of the testes and the photoreceptors of the retina.
This unexpected phenotype could be explained by the role played by
ClC-2 in the interaction (e.g., pH regulation) between the vulnerable
germ cells and photoreceptors and their supporting epithelial cells
that form the blood-testis and blood-retina barriers, respectively
(21).
| |
CFTR |
|---|
|
TOP
ABSTRACT
INTRODUCTION
THE CLC FAMILY OF...
CFTR
CONCLUSIONS AND PERSPECTIVES
REFERENCES
|
|---|
CFTR and CF
The CFTR protein contains 1,480 amino acids and is a member of the ATP-binding cassette (ABC) transporters family. CFTR is composed of two transmembrane domains (TMD1 and TMD2) and two nucleotide-binding domains (NBD1 and NBD2), separated by a large, polar, regulatory (R) domain containing multiple phosphorylation sites (for a review, see Ref. 151). Mutations in the CF gene that encodes CFTR are responsible for CF, the most common autosomal recessive disease in Caucasians (37, 136, 186). CFTR is located primarily in the apical membrane of epithelial cells, where it provides a pathway for transepithelial Cl movement and
regulates other types of channels, such as the outwardly rectifying
Cl channels (ORCC) (148), the
Na+ channel ENaC (164), and the inwardly
rectifying, ATP-sensitive renal K+ channel ROMK2
(112). Thus CFTR is central in determining transepithelial ion transport and fluid movement across epithelia. In the intestine, pancreas, and sweat gland secretory coil, CFTR plays a key role in
fluid and electrolyte secretion, and in sweat gland duct and airway
epithelia it participates in fluid and electrolyte absorption (37, 186). Failure of protein kinase A-mediated
phosphorylation and activation of CFTR Cl channels
accounts for most of the clinical manifestations of CF, which can
include airway disease, pancreatic failure, meconium ileus, male
infertility, and elevated levels of salt in sweat (20,
136).
In addition to impaired Cl channel activity, a defect in
HCO
900 putative disease-causing mutations in the CF gene, there are a few
that are characterized by normal Cl channel activity
(37). A recent study of 16 such mutations expressed in HEK
293 cells has shown that they are associated with a significant
alteration of HCO-coupled HCO
CFTR in the Kidney
Numerous studies have identified CFTR in the mammalian kidney (159). CFTR mRNA is abundantly expressed in the cortex and outer medulla but much less in the inner medulla (116). Segmental analysis by RT-PCR showed that CFTR is expressed in all nephron segments, including PT, thin limbs of Henle, TAL, distal tubules, and CD (116). Immunostaining for CFTR has been detected in the apical domain of both proximal and distal tubules of the mature human kidney (36). CFTR is also expressed early in the embryonic human and rat kidney, with a transient upregulation in the ureteric bud during the late stage of branching morphogenesis (41, 72). This pattern of expression in the mammalian kidney is similar to that observed during branching morphogenesis in pancreas and lung (71). Cl channels with
properties similar to CFTR have been identified in several nephron
segments. A PKA-activated Cl-conductive pathway
attributed to CFTR has been demonstrated in rat kidney cortical
brush-border membrane vesicles (15). In addition,
CFTR-like Cl currents have been detected in cultured
rabbit DCT cells (167), rat and mouse inner medullary CD
cells (74, 179), and mouse cortical CD cells
(98). The exact role of CFTR in the normal renal
epithelium, which primarily reabsorbs most of the filtered NaCl, is not
yet established.
A splice variant of CFTR is expressed in rat and human kidney and shows distinct regulation during nephrogenesis (41, 72). This truncated CFTR variant lacks 145 bp of exons 13 and 14, resulting in a frameshift and a premature stop codon. The truncated, 75-kDa isoform (TNR-CFTR) contains the TMD1, NBD1, and R domains. It is functional when expressed in X. laevis oocytes, with single-channel properties very similar to full-length CFTR but less efficient processing and plasma membrane expression (116).
CFTR and Intracellular Organelles
In parallel with its role in apical membrane Cl
permeability and salt transport, it has been suggested that CFTR may
also function within intracellular organelles of the endosomal and biosynthetic pathways (for a review, see Ref. 22). Studies
in cultured cells have shown that forskolin-stimulated increases in
cAMP result in a significant inhibition of endocytic activity in the
human colonic epithelial cell line T84 (24) and the CF pancreatic cell line expressing wild-type CFTR (23),
suggesting that CFTR may regulate plasma membrane recycling.
Furthermore, CFTR has been identified and partially purified in
endosomes from rabbit kidney (35), in agreement with the
presence of protein kinase A-mediated Cl permeability in
renal endosomes (9). Thus CFTR-mediated Cl
conductance may participate in regulating and maintaining an acidic pH
of intracellular organelles. Consistent with this hypothesis, Barash et
al. (12) demonstrated a significant acidification defect
in trans-Golgi and prelysosomal compartments in both
immortalized respiratory epithelial cells and nasal polyps from CF
patients. The restricted acidification appears to be due to a high
membrane potential (
) secondary to the lack of counterion
conductance (12, 22). It may also explain the decreased
sialylation of mucins and other glycoproteins secreted by CF cells
(12) and the chronic colonization of CF airway by
pathogens such as Pseudomonas, which bind avidly to
asialated glycosphingolipids (95). However, the role of
CFTR in modulating organelle pH has remained controversial, which
probably reflects the use of different cell lines and the assessment of
pH in different organelles (55, 106, 150). Furthermore, recent studies in CF lung epithelial cells showed an hyperacidification of cellubrevin-labeled endosomes that was attributed to a loss of CFTR
inhibitory effect on Na+ transport in this subset of
endosomes (130).
Renal Manifestations of CF
There is a striking discrepancy between the level of CFTR expression in the human kidney and the lack of clearly documented renal phenotype in CF patients (40, 136, 159, 186). Despite structural and obstructive changes in the vas deferens, epididymis, and seminal vesicles, which suggest that the CF disease process operates during the development of the mesonephric duct (171), CF patients do not show overt developmental abnormalities in the urinary excretory system. Children with CF may present dehydration episodes attributed to excessive NaCl loss in sweat, particularly during warm weather and in case of digestive losses (20), but it is unknown whether a reduced ability of the CF kidney to retain NaCl contributes to these episodes. In contrast, clearance studies have suggested that CF patients have a decreased natriuretic response to Na+ loading and a reduction in free-water clearance, suggestive of impaired renal diluting capacity (17, 44). Increased PT Na+ reabsorption has been documented in some studies (8, 161) but not in others (4). Patients with CF show an increased renal clearance of aminoglycosides (99), which may be due to impaired megalin uptake (114) and/or altered endocytosis in PT cells (139). The issue remains to be clarified, because renal handling of2-microglobulin and other
low-molecular-weight proteins is not apparently altered in CF patients
(84; Devuyst O, unpublished observations). Microscopic nephrocalcinosis has been described in a series of neonates with CF
(82), and CF patients are at increased risk of developing kidney stones (63). At variance with Dent's disease
(141, 188), the risk of nephrolithiasis in CF patients is
not attributed to hypercalciuria (16) but rather to
hyperoxaluria and decreased urinary citrate (126).
Taken together, these data suggest that the putative roles of CFTR
during nephrogenesis, and perhaps in the mature kidney, are probably
complemented by alternative pathways for Cl conduction
(cf. Mouse Models of CF and the Concept of Alternative Pathways
for Cl Conduction). It must be stressed that subtle
abnormalities in development, function, and/or morphology of the kidney
may remain clinically silent or, alternatively, be associated with
specific, rare mutations of CFTR.
Mouse Models of CF and the Concept of Alternative Pathways for
Cl Conduction
F508) accounts for >70%
of the CFTR mutations in the human population and induces a
characteristic CFTR-processing defect (37). Three F508
mouse models with a similar CFTR-processing defect have been produced (34, 178, 195). In one additional model, the glycine at
position 551 has been changed to an aspartate (G551D)
(39). The G551D mutation is relatively common in patients
with CF and results in a mutated CFTR protein that is processed
normally but shows a significant reduction in cAMP-regulated
Cl channel activity (187).
As recently reviewed (59), these mouse models gave
previous information on the pathophysiology of CF in different organs. Except for two strains with a milder disease (45, 178),
the most striking phenotype in the CF mice has been demonstrated in the
intestine, with intestinal blockage and perforation, characteristic pathological changes, and ion transport abnormalities, including a lack
of cAMP-mediated Cl secretion (59). In
addition, CFTR plays an important role in the HCO
An attractive explanation for the discrepancy between human and mouse
CF lung phenotypes is that Cl channel(s) active in mouse
airways may constitute an alternative pathway for Cl in
the absence of CFTR (31). For instance, the upregulation of a Ca2+-mediated Cl secretory pathway has
been evidenced in the nasal mucosa of cftrtm1Unc
mice (60) and in the tracheal epithelium of two other CF
mouse models (34, 39). In contrast, intestinal epithelia
lack such apical Ca2+-activated Cl secretion
(31). It has been suggested that the new family of Ca2+-activated Cl channels (CaCC or
alternatively named CLCA) may provide the molecular counterpart for
this Ca2+-activated Cl conductance (53,
79). Other putative candidates to provide apical membrane
Cl transport in the absence of a functional CFTR include
the voltage-activated ClC-2 (119, 147), ORCC
(54), and a volume-activated Cl channel
(6, 163). Recent studies have also suggested that ClC-5,
which is expressed during lung ontogeny and localized along the luminal
surface of the airway epithelium, may also participate in lung
Cl secretion (47). The data summarized
above, which suggest the increased expression of alternative pathways
in the CF mouse, are potentially relevant to all other models of
Cl channel knockouts because they indicate that the
pathophysiological consequences of the loss of a given Cl
channel pathway can be partially compensated for by the induction of
another. This possibility, which has been documented in the lung of the
CF mice, may apply to other epithelia expessing Cl
channels, including the kidney.
Renal Function in CF Mouse Models
Given the lack of renal manifestations in CF, it is not surprising that there is only limited information about kidney development and function in the CF mouse models (Table 1). The cftrtm1Unc mouse (158) was used to investigate whether CFTR is the molecular counterpart of the 9-pS Cl channel identified in the basolateral membrane of the
TAL of mouse kidney (61). Studies in isolated,
microperfused TAL have shown that the basal NaCl reabsorption occurring
in that nephron segment mainly by the operation of the apical NKCC2
cotransporter and the basolateral Na+-K+-ATPase
and ClC-Kb is increased by cAMP (11). However, the study in cftrtm1Unc mice showed that the 9-pS TAL
Cl channel is active at the same frequency and with the
same channel properties in cell-attached patches from cortical TAL
tubules of CFTR+/+, CFTR+/, and
CFTR/ mice. In addition, both the
Cl absorption across isolated TAL epithelium and the
overall Na+-K+-Cl cotransport
activity are identical in all mice (109). These results
indicate that the 9-pS Cl channel in the mouse TAL is
distinct from CFTR and that CFTR has no influence on the
cAMP-stimulated Cl absorption in this nephron segment
(109). The molecular counterpart of this Cl
channel could involve the epithelial basolateral Cl
conductance regulator (EBCR) belonging to the ABC family
(180). However, it remains unclear whether EBCR is a
channel by itself or modulates other endogenous basolateral
Cl channels located in the TAL.
Micropuncture experiments have shown that cAMP stimulates
Cl transport and fluid absorption in rat PT
(185), an effect attributed to the presence of
cAMP-activated Cl channels in that nephron segment
(1, 149). Kibble et al. (85) investigated PT
function in the F508 cftrtm2Cam mouse
(34) and showed that renal Na+ clearances
under basal conditions or after acute extracellular volume expansion
are similar in wild-type and F508 mice. The latter are also
characterized by a lack of increase in glomerular filtration rate
secondary to volume expansion and a relatively less important increase
in proximal Na+ reabsorption than wild-type mice
(85). Despite identification of CFTR expression in the
outer cortex of wild-type mice by RT-PCR, in situ microperfusion showed
that the basal fluid absorptive rate is similar in wild-type and
F508 mice and is not significantly influenced by addition of
forskolin-dibutyryl cAMP to the perfusate. These data may be explained
by a lack of cAMP-dependent whole cell Cl conductance
evidenced in isolated mouse PT cells (85). In addition, Kibble et al. (86) demonstrated that, under chronic
dietary salt restriction, F508 mice are equally able to reduce renal Na+ excretion but display a significantly higher
amiloride-sensitive Na+-reabsorption than wild-type mice.
Thus F508 mice may handle volume expansion and chronic NaCl
restriction similarly to wild-type mice, but that could involve more
distal parts of the nephron (86). The increased
amiloride-sensitive Na+ reabsorption observed in
salt-restricted F508 mice is consistent with the proposed
interaction between CFTR and ENaC in principal cells of the CD
(98). Accordingly, a reduced inhibition of ENaC in the CD
of F508 mice could explain the enhanced Na+ absorption
in that nephron segment, which may provide an advantage during states
of salt deprivation. It must be noted, however, that other mechanisms,
such as enhanced secretion of aldosterone, may participate in distal
Na+ reabsorption in salt-restricted mice (58).
The cftrtm2Cam F508 CF mouse
(34) has also been used to investigate the
K+-sparing diuretic effect of glibenclamide. Glibenclamide
blocks native K+ channels in renal cells (10),
but in vitro studies have demonstrated that coexpression of CFTR with
Kir1.1 (ROMK2) is necessary to confer glibenclamide sensitivity on
Kir1.1 channels (112). The fact that glibenclamide induced
an equivalent diuresis and a similar K+-sparing effect in
wild-type and F508 mice indicates that the formation of Kir1.1/CFTR
complexes is not required to mediate its diuretic effect in mice
(87).
| |
CONCLUSIONS AND PERSPECTIVES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
THE CLC FAMILY OF...
CFTR
CONCLUSIONS AND PERSPECTIVES
REFERENCES
|
|---|
Since the first mouse models of CF described in 1992, the
generation of transgenic mice has become an essential tool for
understanding the pathophysiology of Cl channels in the
kidney. This is particularly true for Cl channels
belonging to the CLC family, which are implicated in selective nephron
segments, where they participate in essential kidney functions such as
endocytosis, NaCl reabsorption, or urinary concentration. There is a
considerable interplay between the characterization of these mouse
models and the discovery of channel diseases in humans. For instance,
the phenotype of patients with Dent's disease due to mutations in
CLCN5 was consistent with a role played by the
Cl channel ClC-5 in proximal tubule endocytosis
(102), a hypothesis that was indeed verified in two
independent knockout mice lacking ClC-5 (129, 184). The
discovery of mutations in genes coding for ClC-Kb and the -subunit
barttin not only completed the understanding of genetic heterogeneity
in Bartter's syndrome but also provided key insights into the
regulation of this class of channels (18, 48).
Furthermore, the nephrogenic diabetes insipidus observed in
Clcnk1/ mice suggests that mutations in
CLCNKA may be responsible for nephrogenic diabetes insipidus
in humans lacking mutations in V2R and AQP2 (172).
There are numerous perspectives for existing and future mouse models
targeting Cl channels. Studies of mice lacking CLC
channels will be important to further define the physiological role of
some of these channels in the kidney and the phenotypical variability
observed in some models. Knockout mice for ClC-k2 or barttin may
provide useful models for Bartter's syndrome and give new information
about the regulation of CLC channels. Gene targeting will also expand
our knowledge of other types of channels that exhibit a
Cl conductance. An interesting example is aquaporin-6
(AQP6), a member of the aquaporin family of water channels that has
been identified in rat kidney (191). AQP6 is located in
podocytes, PT cells, and -type intercalated cells of the CD, and it
appears to be mostly associated with intracellular vesicles
(191). The low basal water permeability of AQP6 can be
increased by exposure to mercurials or low pH. Interestingly, the
latter conditions are also reflected by a significant anion conductance
(190). The colocalization of AQP6 with the vacuolar
H+-ATPase in intracellular vesicles, as well as the rapid
and reversible activation of the anion conductance of AQP6 by acidic
pH, led to the suggestion that AQP6 may play a role along other
Cl channels in vesicle acidification (190).
Further characterization of CF mouse models will certainly be useful to
identify the role of CFTR in specialized kidney functions such as
endocytosis or its interactions with other types of channels and
transporters in the nephron (115, 146). These studies may also be useful in investigating the apparent tissue specificity of
F508 CFTR trafficking (81, 128) and other mechanisms
that could explain the absence of overt renal phenotype in CF. Mating CF mice with other mouse models will also create opportunities to
characterize the role of CFTR in the pathophysiology of renal cyst formation (127). The issue is particularly
relevant in autosomal dominant polycystic kidney disease (ADPKD),
the most common inherited nephropathy. ADPKD is characterized by the
development of multiple cysts in both kidneys, resulting in end-stage
renal disease in 50% of patients by age 60 yr (65).
Genetic studies have shown that ADPKD is due to mutations of two major
genes, PKD1, responsible for ~85% of cases, and
PKD2, which accounts for the vast majority of other cases
(65). Although the complex mechanisms involved in
cystogenesis are not yet understood (27), abnormal fluid secretion is a critical pathogenic mechanism associated with cyst expansion in ADPKD (165). Several lines of evidence have
demonstrated that cAMP-dependent Cl secretion provides
the driving force for fluid accumulation into ADPKD cysts and that CFTR
is implicated in this process (25, 38, 64). The
availability of mice with targeted disruptions of Pkd1
(104) and Pkd2 (189) may give
insights into this intriguing question.
Finally, one should remain careful about the general limitations of mouse models, including the occurrence of phenotypical variability despite similar genetic backgrounds, the necessity for specific environmental conditions to reveal a phenotype, and the frequent species differences in the structure and function of a given organ (125, 132). However, even these limitations could prove useful because the phenotypical variability of mutant mice according to genetic background underlies the potential importance of genetic modifiers (121). Thus mouse models will also be valuable to investigate disease-modifying genes, an approach that may give insights into complex pathways and provide new therapeutic strategies in renal diseases.
| |
ACKNOWLEDGEMENTS |
|---|
We thank Prof. R. Beauwens and Prof. P. Courtoy for fruitful discussions.
| |
FOOTNOTES |
|---|
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-32753, the Fondation Alphonse and Jean Forton, the Fondation pour la Recherche Scientifique Medicale, and Actions de Recherche Concertées 00/05-260.
Address for reprint requests and other correspondence: W. B. Guggino, Johns Hopkins Univ. School of Medicine, Depts. of Physiology and Medicine, 725 North Wolfe St., Baltimore, MD 21205 (E-mail: wguggino{at}bs.jhmi.edu).
10.1152/ajprenal.00184.2002
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
THE CLC FAMILY OF...
CFTR
CONCLUSIONS AND PERSPECTIVES
REFERENCES
|
|---|
1.
Abdulnour-Nakhoul, S,
and
Boulpaep EL.
Transcellular chloride pathways in ambystoma proximal tubule.
J Membr Biol
166:
15-35,
1998[ISI][Medline].
2.
Adachi, S,
Uchida S,
Ito H,
Hata M,
Hiroe M,
Marumo F,
and
Sasaki S.
Two isoforms of a chloride channel predominantly expressed in thick ascending limb of Henle's loop and collecting ducts of rat kidney.
J Biol Chem
269:
17677-17683,
1994
3.
Akizuki, N,
Uchida S,
Sasaki S,
and
Marumo F.
Impaired solute accumulation in inner medulla of Clcnk1/ mice kidney.
Am J Physiol Renal Physiol
280:
F79-F87,
2001
4.
Aladjem, M,
Lotan D,
Boichis H,
Orda S,
and
Katznelson D.
Renal function in patients with cystic fibrosis.
Nephron
34:
84-86,
1983[ISI][Medline].
5.
Alper, SL.
The band 3-related anion exchanger (AE) gene family.
Annu Rev Physiol
53:
549-564,
1991[ISI][Medline].
6.
Anderson, MP,
Sheppard DN,
Berger HA,
and
Welsh MJ.
Chloride channels in the apical membrane of normal and cystic fibrosis airway and intestinal epithelia.
Am J Physiol Lung Cell Mol Physiol
263:
L1-L14,
1992
7.
Aniento, F,
Gu F,
Parton RG,
and
Gruenberg J.
An endosomal beta COP is involved in the pH-dependent formation of transport vesicles destined for late endosomes.
J Cell Biol
133:
29-41,
1996
8.
Assael, BM,
Marra G,
Tirelli AS,
Cavanna G,
Claris Appiani A,
Giunta A,
Amoretti M,
and
Milani S.
Renal function in cystic fibrosis.
Int J Pediatr Nephrol
7:
213-216,
1986[ISI][Medline].
9.
Bae, HR,
and
Verkman AS.
Protein kinase A regulates chloride conductance in endocytic vesicles from proximal tubule.
Nature
348:
637-639,
1990[Medline].
10.
Bailey, MA,
and
Walter SJ.
Renal effects of glibenclamide: a micropuncture study.
J Pharmacol Ex
