Finnish Genome Center, University of Helsinki,
Post Office Box 21, Tukholmankatu 2, 00014,
Finland.Juha.Kere@helsinki.fi
Our kidneys take care of a major water−and−ion−juggling
process, carefully regulated to keep our bodies at constant osmolality, ion
balance and hydration. The process involves both the formation of a large
volume —about 180 litres per day—of primary urine, and the taking
back of over 99% of that volume to yield 1.0−1.5 litres of final product.
Both essential processes, the filtration of plasma in the glomeruli to make
primary urine and its concentration in the kidney medulla, have been characterized
for some time. It is only recently, however, that key molecules that mediate
these processes have been recognized.
Augmenting the swelling body of knowledge of these players, is a study
presented by Yoshihiro Matsumura and colleagues on
page 95 of this issue. Clcnk1 is a member of the large and trendy
chloride channel (CLC) family, containing many members that have been implicated
in human disease. Becker− and Thomsen−type congenital myotonias
are caused by mutation in CLCN1 (2);
Bartter syndrome type 3, by mutation in CLCNKB (3); and Dent disease, X−linked nephrolithiasis, hypercalciuric
nephrocalcinosis and hypophosphataemic rickets by mutations in CLCN5
(4). However, many CLCs, including CLCNKA
(corresponding to mouse Clcnk1), have remained without cognate
diseases, an unfortunate fate for any gene (or its researchers). CLCNKA
and CLCNKB are highly homologous, separated by 11 kb, with 94%
identity across coding regions and a similar genomic organization. It therefore
comes as no surprise that unequal crossing−over, creating a single chimaeric
gene between the two genes, was detected in a Bartter syndrome patient although
this observation gave no insight into a role for CLCNKA3.
The phenotype of mice deficient in Clcnk1, engineered by Matsumura
et al., suggests that mutations in CLCNKA might be the cause of
some cases of nephrogenic diabetes insipidus, an inability of the kidneys
to respond to antidiuretic hormone (ADH) when urine concentration is desired
by the hypophysis. Two other genes (AVPR2, encoding vasopressin V2
receptor, and AQP2, encoding aquaporin water channel) have already
been found to be mutated in some patients with this disease5,
6.
The current study provides inspiration to assess the status of CLCNKA
in people with nephrogenic diabetes insipidus who lack recognized mutations.
So what, exactly, does Clc−k1 do in the mouse? In brief, it helps
the kidney concentrate urine. It has been known for more than 20 years that
the well−devised architecture of the kidney makes the concentration
job much easier. Primary urine passes through the loops of Henle, which make
long capillary hairpin structures starting from the kidney cortex and extending
deep into the kidney medulla (see figure).
Henle's loops in the medulla run parallel with arterial capillaries and collecting
tubules into which urine ultimately drains from Henle's loops. Concentration
of urine depends on the architecture, regulated water permeability and active
chloride transport in the thin ascending part of Henle's loop. Chloride is
removed from the ascending loop to the medullary interstitium by a site−specific
pump that creates a hyperosmolar environment in the medulla. The hyperosmolality
absorbs water from the descending thin loop, thus increasing osmolality of
the urine before it reaches the ascending loop. The chloride pump only needs
to keep a constant gradient across the capillary wall in the ascending part,
as the counter−current architecture creates a high−osmolar environment
deep in the medulla and a gradual decline towards iso−osmolality at
the upper rim of the medulla. This process alone removes up to 90% of the
primary volume. A further volume reduction and concentration is reached when
the collecting tubules descend down to the medullar papillae through the hyperosmolar
medulla and water is removed from urine through permeable tubular walls. All
water removed from urine is transported away from the medullary tissue by
arterial hairpin loops also extending from cortex down to medulla. The results
obtained by Matsumura et al. suggest that Clck−1 is the key chloride
pump in the ascending part of Henle's loop that helps create the hyperosmolar
gradient in the medulla.
Navigating the nephron. Henle's loop, starting from the cortex,
makes a turn down in the hyperosmolar medulla. Clcnk1 transports chloride
in the thin ascending loop, helping to create the osmolality gradient.
Clcnk1−/− mice look normal, but excrete
a volume of urine fivefold that of their wild−type companions, with
a dramatic reduction in osmolality and consistent with diabetes insipidus.
When Clcnk1−/− mice were refused water
ad libitum, they became lethargic. A vasopressin antagonist failed to
reduce urinary volume or concentrate urine, clinching a diagnosis of murine
nephrogenic diabetes insipidus. Wild−type mice express Clc−k1
only in the ascending part of Henle's loops, consistent with a function
in the chloride pump; as expected, Clcnk1−/−
mice showed no Clc−k1 expression. Finally, transepithelial diffusion
potentials of the thin ascending part of Henle's loop were reduced in the
kidneys of Clcnk1−/− mice compared with
wild−type control—a defect consistent with a defunct (or in this
case, absent) chloride transporter.
These results suggest that CLCNKA and CLCNKB have different
functions, with CLCNKA concentrating the urine—and possibly causing
nephrogenic diabetes insipidus when mutated. CLCNKB, which is expressed
in the thick ascending loop of Henle, reclaims most of the chloride that is
left in the urine and, if mutated, causes Bartter syndrome or renal salt−wasting.
Notably, the job of absorbing chloride back into the body seems to be mediated
to a large extent by a protein belonging to the sulfate permease family encoded
a gene (CLD) which is mutated in congenital chloride diarrhea (Refs 7,8). The study reported by
Fiona Karet and colleagues on page 84 of this
issue implicates another kidney gene (ATP6B1) in sensorineural hearing
loss and renal tubular acidosis9. Pendred syndrome, caused by
mutations in PDS, a gene closely related to CLD, is also characterized
by sensorineural hearing loss10. As CLD is involved in the regulation
of body acid−base balance through chloride absorption and bicarbonate
excretion in the gut, these observations, taken together, raise the intriguing
possibility that PDS may, along with ATP6B1, mediate inner ear chloride and
bicarbonate exchange, and thus cochlear pH.
What about the 'upstream' business of the kidney—the filtration process
that derives primary urine? A recent study11 has identified
a gene whose product may mediate filtration through the podocyte filter, which
sits at the interface of the capillary and the nephron in the glomerulus.
This gene (NPHS1) is mutated in congenital nephrosis, a disorder characterized
by leakage of proteins into urine that normally are retained in the blood.
Its protein (nephrin) appears to be specifically expressed in the periphery
of glomeruli, where primary filtration takes place through slit pores between
podocyte processes. It seems likely that nephrin regulates the sizes of filtrated
molecules, but that's another story, doubtless to be told another time. In
the meantime, one can only hope that improvements in genetic manipulation
will match the pace at which the molecular pieces are being fitted together
to form an increasingly detailed picture.
