Mechanisms and Regulation of Sugar Transport
Dietary regulation of the intestinal sodium-dependent glucose cotransporter
(SGLT I )
Soraya P. Shirazi-Beechey*$, Susan M. Gribble*, I: Stuart Wood*, Patrick S. Tarpey*, R. Brian Beechey*, Jane
Dyer*, Dennis Scottt and Patrick J. Barker$
*Epithelial Function and Development Group, Institute of Biological Sciences, University of Wales, Aberystwyth,
Dyfed, SY23 3DD, U.K., +Division of Biochemistry, The Rowett Research Institute, Bucksburn, Aberdeen, AB2 9SB,
U.K. and $Microchemical Facility, The Babraham Institute, Babraham, Cambridge, CB2 4AT, U.K.
Introduction
The intestinal brush-border Na’-glucose
cotransporter (SG1,Tl) functions in the uptake of dietary
I )-glucose and i)-galactose across the microvillus
membrane. Impaired absorption of these sugars
occurs in glucose-galactose malabsorption, an autosoma1 recessive disease caused by a defect in
SGI,Tl [ 11. Studies on the kinetics of glucose transport across the intestinal brush border using intact
tissue, isolated cells and brush-border membrane
vesicles have established that there is one major
transporter that handles all hexoses with an
equatorial hydroxyl group on C-2. These include
the natural dietary sugars i)-glucose and i)-galactose
and the non-metabolized sugars 3-0-methyl a-1)glucopyranoside and methyl a-1)-glucopyranoside
[ 2.31.
It has been shown that SGLT1 is a glycosylated integral brush-border membrane protein with
an apparent molecular mass of 7.5 kDa. cDNAs
encoding rabbit, human, rat and lamb SGLT1 have
been cloned and sequenced. The amino acid
sequences have been deduced and exhibit significant sequence similarity [ 4,q. Antibodies have been
raised to peptides which correspond to extramembranous regions of the rabbit SGLT1
sequence [6,7]. These antibodies have been used to
probe brush-border membrane vesicles isolated
from intestinal tissues from a large number of
species. mKNAs isolated from these tissues have
been probed with rabbit intestinal SGL7’1 cDNA
[H,9]. The combined results indicate that the
S<;I,Tl DNA and protein sequence are highly conserved throughout evolution.
The activity and abundance of SGI.Tl have
been shown to be regulated by dietary carbohydrates in non-ruminant [ 10,111 and ruminant
animals [6,121. Limited information on the molecular mechanisms, site and signals involved in this
regulation is available.
The intestinal tract of ruminant animals, such
as sheep, provides an excellent model system for
studying the dietary regulation of sugar absorption
-
§To whom correspondence should be addressed.
[6,12- 151. In pre-ruminant lambs, 1 )-glucose and I )galactose, derived from milk sugar, lactose, are
absorbed by SGLT1 [ 12,16,17]. Lambs normally
wean at 3-8 weeks, and as the diet changes from
milk to grass the rumen develops; the dietary carbohydrates are fermented to volatile fatty acids in the
rumen, and no hexoses reach the small intestine
[ 181. We have shown that there is a consequent loss
of SGLT1 function and SGLT1 protein from the
apical plasma membrane of the enterocyte and a
decline in the levels of SGI,Tl mRNA [6,12,13].
Maintaining lambs on a milk-replacer diet
beyond the normal weaning period prevented
rumen development and the loss of both the
SGLT1 transport protein and transport function.
Cannulae were placed into the duodenal lumen of
ruminant sheep which were maintained on a
normal roughage diet. The infusion of solutions of
either i)-glucose or methyl a-1)-glucopyranoside
into the intestine through these cannulae resulted in
the expression of functional SGLTI protein in the
brush-border membrane [6,1.5].
The dramatic changes in the activity and the
abundance of SGLTI protein observed in the
brush-border membrane, during development and
with intestinal infusions, were not directly co-ordinated with fluctuations in the levels of SGI,Tl
mRNA [6,13].
The accumulated information implies that: (a)
the presence of glucose in the lumen of the intestine
can lead to stimulation of synthesis of functional
SGLT1, (b) the cellular metabolism of hexoses is
not a pre-requisite for the induction, and (c) the cellular level of regulation is post-transcriptional.
In this paper we report our findings on: the
mechanism by which the lumenal sugar may
stimulate the synthesis of functional SGI,Tl, the
probable cellular site receiving the signal and the
likely mechanism involved in the intracellular synthesis and processing of SGLTl.
Molecular characteristics of inducing
sugars
Various sugar analogues were tested for their abilities to induce functional SGLTl in the intestinal
655
Biochemical Society Transactions
656
brush-border membrane of ruminant sheep. Solutions (30 mM) of u-glucose, 1)-galactose,methyl ai)-glucopyranoside,
3- 0-met hy I
a -1 )-glucopyranoside, u-fructose, 2-deoxy-u-glucose, u-sorbito1 and i)-mannitol were infused through duodenal
cannulae [6,12,15], either for 4 consecutive days or
for 2 h (see below). Brush-border membrane
vesicles were isolated and the SGLTl activity was
assayed [ 12,19,20]. The abundance of the induced
SGLTI-protein was determined using Western blot
analysis; the antibody to a nonadecapeptide segment of SGLT1 was used. Infusion of the intestine
with either i)-glucose or i)-galactose led to the
induction of functional SGLT1, while infusions with
i)-mannitol and i)-sorbitol, (non-transported, nonmetabolizable alditols) did not. The induction of
SGIJTl protein in the brush-border membrane by
the infusion of methyl a-1)-glucopyranoside or 3- 0methyl a-1)-glucopyranoside, non-metabolizable
substrates of SGLT1, indicates that there is no prerequisite for the substrate to be metabolized by the
enterocytes. Induction of functional SGLTl by 1)fructose and 2-deoxy-t)-glucose implies that the
inducing sugar need not necessarily be the substrate of SGLT1.
From these results one can speculate the
presence of a sensing system with a different sugar
recognition specificity to that of SGLT 1. Since
2-deoxy-i)-glucose is not transported by any known
intestinal brush-border membrane protein (S. P.
Shirazi-Heechey and I. S. Wood, unpublished
work) [ 21, the induction by 2-deoxy-u-glucose
implies that the sensing system is localized on the
external face of the luminal membrane.
Infusing the intestinal lumen of the ruminant
sheep, through ileal cannulae, with v-glucose for 4
days induced the expression of SGLT1 protein in
the ileal brush-border membrane. It did not, however, lead to induction of SG1,Tl in the duodenal
brush-border membrane. Furthermore, intravenous
infusion of i)-glucose in ruminant sheep, which led
to an increase in the abundance of basolateral membrane glucose transporter, GLUT2, did not lead to
the induction of SGLTl 0. Dyer and S. P. ShiraziBeechey, unpublished work). The data indicate that
systemic factors may not play a major role as
signals regulating the activity and the expression of
apical SGLT 1.
Epithelial location of the sugar
receptor
New enterocytes are produced in the intestinal
crypts, they mature as they migrate onto villi, to be
shed eventually into the intestinal lumen. This pro-
cess results in a continuous renewal of the intestinal
epithelium every 3-4 days [2 1,221. Experiments
were designed to determine where, along the
crypt-villus axis, the dietary sugar signal is perceived.
In one set of experiments, daily biopsies were
taken through the duodenal cannulae, from the
intestines of ruminant sheep which were infused
continuously for 4 days with i)-glucose, while the
animals were maintained on roughage diet. Brushborder membrane vesicles were prepared from the
biopsy samples and the activity and the presence of
SGLT1 were assayed [23]. The activity was first
detected in brush-border membranes isolated from
mature enterocytes after 3 days; it was maximal
after 4 days.
In a wide-ranging series of experiments, it was
noted that an infusion (through the duodenal
cannulae) of i)-glucose for 2 h into the intestinal
lumen of ruminant sheep had no effect on the ability
of existing mature enterocytes to transport 11glucose. However, the presence of functional
SGLT1 was detected 4 days later in the newly
formed mature enterocytes.
One interpretation of these results is that the
signal-receiving site is localized within the crypt.
The induction or programming is rapid and the
observed lag for the appearance of the SGLT1
activity is correlated with cell migration time along
the crypt-villus axis. This is the simplest model for
interpretation of the above results. However, one
cannot exclude the villus location of the receptor. In
this case, the receptor could be linked to crypt
events via a neural or paracrine mechanism.
The profile of induction of SGLTl
along the crypt-villus axis
The profile of induction of SGLT1 along the cryptvillus of the intestine of ruminant sheep infused with
i,-glucose for 2 h and subsequently maintained for a
further 4 days on the standard roughage diet was
determined. Enterocytes were isolated from the
upper villus, lower villus, and crypt cell populations
[24-261. Crude cellular homogenates and brushborder membrane vesicles were isolated from these
cells. The fractions were probed with the SGLT1
antibody. The abundance of, and enrichment of,
mature 75 kDa protein in vesicle fractions (V) over
the cellular homogenates (H) can be seen in Figure
1, lanes 1-4. There is 25-30-fold enrichment in the
abundance of SGLTl protein in the vesicles (V,
lanes 2 and 4) over the cellular homogenate (H,
lanes 1 and 3), indicating that the SGLT1 is mainly
localized on the brush-border membrane of the
Mechanisms and Regulation of Sugar Transport
Figure 2
Figure I
The
profile of induction of SGLTI
crypt-villus axis
along the
H
V
1
2
Lower
villus
H
3
+
crypt-villus axis
The intestines of ruminant sheep were infused with 30mM
solutions of o-glucose as described in the text Enterocytes
from upper villus, lower villus and crypt cells were isolated
from the proximal region of the intestine (24-261 Crude cellular fractions and brush-border membrane vesicles were isolated from these three cell populations Samples of cell
homogenate (H), and membrane vesicles (V) ( I0 pgAanej were
separated on an 8% ( w h ) polyacrylamide gel containing 0 I%
( w h ) SDS, electrotransferred to nitrocellulose membrane and
blotted with the anti-peptide antibody raised against SGLT I as
described in the text The abundance of the band in each tract
was estimated by densitometry
Upper
villus
Na -dependent D-glucose uptake activity along the
Brush-border membrane vesicles were prepared from enterocytes isolated from upper villus, lower villus and crypt cells of
the same tissue used in the studies described in Figure I Initial
rates (4 s) of Na+-dependent D-glucose uptake into membrane
vesicles was determined by a rapid filtration technique [20,23].
Vesicles were suspended in 300mM mannitol, 20mM Hepes/
Tris, pH 74, 0 I mM magnesium sulphate The incubation
media contained either l00mM NaSCN or KSCN and 0 I mM
['4C]~-glucoseSGLT I activity i s the difference between the
value obtained in the presence of N a + and that in the
presence of K + Results shown are the means k S E M (n = 5)
Crwt
V
4
H
5
V
6
upper and lower villus enterocytes. The same
brush-border membrane vesicles could also transport i)-glucose in a Na+-dependent manner (see
Figure 2), indicating the presence of functional
SG1,Tl along the length of the villus. These findings are in contrast to those of Kinter and Wilson
[27] who have reported that the SGLT1 activity is
confined to the enterocytes at the upper third of the
villus. There is little SGLT1 protein and transport
activity associated with the membranes isolated
from the crypt cells. This distribution of the localization of the SGLT1 protein was confirmed by
immunocytocheniical studies (T. P. King and S. P.
Shirazi-Heechey, unpublished work).
A very characteristic 58 kDa immunoreactive
protein was revealed when the crude cellular homogenate (14) and membrane vesicles (V), isolated
from the crypt cell population from animals that
had been infused with i)-glucose were probed with
anti-SGI,Tl antibody (see Figure 1, lanes 5 and 6).
The abundance of this protein was slightly enriched
in the membrane fraction (V, lane 6) with respect to
the cellular homogenate (I I, lane 5).
Neither the 58 nor 75kDa protein was
detected in the comparable fractions isolated from
the cells along the crypt-villus of the intestine of
ruminant sheep infused with mannitol. The follow-
ing findings make the 58 kDa protein a potential
candidate as a SGI,Tl precursor protein: (a) within
2 h of infusion of the intestine a 58 kDa protein is
synthesized in the crypt region; (b) this protein
cross reacts with the SGLT1 antibody; (c) this protein is not glycosylated; (d) as the enterocyte
migrates out of the crypt the abundance of the
58 kDa protein diminishes, while the level of 75 kDa
SGLT1 increases; and (e) the molecular mass of
SGLT1 when deglycosylated is 58 kDa.
W e propose that the introduction of sugars to
the luminal content of ovine intestine stimulates the
synthesis of functional SGLT1. This stimulation is
perceived through a sugar receptor present in the
crypt leading into synthesis of a 58 kI>a protein.
The properties of this protein suggests a precursor-product relationship between the 58 kDa
protein and SGLT 1.
This work was largely supported by The Agricultural
and Food Research Council grant I,KG/257. The finan-
ti57
Biochemical Society Transactions
cia1 support of The Science and Engineering Research
Council and Tenovus is gratefully acknowledged.
S.P.S.-H.is a Wellcome Trust Senior Lecturer.
658
1 Turk, E., Zabel, B., Mundlos, S., Dyer, J. and Wright,
E. M. (1991) Nature (London) 350, 354-356
2 Hopfer, U. (1987) in Physiology of Gastro-Intestinal
Tract (Johnson, L. R., eds.), 2nd edn., pp. 1499- 1526,
Raven Press, New York, NY
3 Semenza, G., Kessler, M., Hosang, M. and Schmidt,
U. (1984) Biochim. Biophys. Acta 779,343-379
4 Wright, E. M., Hirayama, B., Hazama, A., Loo,
D. D. F., Supplisson, S., Turk, E. and Hager, K. H.
(1993) in Molecular Biology and Function of Carrier
Proteins, pp. 230-241, The Rockefeller University
Press, New York
5 Wood, I. S., Scott, D., Heechey, R. B. and ShiraziBeechey, S. P. (1994) Biochem. SOC.Trans., in the
press
6 Lescale-Matys, I,., Dyer, J., Scott, D., Freeman, T. C.,
Wright, E. M. and Shirazi-Reechey, S. P. (1993)
Biochem. J. 291,435-440
7 Hirayama, H. A,, Wong, H. C., Smith, C. D., Hagenbuch, B. A,, Hediger, M. A. and Wright, E. M. (1991)
Am. J. I’hysiol. 261, C296-C304
8 Pajor, A. M., Hirayama, B. A. and Wright, E. M.
(1992) Am. J. Physiol. 263, R489-K495
9 Vazquex, C. M., Wood, I. S., Dyer, J., Planas, J. M.,
Ilundain, A. and Shirazi-Beechey, S. P. (1993)
Biochem. SOC.Trans. 21,4798
10 Ferraris, R. P. and Diamond, J. M. (1989) Annu. Rev.
Physiol. 51, 125-141
11 Farraris, R. P., Villenas, S. A., Hirayama, B. A. and
Lliamond,J. (1 992) Am. J. Physiol. 262, 1060- 1068
12 Shirazi-Beechey, S. P., Hirayama, B. A., Wang, Y.,
Scott, D., Smith, M. W. and Wright, E. M. (1991) J.
Physiol. (London) 437,699-708
13 Freeman, T. C., Wood, I. S., Sirinathsinghji, D. J. S.,
Beechey, R. B., Dyer, J. and Shirazi-Beechey, S. P.
(1993) Biochim. Biophys. Acta 1146,203-212
14 Shirazi-Beechey, S. P., Smith, M. W., Wang, Y. and
James, P. S. ( 1 99 1 ) J. Physiol. (London) 437, 69 1-698
15 Dyer, J., Scott, D., Reechey, R. B., Care, A. D., Abbas,
K. S.and Shirazi-Beechey,S. P. (1994) in Mammalian
Brush-Border Membrane Proteins, Part I1 (Lentze, M.
J., Grand, R. J. and Naim, H. Y., eds.), pp. 65-72,
Thieme Verlag, Stuttgart and New York
16 Scharrer, E., Liebich, H.-G., Raab, W. and Promberger, N. (1979) Zentralbl. Veterinaermed. Reihe A
26,95-105
17 Shirazi-Beechey, S. P., Kemp, R. B., Dyer, J. and
Beechey, R. B. (1989) Comp. Riochem. Physiol. B:
Comp. Biochem. 94,80 1-806
18 Bassett, J. M. (1975) in Digestion and Metabolism in
Ruminants (McDonald, I. W. and Warner, A. C. I.,
eds.), pp. 383-398, University of New England
Publishing Unit, Armidale
19 Shirazi, S. P., Beechey, R. B. and Butterwoth, P. J.
(1981) Biochem. J. 194,803-809
20 Shirazi-Reechey,S. P., Gorvel, J.-P. and Heechey, R. B.
(1988) J. Bioenerg. Biomembr. 20,273-288
21 Attaix, D. and M e s h , J.-C. (1991) Am. J. Physiol. 261,
R811-R818
22 Gordon, J. I. ( 1 989) J. Cell Biol. 10, 1187- 1 194
23 Shirazi-Beechey, S. P., Davies, A. G., Tebbutt, K.,
Dyer, J., Ellis, A,, Taylor, C. J., Fairclough, P. and
Beechey, R. B. (1990) Gastroenterology 98,676-685
24 Brown, P. D. and Sepulveda, F. V. (1985) J. Physiol.
(London) 363,257-270
25 Flint, N., Cove, F. I,. and Evans, G. S. (1991)
Biochem.J. 280,331-334
26 Meddings,J. B., Desouza, Ll., Goel, M. and Thiesen, S.
(1990)J. Clin. Invest. 85, 1099-1 107
27 Kinter, W. H. and Wilson, T. H. (1965)J. Cell Hiol. 25,
19-39
Received 2 March 1994
Hepatic microsomal glucose transport
Ann Burchell
Department of Obstetrics and Gynaecology, University of Dundee Medical School, Ninewells Hospital,
Dundee DDI 9SY, U.K.
Glucose transport and phosphorylation in cells which cannot make
significant amounts of glucose
Glucose is an important metabolic substrate for
mammalian cells. Most tissues cannot make sufficient glucose to maintain their normal levels of
metabolic function. It is therefore important that tisAbbreviation used: ER, endoplasmic reticulum.
sues receive a steady supply of glucose from the
blood and the transport of blood glucose across the
plasma membrane of mammalian cells is a vital
event. One or more plasma-membrane glucose
transport proteins are present in nearly all mammalian cells.
Two different types of plasma membrane
glucose transport protein have been described.
Sodium-glucose co-transporters [ 1,2] actively