Basic Science Articles

Functional Analysis of NBC1 Mutants Associated with

Proximal Renal Tubular Acidosis and Ocular Abnormalities
Shoko Horita,* Hideomi Yamada,* Jun Inatomi,† Nobuo Moriyama,‡ Takashi Sekine,†
Takashi Igarashi,† Yoko Endo,* Majed Dasouki,§ Mesiha Ekim,储 Lihadh Al-Gazali,¶
Mitsunobu Shimadzu,# George Seki,* and Toshiro Fujita*
Departments of *Internal Medicine and †Pediatrics, Faculty of Medicine, Tokyo University, Tokyo, Japan; ‡Department
of Experimental Nursing, Faculty of Nursing, Fukuoka Prefectural University, Fukuoka, Japan; §Genetics Department,
Children’s Mercy Hospital, Kansas City, Missouri; 储Department of Pediatric Nephrology, Ankara University School of
Medicine, Ankara, Turkey; ¶Clinical Genetics, Faculty of Medicine and Health Science, United Arab Emirates
University, Al-Ain, United Arab Emirates; and #Department of Genetics, Mitsubishi Yuka Bio-Clinical Laboratories
Inc., Tokyo, Japan

Mutations in the Naⴙ-HCO3ⴚ co-transporter (NBC1) cause permanent proximal renal tubular acidosis (pRTA) with ocular
abnormalities. However, little has been known about the relationship between the degree of NBC1 inactivation and the
severity of pRTA. This study identified three new homozygous mutations (T485S, A799V, and R881C) in the common coding
regions of NBC1. Functional analysis of these new as well as the known mutants (R298S and R510H) in Xenopus oocytes
revealed a considerable variation in their electrogenic activities. Whereas the activities of R298S, A799V, and R881C were 15
to 40% of the wild-type (WT) activity, T485S and R510H, as a result of poor surface expression, showed almost no activities.
However, T485S, like R510H, had the transport activity corresponding to approximately 50% of the WT activity in ECV304
cells, indicating that surface expression of T485S and R510H varies between the different in vitro cell systems. Electrophysi-
ologic analysis showed that WT, R298S, and R881C all function with 2HCO3ⴚ to 1Naⴙ stoichiometry and have similar
extracellular Naⴙ affinity, indicating that reduction in Naⴙ affinity cannot explain the inactivation of R298S and R881C. These
results, together with the presence of nonfunctional mutants (Q29X and 2311‚A) in other patients, suggest that at least
approximately 50% reduction of NBC1 activity would be required to cause severe pRTA.
J Am Soc Nephrol 16: 2270 –2278, 2005. doi: 10.1681/ASN.2004080667

he Na⫹-HCO3⫺ co-transporter (NBC1) has multiple

T
functions (1,2). Whereas the kidney-type transporter
(kNBC1) plays an essential role in bicarbonate absorp-
tion from renal proximal tubules, the pancreatic-type trans-
porter (pNBC1) is involved in bicarbonate secretion from pan-
creatic duct cells (3– 6). We showed recently that mutations in
NBC1 cause proximal renal tubular acidosis (pRTA) with ocu-
lar abnormalities (7–9). Because NBC1 is widely expressed in
several ocular tissues (10), its inactivation may potentially ex-
plain the occurrence of ocular abnormalities. However, only a
limited number of NBC1 mutations have been identified so far
(7–9,11–13), and several important questions remain unan-
swered. For example, the exact relationship between the degree
of NBC1 inactivation and the severity of pRTA has not been
established. A study in NHE3-deficient mice suggests that acid
secretion from distal tubules is greatly enhanced in pRTA (14).
Received August 13, 2004. Accepted April 28, 2005.
Published online ahead of print. Publication date available at www.jasn.org.
Address correspondence to: Dr. George Seki, Department of Internal Medicine,
Faculty of Medicine, Tokyo University, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033,
Japan. Phone: ⫹81-3-3815-5411; Fax: ⫹81-3-5800-8806; E-mail: georgeseki-tky@
umin.ac.jp
How effectively such compensatory mechanism works in hu-
man, however, is largely unknown. In addition, the effects of
individual mutations on the transport properties of NBC1 have
not been investigated intensively. These issues would be im-
portant to clarify further the molecular mechanism of pRTA as
well as the physiologic roles of NBC1. In our study, we iden-
tified three new missense mutations in kNBC1 from patients
with pRTA and ocular abnormalities. The functional analysis of
these new as well as the known mutants was performed in
Xenopus oocytes and cultured cells.
Materials and Methods
Patients
Patient 1 (T485S) is a boy from consanguineous parents. He had a
history of failure to thrive and received a diagnosis of severe pRTA and
band keratopathy at 2 yr of age. At 3 yr of age, his height (90 cm) and
weight (11.8 kg) both were less than the third percentile. BP was 79/49
mmHg. Bilateral corneas were cloudy, and a red reflex was bilaterally
absent, suggesting the presence of cataracts. Bone survey was normal
with no evidence of rickets. While he was taking sodium bicarbonate
(9.75 g/d), serum analysis revealed Na⫹ 140 mEq/L, K⫹ 3.8 mEq/L,
Cl⫺ 116 mEq/L, creatinine 0.5 mg/dl, and HCO3⫺ 13 mmol/L. He had
a thyroid function test that was initially suggestive of mild hypothy-
roidism. However, during the follow-up, thyroid function tests were

Copyright © 2005 by the American Society of Nephrology ISSN: 1046-6673/1608-2270

J Am Soc Nephrol 16: 2270 –2278, 2005
within normal limits. Liver function tests and renal ultrasound were
normal. Generalized aminoaciduria was not detected, and serum amy-
lase was not measured.
Patient 2 (A799V) is a girl from consanguineous healthy parents, and
her clinical symptoms were previously reported (15). In brief, she
received a diagnosis of motor and mental retardation at the age of 4 yr.
At the age of 14 yr, her height (130 cm) and weight (28 kg) both were
less than the third percentile. Bilateral band keratopathy, cataracts, and
glaucoma were observed in the eye examination. Biochemical investi-
gation revealed the presence of severe pRTA. Rickets, osteosclerosis,
and nephrocalcinosis were not detected. Serum and blood gas analysis
was as follows: Na⫹ 135 mEq/L, K⫹ 1.6 mEq/L, Cl⫺ 115 mEq/L,
creatinine 0.6 mg/dl, magnesium 2.5 mg/dl, calcium 9.1 mg/dl, inor-
ganic phosphorous 4.2 mg/dl, pH 7.143, HCO3⫺ 6.3 mmol/L, and
pCO2 24 mmHg. Urinary protein, glucose, phosphate, and calcium
levels were within normal limits, and aminoaciduria was not detected.
Serum amylase was not measured. When the metabolic acidosis was
severe (pH 7.13, HCO3⫺ 3.6 mmol/L), the urine pH was 5.5.
Patient 3 (R881C) is a girl from consanguineous parents; their parents
are first-degree cousins. She was delivered after a normal pregnancy
but had developmental delay. At the age of 12 yr, her height (130 cm)
was less than the third percentile. She was operated on for glaucoma
and cataracts and had severe pRTA. Serum and blood gas analysis
revealed Na⫹ 141 mEq/L, K⫹ 2.55 mEq/L, Cl⫺ 108 mEq/L, creatinine
0.6 mg/dl, pH 7.142, HCO3⫺ 10.6 mmol/L, and pCO2 25.7 mmHg.
Serum amylase was significantly elevated to 355 IU/L (normal range 25
to 125). There was no glucosuria, but aminoaciduria was not tested.
Genetic Analysis of pRTA Patients
After full informed consent was obtained, RNA was extracted from
the peripheral blood cells of the patients and their parents, and se-
quence analysis of kNBC1 cDNA was performed as described previ-
ously (8). The sequence analysis of entire kNBC1 cDNA of the patients
revealed three different homozygous mutations as follows. In patient 1
(T485S), A to T transversion at nucleotide 1602 was found, resulting in
a change from tyrosine (T) to serine (S) at codon 485. In patient 2
(A799V), C to T transversion at nucleotide 2545 was found, resulting in
a change from alanine (A) to valine (V) at codon 799. In patient 3
(R881C), C to T transversion at nucleotide 2790 was found, resulting in
a change from arginine (R) to cysteine (C) at codon 881. Genomic DNA
sequencing, determined as described previously (8), confirmed that all
of the patients are homozygous and that the parents are heterozygous
for their respective mutations.
Mutagenesis and Oocyte Preparation
The wild-type (WT) human kNBC1-cDNA construct as well as R298S
and R510H mutants were subcloned into an expression vector
pcDNA3.1 as described (9). The mutants A799V and T485S were pro-
duced by QuikChange Site-Directed Mutagenesis kit (Stratagene, La
Jolla, CA). The mutant R881C was produced by GeneEditor in vitro
Site-Directed Mutagenesis System (Promega, Madison, WI). Incorpora-
tion of these mutations was verified by DNA sequencing. Capped RNA
were synthesized using the linearized cDNA templates and the mMES-
SAGE mMACHINE high-yield Capped RNA Transcription kit (Am-
bion, Austin, TX). Oocytes were removed from Xenopus laevis and
dissociated with collagenase as described previously (9) and were
injected with 50 nl of the WT, each mutant kNBC1 cRNA (0.1 g/L each),
or water. Electrophysiologic studies were performed 3 to 5 d after
injection.
NBC1 Mutations and pRTA 2271
Electrophysiologic Analysis
To determine the transport activity of NBC1, we analyzed at least
three batches of oocytes from different Xenopus, and WT and mutants
were analyzed on the same days. Electrophysiologic studies were per-
formed as described (9). In brief, an oocyte was placed in a perfusion
chamber and superfused with a solution at a rate of approximately 4
ml/min. Nominally HCO3-free ND96 solution contained (in mM) 96
NaCl, 2 KCl, 1 MgCl2, 1.8 CaCl2, and 5 mM HEPES at a pH of 7.4.
HCO3⫺-containing solution was prepared by replacing 30 mM NaCl
with 30 mM NaHCO3 in ND96 and was equilibrated with 5% CO2 in
oxygen (pH 7.4). For Na⫹-free HCO3 solution, Na⫹ was substituted by
equimolar concentrations of either N-methyl-D-glucamine (NMDG⫹)
or choline. Solutions with different concentrations of Na⫹ were pre-
pared by mixing HCO3⫺-containing solution and Na⫹-free HCO3⫺
solution. Solution that contained 15 mM NaHCO3 (pH 7.1, for half
reduction of HCO3⫺) was prepared by mixing ND96 solution and
HCO3⫺-containing solution. The membrane potential (Vm) was mea-
sured by a microelectrode, and a two-electrode voltage clamp method
was used to measure the kNBC1-mediated current with a model CEZ-
1200 dual-electrode voltage clamp amplifier (Nihon Kohden, Tokyo,
Japan) and a Duo773 (WPI, Sarasota, FL) amplifier, controlled by the
Clampex module of pCLAMP software (Axon Instruments, Foster City,
CA,). To monitor the NBC1 currents induced by solution changes from
ND96 to HCO3⫺-containing solution, we set a holding potential to ⫺25
mV, because the baseline current was minimal at this voltage when
oocytes were bathed in ND96 solution.
To obtain the current/voltage (I/V) relationship of NBC1 currents,
we set a holding potential to ⫺60 mV. First, oocytes were bathed in
ND96, then solution was changed to HCO3⫺-containing solution. Dur-
ing the subsequent 4 to 5 min, two to three pulse trains were applied
between Vm ⫽ ⫺160 and 60 mV (in steps of 20 mV) with a pulse
duration of 100 ms, and the obtained data were averaged. Thereafter,
0.4 mM 4, 4⬘-diisothiocyanatostilbene-2, 2⬘-disulphonic acid (DIDS;
Sigma, St. Louis, MO) were applied, and the NBC1 currents were
determined from the difference between the I/V relationship in the
presence and absence of DIDS as described (16).
For determining the apparent Na⫹ affinity, extracellular Na⫹ con-
centrations were sequentially reduced from 96 to 48, 24, 12, and 0 mM,
while the NBC1 current was monitored continuously as reported (17).
The current at 0 mM Na⫹ was subtracted from the raw data to obtain
the corrected currents at the individual Na⫹ concentrations. Damping
Gauss Newton Method was used to fit data into the Michaelis-Menten
equation as described (18).
To determine the transport stoichiometries of electrogenic Na⫹-
HCO3⫺ co-transporter, we used membrane diffusion potentials in re-
sponse to sudden changes in Na⫹ or HCO3⫺ concentrations. As dis-
cussed previously (19), zero-current membrane potentials (⌬V)I ⫽ 0 are
described as follows:
冘
k
1 ti
(⌬V)I ⫽ 0⫽⫺ ⌬␮i (1)
F Zi
i⫽1
where I is the applied current, ⌬␮i is the chemical potential differences
of ions i ⫽ 1, , , , , k across a membrane, and F and zi are Farady constant
and valence, and ti is the transference number defined as charge carried
with the net flux of ion i (Ji) in relation to total current (I):
冘
k
ziFJi
ti ⫽ with ti ⫽ 1 (2)
I
t⫽1
2272 Journal of the American Society of Nephrology
We consider now that a membrane contains an obligatory flux coupling
mechanism for one univalent cation (⫹) and anion (⫺) with the stoi-
chiometric ratio q:
J⫺
q ⫽ (3)
J⫹
From equations (1), (2), and (3), we obtain
1 q
t⫹ ⫽ and t⫺ ⫽ (4)
1⫺q q⫺1
From these equations, we can predict that the negative ratio of the
transference numbers equals the stoichiometric ratio:
t⫺
q ⫽ ⫺ (5)
t⫹
Immunofluorescence Analysis
To examine the surface expression of WT and mutants, we performed
the immunofluorescence analysis as described previously (9). In brief, 3
to 4 d after injection of cRNA, the vitelline membranes were removed
carefully, and oocytes were fixed with buffered 4% paraformaldehyde
and incubated as a whole with the diluted (1:200) anti-kNBC1 antibody
that was raised against the N-terminal region (amino acids 4 to 16) of
kNBC1 (10). The oocytes subsequently were incubated with the mixture
of Alexa Fluor 488 goat anti-rabbit IgG (H ⫹ L; Molecular Probes,
Eugene, OR) and tetramethylrhodamine isothiocyanate-conjugated
wheat germ agglutinin for labeling of the plasma membrane (EY Lab-
oratories Inc., San Mateo, CA). Five oocytes for each group then were
observed with a confocal laser scanning microscope (Radiance 2100/
K-2; Japan Bio-Rad Laboratories, Tokyo, Japan).
Cell pH Measurement in Oocytes
Cell pH (pHi) in oocytes was measured with pH-sensitive dye
bis(carboxyethyl)carboxyfluorescein (BCECF; Molecular Probes). Four
to 5 d after cRNA injection, oocytes were incubated with 20 ␮M
acetoxymethylester form of BCECF for 20 to 30 min, then transferred to
a perfusion chamber mounted on an inverted microscope (IMT-2;
Olympus, Tokyo, Japan) and were superfused with ND96 solution at a
rate of approximately 4 ml/min. pHi was measured with a microscopic
fluorescence photometry system (OSP-10; Olympus) at room tempera-
ture as described (10). The intracellular dye was alternatively exited at
two wavelengths (440 and 490 nm), and emission was measured at a
wavelength of 530 nm. Autofluorescence of oocytes was measured at
the beginning of an experimental day, and these values were subtracted
from the raw data. To obtain the calibration curve for pHi, we exposed
oocytes to HEPES-buffered solution that contained 90 mM K⫹ and 40
␮M nigericin. Solution pH was adjusted at the different levels (from 6.4
to 7.8) with 1 N NaOH, whereas Na⫹ concentration in solution was
kept at 10 mM. For measuring NBC1 activity, solution was switched
from nominally HCO3-free ND96 solution to HCO3⫺-containing solu-
tion equilibrated with 5% CO2 in oxygen. NBC1 activity was measured
as the rate of pHi recovery from acidification as described (20).
Functional Analysis in ECV304 Cells
Functional analysis of NBC1 activity in ECV304 cells was performed
as described previously (6,7). In brief, cells were seeded onto coverslips
and transfected with WT or T485S using LipofectAMINE 2000 (Life
Technologies Inc., Grand Island, NY) according to the manufacturer’s
instruction. Two days after transfection, pHi measurement was per-
formed with BCECF at 37°C. The coverslip was superfused for 30 min
at approximately 5 ml/min with prewarmed Cl⫺-free HCO3⫺ solution
J Am Soc Nephrol 16: 2270 –2278, 2005
that contained (in mM) 144 Na⫹, 5 K⫹, 5 Ca2⫹, 1 Mg2⫹, 132 gluconate⫺,
2 H2PO4⫺, 1 SO42⫺, 25 HCO3⫺, and 5.5 glucose equilibrated with 5%
CO2 in oxygen (pH 7.4). Thereafter, pHi measurement was started with
a microscopic fluorescence photometry system (OSP-10). We used 10
␮M nigericin to acidify the cells. This method has been shown to induce
a more stable and predictable acidification than the NH4Cl-pulse tech-
nique, especially in Cl⫺-free solution (6,7). Because Na⫹/H⫹ exchange
activity assayed by the nigericin acidification was identical to that
assayed by the NH4Cl-pulse technique (6,7), the problem of proton leak
that might be caused by residual nigericin has been shown to be
minimal, if any, in our system. After cell acidification by nigericin, the
coverslip was superfused with Na⫹-free, Cl⫺-free HCO3⫺ solution
without nigericin (Na⫹ was replaced with equimolar concentrations of
NMDG⫹). Thereafter, pHi recovery was induced by Cl⫺-free HCO3⫺
solution that contained 1 mM amiloride that completely inhibits an
endogenous Na⫹/H⫹ exchanger (6,7).
We computed HCO3⫺ fluxes as the product of rate of pHi recovery
and total intracellular buffer capacity (␤T). ␤T is the sum of the buffer
capacity as a result of CO2/HCO3⫺ (␤HCO3⫺) and intrinsic buffer ca-
pacity (␤I). We computed the theoretical ␤HCO3⫺ as 2.3 ⫻ [HCO3⫺]. To
determine ␤I, we measured the pHi increase induced by increasing the
extracellular concentrations of the weak base NH3 as described (21). In
brief, the cells were superfused with HEPES-buffered standard solution
that contained (in mM) 144 Na⫹, 5 K⫹, 5 Ca2⫹, 1 Mg2⫹, 137 Cl⫺, 2
H2PO4⫺, 1 SO42⫺, 25 HCO3⫺, and 5.5 glucose. The cells were acid-
loaded using a solution that contained 20 mM NH3/NH4⫹ (pH 7.4).
Washout of NH3/NH4⫹ typically decreased pHi to approximately 6.4.
Thereafter, the cells were exposed sequentially to solutions that con-
tained 1, 2.5, 5, and 10 mM NH3/NH4⫹ in the presence of 1 mM
amiloride. Because the dissociation constant (pKa) for NH4⫹ is approx-
imately 8.9 at 37°C, the calculated concentrations of NH3 of these
solutions are 31, 77, 153, and 307 ␮M, respectively.
Western Blot Analysis
Membrane-enriched fractions were collected from oocytes and
ECV304 cells as described (6,9). After being dissolved in a sample-
loading buffer that contained 2% SDS, the protein samples were sepa-
rated by SDS-PAGE on 7% acrylamide minigels and transferred to a
nitrocellulose membrane. After incubation in a blocking buffer, the
membrane was treated with the diluted anti-NBC1 antibody that was
raised against the C-terminal region (amino acids 928 to 1035) of
kNBC1 (Chemicon, Temecula, CA) or the anti-kNBC1 antibody (10)
and then with horseradish peroxidase– conjugated anti-rabbit IgG (Bio-
Rad, Richmond, CA) as the secondary antibody. The signal was de-
tected by an ECL Plus system (Amersham, Aylesbury, UK).
Statistical Analyses
The data were represented as mean ⫾ SEM. Significant differences
were determined by applying the paired or unpaired t test as appro-
priate.
Results
Activities of WT and Mutants in Oocytes
We first performed the electrophysiologic analysis in oocytes.
In addition to the three new mutants (T485S, A799V, and
R881C) identified in our study, we examined the known mu-
tants (R298S and R510H), because their transport activities were
previously analyzed in the cultured cell system (7). As shown
in Figure 1, the addition of HCO3⫺/CO2 in oocytes that were
injected with WT induced a large hyperpolarization, consistent
with the transport activity of NBC1 (9,17), but the hyperpolar-
J Am Soc Nephrol 16: 2270 –2278, 2005
Figure 1. Changes in membrane potential (Vm) in response to
solution change from ND96 to HCO3⫺-containing solution.
Typical traces are shown for an oocyte that was injected with
H2O (Cont), R298S mutant (R298S), or wild-type kNBC1 (WT).
Note that hyperpolarization is slightly smaller in R298S than in
WT.
ization in R298S was slightly less than that in WT. Most of the
mutants showed significant hyperpolarization in response to
HCO3⫺/CO2, and the average responses (n ⫽ 8 for each) were
⫺61.1 ⫾ 2.2 mV (R298S), ⫺24.4 ⫾ 2.4 mV (R510H), ⫺40.1 ⫾ 8.1
mV (A799V), and ⫺59.5 ⫾ 4.1 mV (R881C). However, these
changes were significantly (P ⬍ 0.05) smaller than that in WT
(⫺88.1 ⫾ 4.7 mV; n ⫽ 10). By contrast, Vm change in T485S
(⫺3.8 ⫾ 0.8 mV; n ⫽ 8) was not statistically different from that
in control oocytes that were injected with H2O (⫺3.9 ⫾ 0.7 mV;
n ⫽ 8). A small hyperpolarization in response to HCO3⫺/CO2,
observed in control oocytes in this series, was not constantly
observed in other series of experiments and might reflect some
variations of oocytes conditions.
To determine the relative activities of mutants, we performed
the two-electrode voltage clamp experiments. In WT but not in
control oocytes that were injected with H2O, the addition of
HCO3⫺/CO2 induced a large negative charge influx into oo-
cytes (0.40 ⫾ 0.06 ␮A; n ⫽ 10), which originates from NBC1
(9,17). As summarized in Figure 2A, the NBC1 currents, deter-
mined at a holding potential of ⫺25 mV, were significantly
reduced in all of the mutants. The relative activities of R298S,
R510H, A799V, and R881C, calculated by the mean currents,
were 41 ⫾ 3, 4 ⫾ 0.4, 14 ⫾ 2, and 39 ⫾ 4% of the WT activity,
respectively. However, T485S showed no current (n ⫽ 8), con-
firming the results of Vm measurement.
To compare further the activities of WT and mutants, we
determined the I/V relationship by applying step pulses be-
tween Vm ⫽ ⫺160 and 60 mV. As shown in Figure 2B, the I/V
relationship of WT, determined as the difference between the
presence and the absence of DIDS, was linear within the tested
voltage range, and a reversal potential (Erev) was ⫺92 ⫾ 3 mV
(n ⫽ 5). These properties of WT are similar to those reported in
the previous studies (13,17). Although R298S and R881C
showed significantly attenuated currents compared with WT,
the I/V relationship of these mutants was similar to that of WT
as also shown in Figure 2B. In addition, Erev of R298S (⫺94.0 ⫾
4 mV; n ⫽ 4) and R881C (⫺96 ⫾ 6 mV; n ⫽ 4) were statistically
not different from that of WT. However, R510H and A799V
showed only very low currents in the tested voltage range,
which were not sufficient to yield the reliable I/V relationship,
NBC1 Mutations and pRTA 2273
Figure 2. (A) NBC1-mediated currents of WT and mutants
(R298S, R510H, A799V, and R881C) in oocytes. Influxes of
anion charges into oocytes, induced by solution change from
ND-96 to HCO3⫺-containing solution, were determined by the
two-electrode voltage clamp method at a holding potential of
⫺25 mV. Numbers of observation were 10 for WT and eight for
each mutant. *P ⬍ 0.01 versus WT. (B) Current/voltage (I/V)
relationship of WT, R298S, R881C, and control (Cont). Step
pulses between Vm ⫽ ⫺160 and 60 mV were applied in the
presence and the absence of 4, 4⬘-diisothiocyanatostilbene-2,
2⬘-disulphonic acid (DIDS; 0.4 mM). Numbers of observation
were six for WT and four for R298S, R881C, and Cont.
whereas T485S showed no currents at all in the entire voltage
range (data not shown).
Surface Expression in Oocytes
To examine the surface expression of NBC1, we performed
immunofluorescence analysis. As shown in Figure 3, an intense
surface labeling of NBC1 was detected in WT, R298S, A799V,
and R881C. However, only a faint patchy labeling was detected
in T485S and R510H. To obtain a more quantitative estimate of
the protein expression, we performed Western blot analysis
using the membrane-enriched fractions. As shown in Figure 4,
a prominent band of approximately 135 kD, corresponding to
the full-length NBC1 (9), was detected in WT. Although the
intensity was somehow less than in WT, an intense band was
also detected in R298S, A799V, and R881C. However, only a
faint band was detectable in T485S and R510H, consistent with
the results of immunofluorescence analysis.
pHi Measurement in Oocyte
The electrophysiologic analysis suggests that T485S, unlike
the other mutants, has no transport activity in oocytes. To
confirm this view, we performed pHi measurement in oocytes.
As shown in Figure 5, the exposure to HCO3⫺/CO2 induced a
large intracellular acidification as a result of CO2 influx. There-
2274 Journal of the American Society of Nephrology
Figure 3. Surface expression of WT and mutants in oocytes.
NBC1 was detected by the antibody against N-terminal region
of kNBC1 (shown as green), whereas the plasma membrane
was detected by wheat germ agglutinin (WGA; shown as red).
Merge images are shown only for WT and R298S. Note the
absence of NBC1 labeling in control oocytes (Cont). Bars ⫽ 100
␮m.
Figure 4. Western blot analysis of oocyte membrane-enriched
fractions. NBC1 was detected by the antibody against the C-
terminal region of NBC1. Approximately 30 ␮g of protein was
loaded for each lane. Lane 1, control (H2O); lane 2, WT; lane 3,
R298S; lane 4, T485S; lane 5, R510H; lane 6, A799V; lane 7,
R881C.
after, pHi in WT gradually recovered as a result of NBC1-
mediated HCO3⫺ influx (20). The average rate of pHi recovery
in WT was 0.030 ⫾ 0.007 pH unit/min (n ⫽ 7). However, the
rate of pHi recovery in T485S (0.003 ⫾ 0.001 pH unit/min; n ⫽
7) was negligible and not significantly different from that in
control oocytes that were injected with H2O (0.004 ⫾ 0.001 pH
unit/min; n ⫽ 7). As also shown in Figure 5, the exposure to
HCO3⫺/CO2 induced less acidification in WT. Thus, the base-
line pHi in ND96 was similar in WT (7.48 ⫾ 0.03), T485S (7.47 ⫾
0.02), and H2O (7.46 ⫾ 0.03). However, the minimal pHi after
the exposure to HCO3⫺/CO2 in WT (7.08 ⫾ 0.03) was signifi-
cantly (P ⬍ 0.05) higher than that in T485S (6.83 ⫾ 0.03) or H2O
(6.84 ⫾ 0.03), which could be due to the NBC1 activity operat-
ing only in WT. These results confirmed that T485S induced no
measurable transport activity of NBC1 in oocytes.
Functional Analysis of T485S in ECV304 Cells
The results thus far indicate that the poor surface expression
of T485S and R510H precludes the reliable analysis in oocytes.
J Am Soc Nephrol 16: 2270 –2278, 2005
Figure 5. Cell pH (pHi) measurement in oocytes that were
injected with WT, T485S, or H2O (Cont). Solution was changed
from ND96 to HCO3⫺-containing solution (HCO3⫺). Because
the latter solution was equilibrated with 5% CO2, the influx of
CO2 induced large intracellular acidification. Note that signif-
icant pHi recovery was evident only in WT. Note also that the
minimum pHi level was higher in WT than in T485S or Cont.
However, our previous analysis in ECV304 cells showed that
the transport activity of R510H was as much as approximately
50% of the WT activity (7). This prompted us to examine the
transport activity of T485S in ECV304 cells. The cells were
acidified by nigericin in Cl⫺-free HCO3⫺ solution and were
subsequently superfused with Cl⫺-free, Na⫹-free HCO3⫺ solu-
tion as described previously (7). As shown in Figure 6, the
baseline pHi in Cl⫺-free HCO3⫺ solution and the rate of acid-
ification by nigericin were very similar in WT, T485S, and
vector. Furthermore, there was no significant difference in the
minimum pHi values before Na⫹ readdition: WT (6.72 ⫾ 0.05;
n ⫽ 9), T485S (6.70 ⫾ 0.05; n ⫽ 9), and vector (6.74 ⫾ 0.06; n ⫽
9). As also shown in Figure 6, the addition of Na⫹ induced a
prompt pHi recovery in WT, reflecting the NBC1 activity (6,7).
A modest pHi recovery was also detected in T485S, whereas
pHi recovery was negligible in vector alone. In separate exper-
iments, we confirmed that pHi recovery in both WT and T485S
was completely inhibited by 0.3 mM DIDS (n ⫽ 5 for each). The
Figure 6. pHi measurement in ECV304 cells that were trans-
fected with WT, T485S, or vector alone. The cells first were
superfused with Cl⫺-free HCO3⫺ solution (Cl-free) for 30 min
to reduce intracellular Cl⫺ (6). Nigericin (10 ␮M) was added to
induce acidification, then solution was changed to Na⫹-free
Cl⫺-free HCO3⫺ solution (Na-free Cl-free). Thereafter, acidified
cells were exposed to Na⫹-containing solution in the presence
of amiloride (Amil).
J Am Soc Nephrol 16: 2270 –2278, 2005
average rate of DIDS-sensitive pHi recovery was 0.102 ⫾ 0.009,
0.053 ⫾ 0.009, and 0.006 ⫾ 0.002 pH unit/min in WT, T485S,
and vector, respectively (n ⫽ 9 for each). The intrinsic buffer
capacity ␤I, determined at the pHi level of 6.70 to 6.75 as
described in the Materials and Methods section, was similar in
WT (16.1 ⫾ 2.4 mM/pH unit; n ⫽ 7), T485S (16.4 ⫾ 2.3 mM/pH
unit; n ⫽ 6), and vector (14.5 ⫾ 4.3 mM/pH unit; n ⫽ 6). The
total buffer capacity ␤T was also similar in WT (28.5 mM/pH
unit), T485S (27.9 mM/pH unit), and vector (27.3 mM/pH
unit). Accordingly, Na⫹-induced HCO3⫺ influxes, calculated
from the rate of pHi recovery and ␤T, were 2.9 ⫾ 0.3 mM/min
for WT, 1.5 ⫾ 0.3 mM/min for T485S, and 0.2 ⫾ 0.05 mM/min
for vector. Western blot analysis on the membrane-enriched
fractions showed that the protein expression was comparable in
WT and T485S (Figure 7). These results indicate that T485S,
when analyzed in ECV304 cells, has the similar expression
efficiency as WT and displays approximately 50% transport
activity of that of WT.
Stoichiometry and Na⫹ Affinity
We next tried to examine the effects of mutations on the
detailed transport properties of NBC1 in oocytes. For the mu-
tants, we selected only R298S and R881C, because the activities
of the other mutants were very low or absent and were unlikely
to yield reliable results. The transport stoichiometry was deter-
mined by comparing the initial Vm changes in response to half
reduction of extracellular HCO3⫺ or Na⫹ concentrations as
described (22,23). In oocytes expressing NBC1, the Vm changes
were shown to be fast enough compared with the changes in
intracellular ion concentrations (17), supporting the validity of
this approach (24). In addition, as shown in Figure 8, the
depolarizing responses to HCO3⫺ or Na⫹ reduction were al-
most completely inhibited by 0.4 mM DIDS. Furthermore, the
depolarizing responses to HCO3⫺ or Na⫹ reduction were com-
pletely absent in control oocytes. By comparing the DIDS-
sensitive changes in membrane potential (⌬Vm) in response to
HCO3⫺ or Na⫹ reduction, we obtained estimated values of the
stoichiometry as shown in Table 1. Our data indicate that R298S
and R881C, like WT (16,17), function with 2HCO3⫺ to 1Na⫹
stoichiometry in oocytes.
For determining the apparent Na⫹ affinity, the current was
monitored continuously by the two-electrode voltage clamp
method, whereas extracellular Na⫹ concentrations were
changed sequentially from 96 mM to 48, 24, 12, and 0 mM. By
subtracting the current at 0 mM Na⫹ from raw data and fitting
these data into the Michaelis-Menten equation through nonlin-
Figure 7. Western blot analysis of ECV304 cell membrane–
enriched fractions. NBC1 was detected by the antibody against
the C-terminal region of NBC1. Approximately 20 ␮g of protein
was loaded for each lane. Lane 1, vector; lane 2, WT; lane 3,
T485S.
NBC1 Mutations and pRTA 2275
Figure 8. Changes in membrane potential to extracellular
HCO3⫺ or Na⫹ reduction. Oocytes were superfused with
HCO3⫺-containing solution, and HCO3⫺ or Na⫹ concentra-
tions were repeatedly reduced by half as indicated by bars in
the absence and the presence of DIDS. Solution pH was
reduced from 7.4 to 7.1 in case of HCO3⫺ reduction. Note
that depolarizing responses as a result of HCO3⫺ or Na⫹
reduction were almost completely inhibited by DIDS. A rep-
resentative response in oocyte that was injected with WT is
shown.
Table 1. DIDS-sensitive changes in membrane potentials
in response to HCO3⫺ or Na⫹ reductiona
⌬Vm to 1/2 ⌬Vm to 1/2
HCO3⫺ Na⫹ Ratio
(mV) (mV)
WT 20.5 ⫾ 1.0 10.7 ⫾ 0.5 1.93 ⫾ 0.04
R298S 13.4 ⫾ 2.0 6.9 ⫾ 1.2 2.00 ⫾ 0.11
R881C 16.6 ⫾ 1.6 8.2 ⫾ 1.1 2.04 ⫾ 0.16
a
Ratio was calculated by the individual membrane
potential changes on the same oocytes. Numbers of
observation are six (WT), five (R298S), and five (R881C).
DIDS, 4, 4⬘-diisothiocyanatostilbene-2, 2⬘-disulphonic acid;
Vm, change in membrane potentials.
ear regression analysis, we obtained an individual curve for
WT and the mutants as shown in Figure 9. The calculated Km
values of R298S (29 mM) and R881C (26 mM) were not statis-
tically different from that of WT (29 mM).
Discussion
This study identified the three new homozygous muta-
tions (T485S, A799V, and R881C) in the common coding
regions of NBC1 from patients with pRTA and ocular abnor-
malities. Functional analysis of these new as well as the two
known mutants (R298S and R510H) suggests that at least
approximately 50% reduction in the transport activity of
NBC1 may be required to cause severe pRTA, which is
defined as reduction of blood HCO3⫺ concentration to ⬍50%
of the normal range (or less than approximately 13 mM). The
relative activities of R298S, R510H, A799V, and R881C, as
determined by the NBC1-mediated currents in Xenopus oo-
cytes, were 41, 4, 14, and 39% of the WT activity, respec-
2276 Journal of the American Society of Nephrology
Figure 9. Apparent Na⫹ affinity for WT, R298S, and R881C. E,
mean data at 12, 24, 48, and 96 mM Na⫹ with SEM. A holding
potential was set at ⫺25 mV, and currents at 0 mM Na⫹ were
subtracted from raw data to yield the individual data. Fitting
was performed as described in the text. Numbers of observa-
tions were six for WT and five for R298S and R881C.
tively. Whereas T485S showed no electrogenic activity in
oocytes, its transport activity was approximately 50% of the
WT activity in ECV304 cells. T485S, which lies at the inner
surface of the third transmembrane domain (25), is of special
interest because the corresponding codon in the electroneu-
tral anion exchangers (AEs1 to 3) is serine (2). However, the
results of pHi measurement confirmed that T485S induced
no measurable transport activity in oocytes. These results
together with our previous findings (7) indicate that the
surface expression of T485S and R510H is very poor in
oocytes but well preserved in the cultured cells. The results
of immunofluorescence and Western blot analysis also sup-
ported this view. It has been known that trafficking behavior
of mutant proteins varies between different in vitro cell sys-
tems (26). Future studies will be required to examine
whether the NBC1 mutants can be expressed in the proper
cell surface in renal epithelial cells.
By comparing membrane potentials to sudden reduction in
Na⫹ or HCO3⫺ concentrations on the same oocytes, we
showed that WT, R298S, and R881C all functioned with
2HCO3⫺ to 1Na⫹ stoichiometry. The depolarizing responses
to Na⫹ or HCO3⫺ reduction, which were absent in control
oocytes, were completely inhibited by DIDS. We could not
exclude a possibility that the Vm changes by HCO3⫺ reduc-
tion may have a small systematic error originating from the
reduction in extracellular pH. However, the similar Erev
values of WT, R298S, and R881C, determined by the I/V
relationship, also supported our conclusion. Although
kNBC1 is thought to function with 3HCO3⫺ to 1Na⫹ stoichi-
ometry in renal proximal tubules in vivo (27), recent studies
indicate that the stoichiometry is not fixed but varies accord-
ing to cell conditions or cell types (19,23,28,29). In oocytes,
kNBC1 was reported to function with 2HCO3⫺ to 1Na⫹
stoichiometry (16,17), consistent with our analysis. We also
showed that R298S and R881C have similar Na⫹ affinity as
WT. Sciortino and Romero (17) reported that a Km value of
rat kNBC1 is approximately 30 mM. Our results indicate that
human kNBC1 has a very similar Km value. They further
J Am Soc Nephrol 16: 2270 –2278, 2005
indicate that the reduction in extracellular Na⫹ affinity can-
not explain the inactivation of R298S and R881C.
Previously we reported two other kNBC1 mutations: Q29X
(8) and a single nucleotide deletion, 2311‚A (9), both of
which would result in nonfunctional proteins. However,
blood HCO3⫺ concentrations reported in the patients with
Q29X (9.4 mM) and 2311‚A (13.2 mM) are very similar to the
concentrations in the patients with functional mutants. We
do not known the exact reason for why there is no tight
relationship between the degree of NBC1 inactivation and
the severity of pRTA. Probably the differences in environ-
mental factors or the ability of distal nephron compensation
may be involved. Alternatively, some functional mutants,
through the improper trafficking, may have additional dete-
riorating effects on the acid-base homeostasis as shown pre-
viously for AE1 (26).
Typically, patients with mutations in the common coding
regions of NBC1 have band keratopathy, cataracts, and glau-
coma (9,11). The new patients identified in this study also
presented most of these ocular abnormalities, confirming the
important roles of NBC1 in several ocular tissues (10,30).
Unlike the Q29X mutation that inactivates only kNBC1 (8),
mutations in the common coding regions of NBC1 will affect
the functions of both kNBC1 and pNBC1 (6), which originate
from the same SLC4A4 gene (31). Because pNBC1 may me-
diate bicarbonate uptake into pancreatic duct cells, its inac-
tivation could theoretically interfere with pancreatic secre-
tion as discussed previously (6,32,33). In addition, NBC1 is
known to be widely expressed in brain and may play critical
roles during the later stages of brain development (34,35).
However, previous studies (7–9) as well as ours indicate that
abnormal pancreatic function tests or mental retardation are
found only in some patients. In particular, a patient who had
2311‚A mutation did not present mental retardation, al-
though this mutation was thought to result in the complete
loss of function (9). For clarifying the molecular mechanism
of these rare phenotypes, more studies will be warranted to
investigate the physiologic roles of NBC1 in several tissues,
such as pancreas and brain.
Acknowledgments
This study was supported by the Ministry of Education, Science and
Culture of Japan (grants 14571013 and 16590779).
We thank the patients and their families for participation, Dr. Masa-
nori Taira (Department of Biologic Sciences, Tokyo University) for
providing a facility for Xenopus laevis, and Dr. Eberhard Frömter (Zen-
trum der Physiology, JW Goethe University) for helpful discussion. The
anti-kNBC1 antibody was supplied by Kumamoto Immunochemical
Laboratory Co. (Kumamoto, Japan).
References
1. Romero MF, Hediger MA, Boulpaep EL, Boron WF:
Expression cloning and characterization of a renal elec-
trogenic Na⫹/HCO3⫺ cotransporter. Nature 387: 409 –
413, 1997
2. Romero MF, Boron WF: Electrogenic Na⫹/HCO3⫺ co-
J Am Soc Nephrol 16: 2270 –2278, 2005
transporters: Cloning and physiology. Annu Rev Physiol
61: 699 –723, 1999
3. Burnham CE, Amlal H, Wang Z, Shull GE, Soleimani M:
Cloning and functional expression of a human kidney
Na⫹:HCO3⫺ cotransporter. J Biol Chem 272: 19111–19114,
1997
4. Yamada H, Yamazaki S, Moriyama N, Hara C, Horita S,
Enomoto Y, Kudo A, Kawakami H, Tanaka Y, Fujita T,
Seki G: Localization of NBC-1 variants in human kidney
and renal cell carcinoma. Biochem Biophys Res Commun
310: 1213–1218, 2003
5. Abuladze N, Lee I, Newman D, Hwang J, Boorer K,
Pushkin A, Kurtz I: Molecular cloning, chromosomal
localization, tissue distribution, and functional expres-
sion of the human pancreatic sodium bicarbonate co-
transporter. J Biol Chem 273: 17689 –17695, 1998
6. Satoh H, Moriyama N, Hara C, Yamada H, Horita S,
Kunimi M, Tsukamoto K, Iso ON, Inatomi J, Kawakami
H, Kudo A, Endou H, Igarashi T, Goto A, Fujita T, Seki
G: Localization of Na⫹-HCO3⫺ cotransporter (NBC-1)
variants in rat and human pancreas. Am J Physiol Cell
Physiol 284: C729 –C737, 2003
7. Igarashi T, Inatomi J, Sekine T, Cha SH, Kanai Y, Kunimi
M, Tsukamoto K, Satoh H, Shimadzu M, Tozawa F, Mori
T, Shiobara M, Seki G, Endou H: Mutations in SLC4A4
cause permanent isolated proximal renal tubular acido-
sis with ocular abnormalities. Nat Genet 23: 264 –266,
1999
8. Igarashi T, Inatomi J, Sekine T, Seki G, Shimadzu M,
Tozawa F, Takeshima Y, Takumi T, Takahashi T, Yo-
shikawa N, Nakamura H, Endou H: Novel nonsense
mutation in the Na⫹/HCO3⫺ cotransporter gene
(SLC4A4) in a patient with permanent isolated proximal
renal tubular acidosis and bilateral glaucoma. J Am Soc
Nephrol 12: 713–718, 2001
9. Inatomi J, Horita S, Braverman N, Sekine T, Yamada H,
Suzuki Y, Kawahara K, Moriyama N, Kudo A,
Kawakami H, Shimadzu M, Endou H, Fujita T, Seki G,
Igarashi T: Mutational and functional analysis of
SLC4A4 in a patient with proximal renal tubular acido-
sis. Pflugers Arch 448: 438 – 444, 2004
10. Usui T, Hara M, Satoh H, Moriyama N, Kagaya H,
Amano S, Oshika T, Ishii Y, Ibaraki N, Hara C, Kunimi
M, Noiri E, Tsukamoto K, Inatomi J, Kawakami H, En-
dou H, Igarashi T, Goto A, Fujita T, Araie M, Seki G:
Molecular basis of ocular abnormalities associated with
proximal renal tubular acidosis. J Clin Invest 108: 107–
115, 2001
11. Igarashi T, Sekine T, Inatomi J, Seki G: Unraveling the
molecular pathogenesis of isolated proximal renal tubu-
lar acidosis. J Am Soc Nephrol 13: 2171–2177, 2002
12. Alper SL: Genetic diseases of acid-base transporters.
Annu Rev Physiol 64: 899 –923, 2002
13. Dinour D, Chang MH, Satoh J, Smith BL, Angle N,
Knecht A, Serban I, Holtzman EJ, Romero MF: A novel
missense mutation in the sodium bicarbonate cotrans-
porter (NBCe1/SLC4A4) causes proximal tubular acido-
sis and glaucoma through ion transport defects. J Biol
Chem 279: 52238 –52246, 2004
14. Schultheis PJ, Clarke LL, Meneton P, Miller ML, So-
leimani M, Gawenis LR, Riddle TM, Duffy JJ,
Doetschman T, Wang T, Giebisch G, Aronson PS, Lorenz
NBC1 Mutations and pRTA 2277
JN, Shull GE: Renal and intestinal absorptive defects in
mice lacking the NHE3 Na⫹/H⫹ exchanger. Nat Genet
19: 282–285, 1998
15. Deda G, Ekim M, Guven A, Karagol U, Tumer N: Hy-
popotassemic paralysis: A rare presentation of proximal
renal tubular acidosis. J Child Neurol 16: 770 –771, 2001
16. Heyer M, Müller-Berger S, Romero MF, Boron WF,
Frömter E: Stoichiometry of the rat kidney Na⫹-HCO3⫺
cotransporter expressed in Xenopus laevis oocytes.
Pflugers Arch 438: 322–329, 1999
17. Sciortino CM, Romero MF: Cation and voltage depen-
dence of rat kidney electrogenic Na⫹-HCO3⫺ cotrans-
porter, rkNBC, expressed in oocytes. Am J Physiol Renal
Physiol 46: F611–F623, 1999
18. Morita N, Kusuhara H, Sekine T, Endou H, Sugiyama Y:
Functional characterization of rat organic anion trans-
porter 2 in LLC-PK1 cells. J Pharmacol Exp Ther 298:
1179 –1184, 2001
19. Seki G, Coppola S, Yoshitomi K, Burckhardt BC, Samar-
zija I, Müller-Berger S, Frömter E: On the mechanism of
bicarbonate exit from renal proximal tubular cells. Kid-
ney Int 49: 1671–1677, 1996
20. Xu J, Wang Z, Barone S, Petrovic M, Amlal H, Conforti
L, Petrovic S, Soleimani M: Expression of the Na⫹-
HCO3⫺ cotransporter NBC4 in rat kidney and charac-
terization of a novel NBC4 variant. Am J Physiol Renal
Physiol 284: F41–F50, 2003
21. Bevensee MO, Weed RA, Boron WF: Intracellular pH
regulation in cultured astrocytes from rat hippocampus.
I. Role of HCO3. J Gen Physiol 110: 453– 465, 1997
22. Seki G, Coppola S, Frömter E: The Na⫹-HCO3⫺ cotrans-
porter operates with a coupling ratio of 2 HCO3⫺ to 1
Na⫹ in isolated rabbit renal proximal tubule. Pflugers
Arch 425: 409 – 416, 1993
23. Müller-Berger S, Nesterov VV, Frömter E: Partial recov-
ery of in vivo function by improved incubation condi-
tions of isolated renal proximal tubule. II. Change of
Na-HCO3 cotransport stoichiometry and of response to
acetazolamide. Pflugers Arch 434: 383–391, 1997
24. Burckhardt BC, Cassola AC, Frömter E: Electrophysio-
logical analysis of bicarbonate permeation across the
peritubular cell membrane of rat kidney proximal tu-
bule. II. Exclusion of HCO3⫺-effects on other ion per-
meabilities and of coupled electroneutral HCO3⫺-trans-
port. Pflugers Arch 401: 43–51, 1984
25. Alvarez BV, Loiselle FB, Supuran CT, Schwartz GJ, Ca-
sey JR: Direct extracellular interaction between carbonic
anhydrase IV and the human NBC1 sodium/bicarbonate
co-transporter. Biochemistry 42: 12321–12329, 2003
26. Devonald MA, Karet FE: Renal epithelial traffic jams and
one-way streets. J Am Soc Nephrol 15: 1370 –1381, 2004
27. Yoshitomi K, Burckhardt BC, Frömter E: Rheogenic so-
dium-bicarbonate cotransport in the peritubular cell
membrane of rat renal proximal tubule. Pflugers Arch
405: 360 –366, 1985
28. Gross E, Hawkins K, Abuladze N, Pushkin A, Cotton
CU, Hopfer U, Kurtz I: The stoichiometry of the electro-
genic sodium bicarbonate cotransporter NBC1 is cell-
type dependent. J Physiol 531: 597– 603, 2001
29. Kunimi M, Seki G, Hara C, Taniguchi S, Uwatoko S,
Goto A, Kimura S, Fujita T: Dopamine inhibits renal
Na⫹:HCO3⫺ cotransporter in rabbits and normotensive
2278 Journal of the American Society of Nephrology
rats but not in spontaneously hypertensive rats. Kidney
Int 57: 534 –543, 2000
30. Usui T, Seki G, Amano S, Oshika T, Miyata K, Kunimi M,
Taniguchi S, Uwatoko S, Fujita T, Araie M: Functional
and molecular evidence for Na⫹-HCO3⫺ cotransporter
in human corneal endothelial cells. Pflugers Arch 438:
458 – 462, 1999
31. Abuladze N, Song M, Pushkin A, Newman D, Lee I,
Nicholas S, Kurtz I: Structural organization of the hu-
man NBC1 gene: kNBC1 is transcribed from an alterna-
tive promoter in intron 3. Gene 251: 109 –122, 2000
32. Thévenod F, Roussa E, Schmitt BM, Romero MF: Clon-
ing and immunolocalization of a rat pancreatic Na⫹
bicarbonate cotransporter. Biochem Biophys Res Commun
264: 291–298, 1999
J Am Soc Nephrol 16: 2270 –2278, 2005
33. Shumaker H, Amlal H, Frizzell R, Ulrich CD 2nd, So-
leimani M: CFTR drives Na⫹-nHCO3⫺ cotransport in
pancreatic duct cells: A basis for defective HCO3⫺ se-
cretion in CF. Am J Physiol Cell Physiol 276: C16 –C25,
1999
34. Schmitt BM, Berger UV, Douglas RM, Bevensee MO,
Hediger MA, Haddad GG, Boron WF: Na/HCO3 co-
transporters in rat brain: Expression in glia, neurons,
and choroid plexus. J Neurosci 20: 6839 – 6848, 2000
35. Giffard RG, Papadopoulos MC, van Hooft JA, Xu L,
Giuffrida R, Monyer H: The electrogenic sodium bicar-
bonate cotransporter: Developmental expression in rat
brain and possible role in acid vulnerability. J Neurosci
20: 1001–1008, 2000