Download invited review

Am J Physiol Renal Physiol
280: F10–F18, 2001.
invited review
Renal Na⫹-glucose cotransporters
ERNEST M. WRIGHT
Department of Physiology, University of California Los Angeles School of Medicine,
Los Angeles, California 90095-1751
Wright, Ernest M. Renal Na⫹-glucose cotransporters. Am J Physiol
Renal Physiol 280: F10–F18, 2001.—In humans, the kidneys filter ⬃180
g of D-glucose from plasma each day, and this is normally reabsorbed in
the proximal tubules. Although the mechanism of reabsorption is well
understood, Na⫹-glucose cotransport across the brush-border membrane
and facilitated diffusion across the basolateral membrane, questions
remain about the identity of the genes responsible for cotransport across
the brush border. Genetic studies suggest that two different genes
regulate Na⫹-glucose cotransport, and there is evidence from animal
studies to suggest that the major bulk of sugar is reabsorbed in the
convoluted proximal tubule by a low-affinity, high-capacity transporter
and that the remainder is absorbed in the straight proximal tubule by a
high-affinity, low-capacity transporter. There are at least three different
candidates for these human renal Na⫹-glucose cotransporters. This
review will focus on the structure-function relationships of these three
transporters, SGLT1, SGLT2, and SGLT3.
sodium-glucose cotransporters; genes; renal brush borders
in the regulation of
plasma glucose levels, and ever increasing attention is
now being given to renal glucose transporters as drug
targets in the treatment of patients with diabetes mellitus. Each day, ⬃180 g of D-glucose are filtered from
plasma by the kidneys, and this is all normally reabsorbed back into the blood in the proximal tubules. The
model for glucose transport across the tubule is similar
to that first proposed for the small intestine; i.e., glucose is first accumulated within the epithelium by a
Na⫹-glucose cotransporter (SGLT) in the brush-border
membrane and then is transported out of the cell
across the basolateral membrane by a facilitated sugar
transporter (see Ref. 72).
There is compelling genetic evidence that there are
two SGLTs in the human kidney (7, 8). Namely, patients with familial renal glycosuria (OMIM 233100;
online, Mendelian Inheritance in Man, http://www.
ncbi.nlm.nih.gov/OMIM) have no defect in intestinal
glucose absorption, whereas patients with familial intestinal glucose-galactose malabsorption (OMIM 182380)
only have a mild renal glycosuria. This suggests that
THE KIDNEYS PLAY A MAJOR ROLE
Address for reprint requests and other correspondence: E. Wright,
Dept. of Physiology, UCLA School of Medicine, Los Angeles, CA
90095-1751 (E-mail: [email protected]).
F10
the major intestinal Na⫹/glucose transporter only
plays a minor role in renal glucose reabsorption. Animal studies have supported this view. For example, in
rabbit isolated perfused renal proximal tubules, the
maximum rate of active glucose transport in the convoluted tubule (S1/S2) was an order of magnitude
greater than that in the late straight tubule. The
Michaelis-Menten coefficient value (Km) for glucose
was ⬃2 mM in S1/S2 and ⬃0.5 mM in S3 (1). This is
consistent with sugar uptakes reported for human renal brush-border membrane vesicles (68), where there
is some indication of high (Km ⬃0.5 mM)- and low (Km
6 mM)-affinity transporters.
In rabbit, the two SGLTs were spatially resolved by
Turner and Moran (66, 67), who isolated brush-border
membranes from the outer cortex (S1/S2) and outer
medulla (S3). At 17°C with a Na⫹ gradient of 40 mM,
the Km values were 6 mM in the outer cortex and 0.3
mM in the inner medulla. The high-affinity S3 transporter was more sensitive to phlorizin but more sensitive to competition by D-galactose than low-affinity
S1/S2 transporter. Finally, Turner and Moran conThe costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
0363-6127/01 $5.00 Copyright © 2001 the American Physiological Society
http://www.ajprenal.org
F11
INVITED REVIEW
cluded that the Na⫹/glucose stoichiometry, as judged
by indirect methods, was 1:1 for the S1/S2 and 2:1 for
the S3 transporters. Although the kinetic heterogeneity of Na⫹/glucose uptake by rabbit renal brush borders
has been examined by others, alterative interpretations of the data have been put forward (41).
The presently accepted dogma is that the bulk of the
filtered glucose is reabsorbed in the proximal convoluted tubule by a low-affinity, high-capacity SGLT, frequently referred to as “SGLT2,” and that the remainder
is reabsorbed by the high-affinity cotransporter SGLT1.
SGLTS
Table 1. Na⫹-glucose cotransporter genes
SGLT1
Locus name
Locus ID
Alternatives
Unigene
Accession nos. (refseq)
OMIM
Chromosome (refseg)
Size
Exons (no.)
Protein (residues)
SLC5A1
6523
D22S675
Hs 1964
NP00334
182380
22p13.1
72 kb
15
664
SGLT2
SLC5A2
6524
Hs 9003
NP003032
182381
16p11.2
672
SGLT3
SLC5A4
6527
DJ90G24.4/SAAT1
Hs 130101
AAB61732
22p12.1–12.3
37 kb
15
659
⫹
ID, identification; SGLT, Na -glucose cotransporter. Web address:
http://www.ncbi.nlm.nih.gov80/LocusLink.
Protein
Definitive identification of SGLTs came with our use
of fluorescence group-specific reagents on rabbit intestinal brush-border membranes (51–55). This was
based on our observation that isothiocyanates were
potent irreversible blockers of cotransport and that
this inhibition was prevented by the presence of Na⫹
and D-glucose. Nonspecific amino groups in brush borders were first labeled with phenyl isothiocyanate in
the presence of Na⫹ and D-glucose, and then the transporters were labeled with fluorescein isothiocyanate in
the absence of sugar. The SGLT was identified as a
75-kDa protein with a pI of 5.3, and additional studies
with the fluorescent-labeled transporter demonstrated
that Na⫹ opens up access to the sugar-binding site (see
also Fig. 2).
Genes
In 1985, I set out to clone the intestinal SGLT but
was initially unsuccessful because of our difficulty in
obtaining protein microsequence information for synthesis of DNA probes. Then, along with Mike Coady,
Tyson Ikeda, and Matthias Hediger, we developed a
novel cloning strategy, expression cloning, to successfully isolate a cDNA coding for the rabbit high-affinity
transporter (18, 19). When expressed in a variety of
heterologous expression systems, this clone exhibits all
the properties of the brush-border transporter, and
there is clear evidence that this protein (SGLT1) is
largely responsible for glucose and galactose transport
across the intestinal brush border. Expression cloning
soon became the preferred method for cloning new
transporters, channels, and receptors and has been
successfully exploited by my trainees (e.g., M. A. Hediger, A. M. Pajor, and B. A. Hagenbuch), colleagues,
and others to clone new transporters and channels in
the intestine, kidney, liver, and brain. This, and other
methods (homology cloning), has resulted in the identification of three human SGLTs (SGLT1, SGLT2, and
SGLT3) and over 55 other members of the SGLT1 gene
family in bacteria, yeast, invertebrates, and vertebrates (see 65). These family members include the
human myoinositol, iodide, and panthothenate and the
bacterial proline, urea, and galactose cotransporters.
Table 1 lists the features of the three identified
human SGLT genes. Human SGLT1 was first isolated
from an intestinal cDNA library by using the rabbit
SGLT1 as a probe (20) and was assigned to chromosome 22 q13.1 (17, 63). The gene spans 72 kb of
genomic DNA on chromosome 22 (6, 64). Northern blot
analysis revealed that the SGLT1 gene was transcribed in the human ileum and in rabbit small intestine, renal cortex, and outer renal medulla (42). SGLT2
was isolated from a human kidney library, mapped to
chromosome 16 p11.2, and this gene is transcribed in
human kidney cortex, and, to a much lesser degree, in
the ileum (70, 71). Rat kidney SGLT1 and SGLT2
genes are transcribed in the cortex and the outer medulla, respectively (74). SGLT3 was isolated from a pig
renal cell line by using our rabbit SGLT1 probe, and it
was designated SAAT1 on the basis of expression experiments in oocytes, where it produced a modest stimulation of amino acid transport uptake (29). We subsequently demonstrated that it was a low-affinity glucose
(not galactose)-Na⫹ cotransporter and referred to it
initially as pig SGLT2 (37, 38). The human isoform was
also identified on chromosome 22 (6). Interestingly,
SGLT3 is located downstream of SGLT1 (within 0.15
mb) and has a similar intron-exon organization, suggesting an ancient gene duplication. Pig SGLT3 is
transcribed in kidney, intestine, liver, and spleen (29).
There is little direct information about the translation of SGLT mRNAs in the human kidney, because of
the lack of suitable antibodies to discriminate among
the three proteins. Several polyclonal antipeptide antibodies have been raised against SGLT1 that do recognize proteins in brush borders of the rodent cortical
tubules and small intestine (3, 24, 25, 61). However,
these antibodies may also recognize other SGLTs because there is close similarity between the peptide
sequences used to raise the antibodies. Progress requires the development of specific antibodies to each of
the isoforms.
Orphan Clones
In addition to the SGLTs of known function, there is
at least one SGLT human gene whose function is unknown (6), and at least six orphan cDNAs have been
isolated from rabbit and human kidney libraries (42
and M. G. Martin and E. M. Wright, unpublished
F12
INVITED REVIEW
observations). The chromosome 22 SGLT-related gene
AP000362.1 (6) is reported to code for a 249-residue
polypeptide with similarity to SGLT2. Close inspection
suggests that exons 2, 3, and 5 code for peptides with
significant homologies to the corresponding SGLT 1–3
sequences. However, other exons appear to be missing.
Given that chromosome 22 comprises ⬍2% of the
genomic DNA and that the SGLT1 gene family is quite
large, it is to be expected that more SGLT1 genes will
emerge from the genome project. Only one SGLT1
family member gene, SMIT, was found on chromosome
21 (15). The Sanger Centre has discovered a new member of the SGLT family during the sequencing of chromosome 1 (gene dJ1024N4, http://webace.sanger.
ac.uk. accession CAC00574. GI: 9588428). Inspection
of the genomic sequence (E. Turk and E. M. Wright,
unpublished observations) shows that this SGLT1-related gene, at position 1p32.1-33, has 14 exons spanning 20 kb of genomic DNA. There is one less exon than
SGLT1 and SGLT3, due to the elimination of the
intron corresponding to the one between exons 4 and 5
in SGLT1. The gene codes for a protein of 674 residues
with high similarity to SGLT1. As yet, there is no
information available about either the tissue distribution of the transcript or the function of the protein. The
expression and function of SGLT-related genes in the
kidney and other tissues will continue to be an active
area of investigation.
SGLT Proteins
The SGLT1–3 genes code for proteins of 659–672
residues, with a predicted mass of 73 kDa. Amino acid
sequence alignments show that SGLT3 and SGLT2
have identities of 70 and 59% to SGLT1. The secondary
structure model for SGLT1 is shown in Fig. 1, and,
although there have been few or no experimental studies, SGLT2 or SGLT3 is predicted to have the same
secondary structure profile. The model contains 14
transmembrane a-helices (TMH) with both the hydrophobic NH2 terminus and the COOH terminus of TMH
14 facing the extracellular solution. All probably use
the consensus N-linked glycosylation site (N 248) between TMH 6 and 7. The model is based on 1) experimental studies on SGLT1 (see 65); 2) experimental
studies on related SGLT1 family members (Na⫹/iodide,
32; Na⫹/proline, 26, 69); and 3) a computational analysis of SGLT1, SGLT2, and pig SGLT3 using algorithms such as PredictProtein and Memsat (65).
On the basis of freeze-fracture electron microscopic
examination of recombinant membrane proteins expressed in Xenopus laevis oocytes, we have established
that SGLT1 functions as a monomer (11, 75). The
Fig. 1. The amino acid sequence and secondary structure of human (h) Na⫹-glucose cotransporter SGLT1. The
664-residue protein is shown as a pearl chain, with each residue indicated by a single-letter code. The secondary
structure model shows 14 transmembrane helices (TMH) with both the NH2 and COOH termini facing the
extracellular solution (65).
F13
INVITED REVIEW
density of intramembrane particles in the plasma
membrane of oocytes expressing SGLT1 is directly
proportional to SGLT1 transport activity, and the area
of the particle in the membrane is consistent with
TMH 14. The relationship between area and number of
TMHs was established by expressing membrane proteins of known structure in oocytes and comparing
their area with those obtained from two-dimensional
projection maps. We also concluded that the 14 TMHs
are arranged in an elliptical array with an x-to-y ratio
of 1.2. This quaternary structure was confirmed in a
recent study of a purified SGLT protein reconstituted
into liposomes (62). These results are in contrast to our
previous irradiation inactivation experiments on rabbit brush-border membranes, where the rate of transport decreased with increasing radiation dose with an
apparent size of a 290-kDa homotetramer (60). However, radiation decreased the size of SGLT1 protein
(detected on Western blots) with an apparent size of 70
kDa. The explanation for the higher estimate (290
kDa) by irradiation inactivation is not clear, but similar discrepancies have been reported for other transporters.
The only information available about the packing
arrangement of the 14 TMHs in these proteins comes
from cross-linking on the bacterial homolog vSGLT
(73). In these experiments, vSGLT was split into two
halves, the NH2- and COOH-terminal halves, and each
half was expressed in bacterial cells. The split proteins
were only functional when expressed in the same cell,
but the proteins migrated on SDS-PAGE gels at onehalf the expected mobility of the full-length transporter. However, when each half contained a single
cysteine residue (Cys149 and Cys423), the proteins migrated as the full-length protein after cross-linking
with bis-maleimidoethane or iodine. We conclude TMH
4 and 5 lie in close proximity to TMH 10 and 11 (⬍8 Å).
Further studies are in progress to obtain a TMH helical
packing map, as has already been done for lactose
permease (see 27). Additional studies (see STRUCTURE
AND FUNCTION) indicate that TMH 10–13 lie in close
proximity to form the sugar translocation pathway
through SGLT1.
TRANSPORT ACTIVITY OF SGLTS
Although the SGLT clones have been expressed in a
number of heterologous expression systems ranging
from insect to primate cells (e.g., 2, 59), the most
complete characterizations have been carried out in X.
laevis oocytes. A summary of the salient features is
presented in Table 2. There are no functional studies
on human SGLT3,1 so the features of SGLT3 refer to
work on the pig clone. At least with SGLT1 and
SGLT2, there is a fair convergence of kinetic parameters among different species (human, rabbit, rat; 23,
1
Preliminary functional studies of human SGLT3 in oocytes indicate that both the selectivity and kinetics are similar to those
reported for pig SGLT3 (C. Volk, A. Diez-Sampedro, B. A. Hirayama,
E. M. Wright, and H. Koepsell, unpublished observations).
Table 2. Transport properties of SGLTs
K0.5 (D-glucose; mM)
K0.5 (Na⫹; mM)†
Coupling (Na⫹/glucose)‡
Turnover no., s⫺1§
Phlorizin Ki, ␮M
Sugar selectivity
Na⫹ uniport
Water cotransport
SGLT1
SGLT2
SGLT3*
0.4
3
2
60
0.22
D-glc ⬃ D-gal
⫹
⫹
2
100
(1)
6
1.5
2
60
9
D-glc Ⰷ D-gal
⫹
⫹
1
D-glc Ⰷ D-gal
K0.5, affinity value; glc, glucose; gal, galactose. * Properties of pig
SGLT3. Those for SGLT1 and SGLT2 are for human. † Determined
at ⫺150 mV, but the voltage for SGLT2 was not specified. ‡ Determined using Na⫹ and sugar fluxes under voltage clamp, except for
SGLT2. § Taken from the maximum rate of Na⫹/glucose cotransport
at saturating voltages (⫺150 mV) and the number of transporters
estimated from SGLT charge movements (Qmax).
28, 74). The reader should be aware that SGLT2, unlike SGLT1 and 3, expresses very poorly in oocytes and
COS-7 cells; e.g., in SGLT2 oocytes, the Na⫹-dependent uptake of sugar and the Na⫹/sugar currents were,
at most, only 2–4 times greater than those in control
oocytes (28), whereas the sugar fluxes and currents in
SGLT1 and SGLT3 oocytes were normally 500–2,000
times greater than those in controls. This means that
the reported SGLT2 kinetics parameters are subject to
uncertainty.
All three SGLTs transport ␣-methyl-D-glucoside (␣MDG), a nonmetabolized, model substrate, in a sodium-dependent manner with apparent affinity (K0.5)
values from 0.4 (SGLT1) to 2 mM (SGLT2 and SGLT3).
For D-glucose the K0.5 values are 0.4, 2, and 6 mM,
respectively. Although SGLT1 does not discriminate
among ␣-MDG, glucose, and galactose, neither SGLT2
nor SGLT3 transports D-galactose efficiently (K0.5 ⬎ 20
mM). It appears the S1/S2 transporter in rabbit kidney
does not transport D-galactose efficiently either (see the
beginning of this review). Another model substrate for
SGLT1, 3-O-methyl-D-glucose, is also not transported
by either SGLT2 or 3. Phlorizin is a potent competitive
inhibitor of all the transporters, with inhibitor constants (Ki) of 0.2 mM for SGLT1 and 10 ␮M for SGLT3,
and an apparent Ki of 1 ␮M for SGLT2. This variation
in Ki between transporters is no greater than that
among the species for SGLT1, with Ki values ranging
from 0.01 ␮M in rat to 0.75 ␮M in rabbit (23).
The apparent sodium affinities for SGLT1 and 3 are
voltage dependent, but at hyperpolarizing membrane
potentials (⫺150 mV) the K0.5 asymptotes to ⬃3 mM
(23, 37, 38). The Na⫹ K0.5 for SGLT2 is thought to be
⬎100 mM (28, 74). The stoichiometry of Na⫹ to sugar
transport, when measured directly with radioactive
fluxes under voltage-clamp conditions, is two for both
SGLT1 and 3 (4, 39). Indirect methods were used in the
case of SGLT2, but, given the signal-to-noise problems
with both the rat and human clones, there is uncertainty in the reported 1:1 stoichiometry (28, 74). Other
cations such as H⫹ and Li⫹ can drive sugar transport
through SGLT1 and SGLT3 (22, 38), but this has not
been tested for SGLT2.
F14
INVITED REVIEW
In summary, the major differences in function between SGLT1 on one hand and SGLT2/SGLT3 on the
other relate to differences in the selectivity for sugar
transport. SGLT1 accepts the natural sugars D-glucose
and D-galactose with comparable affinities and a variety of synthetic sugars such as ␣-MDG and 3-O-methyl-D-glucose. SGLT2 and 3 are much more restrictive in
that they prefer D-glucose to D-galactose or 3-O-methylD-glucose. The affinity of SGLT1 for D-glucose is 10-fold
higher than for SGLT2 or 3. The Na⫹/glucose coupling
ratio is two for SGLT1 and SGLT3, and there is a hint
that it is one for SGLT2. All three are quite sensitive to
the specific inhibitor phlorizin.
TRANSPORT MECHANISMS
The kinetics of rabbit, rat, and human SGLT1 transport has been examined in considerable detail by measuring steady-state and pre-steady-state methods (16,
22, 23, 34, 35, 39, 43–50b). A limited series of experiments has also been conducted on pig SGLT3 (37, 38),
but there is little information available about SGLT2,
owing to its low level of expression (see TRANSPORT
ACTIVITY OF SGLTS). However, the similarity in kinetics of
SGLT1, SGLT3, and other members of the SGLT1
family of proteins [SMIT (14) and NIS (9)] points to a
common transport mechanism.
A six-state ordered kinetic model with “mirror” symmetry is shown in Fig. 2A. It is assumed that the carrier
is divalent and that two external Na ions bind in a single
reaction step before sugar (or phlorizin) to neutralize this
charge. There are two transport loops linking the six
discrete states of the carrier: one carries out the transport
of Na ions in the absence of sugar, and the other carries
out the coupled transport of two Na ions and one sugar
molecule. Of the three transmembrane state transitions,
there is only one charge translocation step (1–6); thus
SGLT1 currents are carried by state transitions of the
unloaded carrier. Rapid equilibrium was not assumed,
and a formal description of the reaction scheme has been
obtained. We have carried out extensive “iterative” numerical simulations of both our pre-steady-state and
steady-state data and have arrived at a set of rate constants that account for the global properties of the transporter (see 16, 33, 50b).
Substantial experimental evidence has been gathered in favor of this model. For example, 1) the coupling coefficient for Na⫹ and glucose transport is 2:1; 2)
in the absence of sugar, Na⫹ is transported with kinetics similar to those of Na⫹-glucose cotransport; i.e., two
Na ions are transported per cycle; 3) the maximum rate
of Na⫹-glucose cotransport at saturating membrane
potentials is independent of sodium concentration; and
4) rapid voltage-jump experiments in the absence of
sugar show SGLT1 pre-steady-state currents that represent the redistribution of the states of the transporter between C2 and C6. The cotransport turnover
time is ⬃60 s at 22°C, and, for the Na⫹ uniport, is ⬃10
s. The slow steps in the cycle appear to be the internal
dissociation of Na⫹ from the carrier (C5-C6) and the
conversion of C6 to C1, but which of these is the
Fig. 2. A kinetic model for Na⫹-glucose cotransport. A: the 6-state
model with ordered ligand binding (44). At negative membrane
potentials, the divalent membrane protein is in the C1 state, and in
the presence of high extracellular 2 Na ions bind to form the C2 state.
In the absence of sugar, the protein facilitates inward Na⫹ transport
(Na⫹ leak), but, in the presence of outside sugar, 2 Na ions and 1
sugar are transported across the membrane. Sugar and then Na⫹
dissociate at the inner surface, and the unloaded carrier completes
the cycle, C6–C1. Na⫹ is shown to bind to C1 at the bottom of a well
30% of the way across the membrane field. B: proposed structural
basis for Na⫹-dependent conformational changes in SGLT1. TMH
10–13 are postulated to form the sugar translocation path through
SGLT1. Cysteines at positions 457, 468, and 499 are not accessible to
methanethiosulfonate (MTS) reagents in state C1, but, on Na⫹
binding, these become accessible, possibly by tilting of TMH 13. This
crevice between TMH 10–13 may also be a sugar-binding site, as
access by MTS reagents is blocked by the presence of phlorizin (C3).
rate-limiting step depends critically on the membrane
potential. Recently, two independent studies using the
giant excised-patch technique (10, 58) have provided
additional support by showing SGLT1 can work in the
reverse mode (transporting sugar from cytoplasm to
the exterior). As predicted by our model, the affinities
for sugar and Na⫹ on the cytoplasmic face of the
membrane are orders of magnitude lower than those
on the outward face of the membrane.
Berteloot and colleagues (12) have critically examined our six-state model and have pointed out that we
have oversimplified the reaction scheme by assuming
Na⫹ binding can be described as a single reaction
(either the binding of the first Na⫹ is much faster than
the second or the two Na⫹ bind with identical rate
constants to similar sites; 44). According to Berteloot et
al. (12), this may only be valid over a restricted range
INVITED REVIEW
of Na⫹ concentrations or when there is very strong
positive cooperativity between the Na⫹-binding sites.
Evidence that there is strong interaction between the
two Na⫹-binding sites comes from the near identity of
the Hill coefficient and the coupling coefficient (39). On
balance, the gain in simplifying the model from 20 to
14 rate constants outweighs the more rigorous approach.
SGLT1 and other cotransporters also behave as
channels and pumps for urea and water (31, 35, 36, 40),
but this is beyond the scope of the present review.
STRUCTURE AND FUNCTION
The present hypothesis is that Na⫹ and sugar have
two separate pathways through SGLT1: Na⫹ permeates through the NH2-terminal half of the protein, and
sugar permeates through the COOH-terminal half
(Fig. 3). Na⫹ binding to the NH2-terminal domain then
causes a long-range conformational change in the protein to permit sugar binding and translocation. The
experimental basis for this hypothesis comes from
studies of SGLT1/SGLT3 chimeras that suggested that
the COOH-terminal domain (residues 407–662 containing TMHs 10–14) determined sugar affinity and
selectivity (47). This was reinforced by experiments
showing that the truncated protein containing residues
407–662 (C5) behaved as a sugar uniporter either
when expressed in oocytes or when reconstituted into
proteoliposomes (45, 46). We speculate that TMHs 10–
13 form the sugar translocation pathway through
SGLT1 (Fig. 2B), as TMH 14 is missing from some
members of the SGLT1 family and deletion of this
TMH does not eliminate sugar transport. Although we
have not yet been able to obtain full Na⫹-glucose cotransport activity by coexpressing separate N9 and C5
cRNAs in oocytes, we have been successful in expressing a split Vibrio transporter (vSGLT) in bacteria (73).
With another member of the gene family, the bacterial
Na⫹-proline cotransporter (PutP), there is further ev-
F15
idence that the NH2-terminal region of the protein is
involved in Na⫹ binding and coupling (56, 69). This is
reasonable in view of the fact that the amino acid
sequences of the NH2-terminal halves of SGLT1 family
members are quite similar, whereas the sequences of
the COOH-terminal half are quite divergent.
Additional evidence for both the alternating access
model for Na⫹-glucose cotransport and the location of
the sugar transport pathway comes from studies on the
effect of thiol reagents on cysteine mutants (B. A.
Hirayama, D. D. F. Loo, E. M. Gallardo, E. Turk, and
E. M. Wright, unpublished observations; 31, 34). The
experimental approach is to replace residues on TMH
10–13 with cystines and then determine the effect of
methanethiosulfonate (MTS) reagents on Na⫹/glucose
transport. So far, we have found that cysteines at
positions 457, 468, and 499 render the transporter
sensitive to MTS reagents; e.g., 457C SGLT is reversibly blocked by exposure to 1 ␮M MTSEA for 80 s. In
contrast, cysteines at 530, 535, 497, 544, and 665 are
insensitive. The wild-type transporter is also insensitive to MTS reagents, and cysteineless vSGLT is fully
functional (73).
A common feature of SGLT1 with cysteines at 457,
468, and 499 is that they are only sensitive to MTS
reagents when the transporter is in the C2 conformation (Fig. 2A). MTS accessibility to these residues is
blocked when the reaction is carried out 1) in the
absence of Na⫹, 2) in the presence of Na⫹ and either
glucose or phlorizin, and 3) in the presence of Na⫹
when the membrane potential is depolarized. The access of MTS reagents to the residues is directly proportional to the probability of the protein being in the C2
state. One interpretation of these observations is that
residues 457, 468, and 499 line the sugar translocation
pathway through the transporters and that this pathway is only accessible in the C2 conformation (Fig. 2B).
Given the dimensions of the MTS reagents that can
gain access, phlorizin that can block access, and the
Fig. 3. Secondary structure model for SGLT1
showing the location of cysteine mutations. The
sugar-binding/translocation domain, residues 407–
664, is referred to as C5. TMH 5 and 6 are indicated
to be within 8 Å of TMH 10–11 (73).
F16
INVITED REVIEW
sugar analogs that can be transported, we suggest that
sodium produces a large change in the tilt or rotation of
one or more TMHs in the 10–13 helical bundle (Fig.
2B). Direct evidence for motion of 457C during the
C1-C2 transitions comes from monitoring fluorescence
of rhodamine-labeled protein during voltage-jump experiments (35). Both the level and time course of the
change in fluorescence closely followed the charge
movements that are a hallmark of the voltage-induced
changes in state of the protein. We further suggest that
sugar translocation to the cell interior requires further
large changes in helical tilt and/or rotation and that it
is these changes in conformation that account for the
cotransport of water and urea (2 Na⫹:1 glucose:250
water; 31, 36, 40). Presently, we are extending these
cysteine-scanning experiments to map the conformational changes for the cotransporter during other state
transitions and to determine the TMH packing.
We postulate that the structural basis for cotransport is similar among the SGLTs (SGLT1–3) and
among the other members of the large SGLT superfamily.
IDENTITY OF RENAL SGLTS
A question still to be resolved is the identity of the
genes responsible for glucose reabsorption across the
brush border of the human proximal tubule. Although
it is clear from genetic studies that SGLT1 plays only a
minor role, perhaps in the late tubule, the identity of
the low-affinity renal SGLT (“SGLT2”) is still unclear.
There are at least two low-affinity SGLTs, SGLT2 and
SGLT3, and there are likely to be others. Another
SGLT is indicated by the results of the chromosome 22
sequencing project (6), and our screening of human
renal cDNA libraries has identified four other clones
(M. G. Martin and E. M. Wright, unpublished observations). A similar problem exists in rodent models, and
the information that does exist about the expression of
SGLTs, e.g., SGLT1 and SGLT2 in the rat kidney (74),
only relates to mRNA levels. In the absence of information about SGLT protein in renal brush borders, it is
difficult to draw definitive conclusions, especially when
there are examples where there is a divergence between SGLT mRNA and protein levels (30). One powerful approach with animal models is to use geneknockout technology to explore the physiological role of
SGLTs, similar to that employed already with the
basolateral facilitated transporter GLUT-2 (13).
GLUT-2-null mice show a severe renal glycosuria.
However, extrapolation of results from animal models
to humans may not always be justified. One useful
strategy with humans is to determine which SGLT
gene is responsible for familial renal glycosuria, but
this may not be foolproof as it has already been reported that mutations in GLUT2 may cause renal
glycosuria in patients with the Fanconi-Bickel syndrome (57). Finally, it should be noted that in a linkage
study of 25 patients in 5 unrelated families with renal
glycosuria (3a), the gene for the renal glycosuria
(GLYS1, OMIM 233100) was segregated with the Hu-
man Leukocyte Antigen complex. This suggests that
GLYS1 is on chromosome 6p21.3, but so far the gene
has not been isolated. However, I do anticipate that the
gene coding for the major renal bush-border Na⫹-glucose cotransporter will soon be positively identified,
and this will open a new era in the study of the renal
handling of sugars.
I thank Debra Moorehead, Matthias Quick, and Bruce Hirayama
for assistance with the figures.
The studies reported from this laboratory were made possible by
the work of many talented colleagues, postdoctoral fellows, students,
and research assistants, and by the financial support of National
Institute of Diabetes and Digestive and Kidney Diseases Grants
DK-19567, DK-44602, and DK-44582.
REFERENCES
1. Barfus DW and Schafer JA. Differences in active and passive
glucose transport along the proximal nephron. Am J Physiol
Renal Fluid Electrolyte Physiol 240: F322–F332, 1981.
2. Birnir B, Lee HS, Hediger MA, and Wright EM. Expression
and characterization of the intestinal Na⫹/glucose cotransporter
in COS-7 cells. Biochim Biophys Acta 1048: 100–104, 1990.
3. Cramer SC, Pardridge WM, Hirayama BA, and Wright EM.
Colocalization of GLUT-2 glucose transporter, sodium/glucose
cotransporter, and ␥ glutamyl transpeptidase in rat kidney using
double peroxidase immunocytochemistry. Diabetes 41: 766–770,
1992.
3a.De Marchi S, Cecchin E, Basile A, Proto G, Donadon W,
Jengo, Schinella D, Jus A, Villalta D, de Paoli, P, Santini
G, and Tesio F. Close genetic linkage between HLA and renal
glycosuria. Am J Nephrol 4: 280–286, 1984.
4. Diez-Sampedro A, Eskandari S, Wright EM, and Hirayama
BA. Na⫹ to sugar stoichiometry of SGLT3. Am J Physiol Renal
Physiol. In press.
5. Diez-Sampedro A, Lostao MP, Wright EM, and Hirayama
BA. Glycoside binding and translocation in Na⫹-dependent glucose cotransporters: comparison of SGLT1 and SGLT3. J Membr
Biol 176: 111–117, 2000.
6. Dunham I, Shimizu N, Roe BA, and Chissoe S. The DNA
sequence of human chromosome 22. Nature 402: 489–493. 1999.
7. Elsas LJ, Hillman RE, Patterson JH, and Rosenberg LE.
Renal and intestinal hexose transport in familial glucose-galactose malabsorption. J Clin Invest 49: 576–585, 1970.
8. Elsas LJ and Rosenberg LE. Familial renal glycosuria: a
genetic reappraisal of hexose transport by kidney and intestine.
J Clin Invest 48: 1845–1854, 1969.
9. Eskandari S, Loo DDF, Dai G, Levy R, Wright EM, and
Carrasco N. Thyroid Na⫹/I symporter: mechanisms, stoichiometry, and specificity. J Biol Chem 272: 27230–27238, 1997.
10. Eskandari S, Loo DDF, and Wright EM. Functional asymmetry of the sodium/glucose cotransporter. Experimental Biology 99, Washington, DC. FASEB J 13: A399, 1999.
11. Eskandari S, Wright EM, Kreman M, Starace DM, and
Zampighi GA. Structural analysis of cloned membrane proteins
by freeze-fracture electron microscopy. Proc Natl Acad Sci USA
95: 11235–11240, 1998.
12. Falk S, Guay A, Chenu C, Patil S, and Berteloot A. Reduction of an eight-state mechanism of cotransport to a six-state
model using a new computer program. Biophys J 74: 816–830.
1998.
13. Guillam MT, Hummler E, Schaerer Y, Ji E, Birnbaum JJ,
Beermann F, Schmidt A, Deriaz N, and Torens B. Early
diabetes and abnormal postnatal pancreatic islet development in
mice lacking GLUT-2. Nature Genet 17: 327–330, 1997.
14. Hager K, Hazama A, Kwon HM, Loo DDF, Handler JS, and
Wright EM. Kinetics and specificity of the renal Na⫹/myoinositol cotransporter expressed in Xenopus oocytes. J Membr
Biol 143: 103–113, 1995.
15. Hattori M, Fujiyama A, Taylor TD, Watanabe H, Yada T,
Park HS, Toyoda A, Ishii K, Totoki Y, Choi DK, Soeda E,
Ohki M, Takagi T, Sakaki Y, Taudien S, Blechschmidt K,
INVITED REVIEW
16.
17.
18.
19.
20.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
Polley A, Menzel U, Delabar J, Kumpf K, Lehmann R,
Patterson D, Reichwald K, Rump A, Schillhabel M, and
Schudy A. The DNA sequence of human chromosome 21. The
chromosome 21 mapping and sequencing consortium. Nature
405: 311–319, 2000.
Hazama A, Loo DDF, and Wright EM. Presteady-state currents of the Na⫹/glucose cotransporter (SGLT1). J Membr Biol
155: 175–186, 1997.
Hediger MA, Budarf ML, Emanuel BS, Mohandas TK, and
Wright EM. Assignment of the human intestinal Na⫹/glucose
cotransporter gene (SGLT1) to the q11.2 3 qter region of chromosome 22. Genomics 4: 297–300. 1989.
Hediger MA, Coady MJ, Ikeda TS, and Wright EM. Expression cloning and cDNA sequencing of the Na⫹/glucose cotransporter. Nature 330: 379–381, 1987.
Hediger MA, Ikeda T, Coady M, Gundersen CB, and
Wright EM. Expression of size selected mRNA encoding the
intestinal Na/glucose cotransporter in Xenopus laevis oocytes.
Proc Natl Acad Sci USA 84: 2634–2637, 1987.
Hediger MA, Turk E, and Wright EM. Homology of the
human intestinal Na⫹/glucose and E. coli Na⫹/proline cotransporters. Proc Natl Acad Sci USA 86: 5748–5752, 1989.
Hirayama BA, Loo DDF, and Wright EM. Cation effects on
protein conformation and transport in the Na⫹/glucose cotransporter. J Biol Chem 272: 2110–2115, 1997.
Hirayama BA, Lostao MP, Panayotova-Heiermann M, Loo
DDF, Turk E, and Wright EM. Kinetic and specificity differences between rat, human and rabbit Na/glucose cotransporters
(SGLT1). Am J Physiol Gastrointest Liver Physiol 270: G919–
G926, 1996.
Hirayama BA, Wong HC, Smith CD, Hagenbuch BA, Hediger MA, and Wright EM. Intestinal and renal Na⫹/glucose
cotransporters share common structures. Am J Physiol Cell
Physiol 261: C296–C304, 1991.
Hwang ES, Hirayama BA, and Wright EM. Distribution of
the SGLT1 Na⫹/glucose cotransporter and mRNA along the
crypt-villus axis of rabbit small intestine. Biochem Biophys Research Commun 181: 1208–1217, 1991.
Jung H, Rubenhagen R, Tebbe S, Leifker K, Tholema N,
Quick M, and Schmid R. Topology of the Na⫹/proline transporter of Escherichia coli. J Biol Chem 273: 26400–26407, 1998.
Kaback HR. Structure/function studies on the lactose permease
of Escherichia coli. Acta Physiol Scand Suppl 643: 21–33, 1998.
Kanai Y, Lee WS, You G, Brown D, and Hediger MA. The
human kidney low affinity Na⫹/glucose cotransporter SGLT2.
J Clin Invest 93: 397–404, 1994.
Kong CT, Yet SF, and Lever JE. Cloning and expression of a
mammalian Na⫹/amino acid cotransporter with sequence similarity to Na⫹/glucose cotransporters. J Biol Chem 268: 1509–
1512, 1993.
Lescale-Matys L, Dyer J, Scott D, Wright EM, and ShiraziBeechey SP. Regulation of ovine intestinal Na⫹/glucose cotransporter (SGLT1) by sugar is dissociated from mRNA abundance. Biochem J 291: 435–440, 1993.
Leung DW, Loo DDF, Hirayama BA, Zeuthen T, and
Wright EM. Urea transport by cotransporters. J Physiol (Lond)
528.2: 251–257, 2000.
Levy O, DelaVieja A, Ginter CS, Riedel C, Dai G, and
Carrasco N. N-linked glycosylation of the thyroid Na⫹/I-symporter (NIS). J Biol Chem 273: 22657–22663, 1998.
Loo DDF, Hazama A, Supplisson S, Turk E, and Wright
EM. Relaxation kinetics of the Na⫹/glucose cotransporter. Proc
Natl Acad Sci USA 90: 5767–5771, 1993.
Loo DDF, Hirayama BA, Gallardo EM, Lam JT, Turk E,
and Wright EM. Conformational changes couple Na⫹ and glucose transport. Proc Natl Acad Sci USA 95: 7789–7794, 1998.
Loo DDF, Hirayama BA, Meinild AK, Chandy G, Zeuthen
Z, and Wright EM. Passive water and ion transport by cotransporters. J Physiol (Lond) 518: 195–202, 1999.
Loo DDF, Zeuthen T, Chandy G, and Wright EM. Cotransport of Water by the Na⫹/glucose cotransporter. Proc Natl Acad
Sci USA 93: 13367–13370, 1996.
Mackenzie B, Panayotova-Heiermann M, Loo DDF, Lever
JE, and Wright EM. SAAT1 is a low affinity Na⫹/glucose
F17
cotransporater and not an amino acid transporter. J Biol Chem
269: 22488–22491, 1994.
38. Mackenzie BM, Loo DDF, Panayotova-Heiermann M, and
Wright EM. Biophysical characteristics of the pig kidney Na⫹/
glucose cotransporter SGLT2 reveal a common mechanism for
SGLT1 and SGLT2. J Biol Chem 271: 32678–32683, 1996.
39. Mackenzie B, Loo DDF, and Wright EM. Relations between
Na⫹/glucose cotransporter (SGLT1) currents and fluxes. J
Membr Biol 162: 101–106, 1998.
40. Meinild AK, Klaerke D, Loo DDF, Wright EM, and Zeuthen
T. The human Na⫹/glucose cotransporter is a molecular water
pump. J Physiol (Lond) 508: 15–21, 1998.
41. Oulianova N and Berteloot A. Sugar transport heterogeneity
in the kidney: two independent transporters or different transport modes through an oligomeric protein. 1. Glucose transport
studies. J Membr Biol 153: 181–194, 1996.
42. Pajor AM and Wright EM. Sequence, tissue distribution and
functional expression of a mammalian Na⫹/nucleoside cotransporter. J Biol Chem 267: 3557–3560, 1992.
45. Panayotova-Heiermann M, Eskandari S, Zampighi GA,
and Wright EM. Five transmembrane helices form the sugar
pathway through the Na⫹/glucose transporter. J Biol Chem 272:
20324–20327, 1997.
46. Panayotova-Heiermann M, Leung DW, Hirayama BA, and
Wright EM. Purification and functional reconstitution of a truncated human Na⫹/glucose cotransporter (SGLT1) expressed in
E. coli. FEBS Lett 459: 386–290, 1999.
47. Panayotova-Heiermann M, Loo DDF, Klong CT, Lever JE,
and Wright EM. Sugar binding to Na⫹/glucose cotransporters
is determined by the C-terminal half of the protein. J Biol Chem
271: 10029–10034, 1996.
48. Panayotova-Heiermann M, Loo DDF, Lostao PM, and
Wright EM. Sodium-D-glucose cotransporter charge movements
involve polar residues. J Biol Chem 269: 21016–21020, 1994.
49. Panayotova-Heiermann M, Loo DDF, and Wright EM. Kinetics of steady-state currents and charge movements associated
with the rat Na⫹/glucose cotransporter. J Biol Chem 270:
27099–27105, 1995.
50. Panayotova-Heiermann M, Loo DDF, and Wright EM. Neutralization of conservative charged transmembrane residues in
the Na⫹/glucose cotransporter SGLT1. Biochemistry 37: 10522–
10528, 1998.
50a.Parent L, Supplisson S, Loo DDF, and Wright EM. Electrogenic properties of the cloned Na⫹/glucose cotransporter. Part I.
Voltage-clamp studies. J Membr Biol 125: 49–62, 1992.
50b.Parent L, Supplisson S, Loo DDF, and Wright EM. Electrogenic properties of the cloned Na⫹/glucose cotransporter. Part II.
A transport model under nonrapid equilibrium conditions. J
Membr Biol 125: 63–79, 1992.
51. Peerce BE and Wright EM. Conformational changes in the
intestinal brush border Na-glucose cotransporter labeled with
fluorescein isothiocyanate. Proc Natl Acad Sci USA 81: 2223–
2226, 1984.
52. Peerce BE and Wright EM. Sodium induced conformational
changes in the glucose transporter of intestinal brush borders.
J Biol Chem 259: 14105–14112, 1984.
53. Peerce BE and Wright EM. Evidence for tyrosyl residues at
the Na⫹ site on the intestinal Na⫹/glucose cotransporter. J Biol
Chem 260: 6026–6031, 1985.
54. Peerce BE and Wright EM. Distance between substrate sites
on the Na/glucose cotransporter by fluorescence energy transfer.
Proc Natl Acad Sci USA 83: 8092–8096, 1986.
55. Peerce BE and Wright EM. Examination of the Na⫹-induced
conformational change of the intestinal brush border Na/glucose
symporter using fluorescent probes. Biochemistry 26: 4272–4279,
1987.
56. Quick M, Stolting S, and Jung H. Role of conserved Arg40 and
Arg117 in the Na⫹/proline transporter of Escherichia coli. Biochemistry 38: 13523–13529, 1999.
57. Santer R, Schneppenheim A, Dombrowski H, Götze B,
Steinmann B, and Schaub J. Mutations in GLUT2, the gene
for the liver-type glucose transporter, in patients with FanconiBickel syndrome. Nature Genet 17: 324–326, 1997.
F18
INVITED REVIEW
58. Sauer GA, Nagel G, Hoepsell H, Bamberg E, and Hartung
K. Voltage and substrate dependence of the inverse transport
mode of the rabbit Na⫹/glucose cotransporter (SGLT1). FEBS
Lett 469: 98–100, 2000.
59. Smith CD, Hirayama BA, and Wright EM. Baculovirus mediated expression of the Na⫹/glucose cotransporter in Sf9 cells.
Biochim Biophys Acta 1104: 151–159, 1992.
60. Stevens BR, Fernandez A, Hirayama B, Wright EM, and
Kempner ES. Intestinal brush border membrane Na⫹/glucose
cotransporter functions in situ as a homotetramer. Proc Natl
Acad Sci USA 87: 1456–1460, 1990.
61. Takata K, Kasahara T, Kasahara M, Ezaki O, and Hirano
H. Localization of Na⫹-dependent active types and erythrocyte/
HepG2-type glucose transporters in rat kidney: immunofluorescence and immunogold study. J Histochem Cytochem 39: 287–
298, 1991.
62. Turk E, Kim O, le Coutre J, Whitelegge JP, Eskandari S,
Lam JT, Kreman M, Zampighi G, Faull KF, and Wright
EM. Molecular characterization of Vibrio parahaemolyticus vSGLT: a model for sodium-coupled sugar cotransporters. J Biol
Chem 275: 25711–25716, 2000.
63. Turk E, Klisak I, Sparkes RS, and Wright EM. Assignment
of the human Na⫹/glucose cotransporter gene SGLT1 to chromosome 22q13.1. Genomics 17: 752–754, 1993.
64. Turk E, Martı́n MG, and Wright EM. Structure of the human
Na⫹/glucose cotransporter gene SGLTI. J Biol Chem 269:
15204–15209, 1994.
65. Turk E and Wright EM. Membrane topological motifs in the
SGLT cotransporter family. J Membr Biol 159: 1–20, 1997.
66. Turner RJ and Moran A. Heterogeneity of sodium-dependent
D-glucose transport sites along the proximal tubule: evidence
form vesicle studies. Am J Physiol Renal Fluid Electrolyte
Physiol 242: F406–F411, 1982.
67. Turner RJ and Moran A. Further studies of proximal tubular
brush border membrane D-glucose transport heterogeneity. J
Membr Biol 70: 37–45, 1982.
68. Turner RJ and Silverman M. Sugar uptake into brush border
vesicles from normal human kidney. Proc Natl Acad Sci USA 74:
2825–2829, 1997.
69. Wegener C, Tebbe S, Steinhoff HJ, and Jung H. Spin labeling analysis of structure and dynamics of the Na⫹/proline transporter of Escherichia coli. Biochemistry 39: 4831–4837, 2000.
70. Wells RG, Kanai Y, Pajor AM, Turk E, Wright EM, and
Hediger MA. The cloning of a human kidney cDNA with similarity to the sodium/glucose cotransporter. Am J Physiol Renal
Fluid Electrolyte Physiol 263: F459–F465, 1992.
71. Wells RG, Mohandas TK, and Hediger MA. Localization of
the Na⫹/glucose cotransporter gene SGLT2 to human chromosome 16 close to the centromere. Genomics 17: 787–789, 1993.
72. Wright EM. Glucose galactose malabsorption. Am J Physiol
Gastrointest Liver Physiol 275: G879–G882, 1998.
73. Xie Z, Turk E, and Wright EM. Characterization of the Vibrio
parahaemolyticus Na⫹/glucose cotransporter: a bacterial member of the sodiuym/glucose transport (SGLT) family. J Biol Chem
275: 25959–25964, 2000.
74. You GF, Leer WS, Barros EJG, Kanai Y, Huo TL, Khawaja
S, Wells RG, Nigam SK, and Hediger MA. Molecular characteristics of Na⫹-coupled glucose transporters in adult and embryonic rat kidney. J Biol Chem 270: 29365–29371, 1995.
75. Zampighi GA, Kreman M, Boorer KJ, Loo DDF, Bezanilla
F, Chando G, Hall JE, and Wright EM. A method for determining the unitary functional capacity of cloned channels and
transporters expressed in Xenopus laevis oocytes. J Membr Biol
148: 65–78, 1995.