Enhanced glucose absorption in the rat small intestine following repeated doses of 5-fluorouracil

Many studies demonstrated that 5-fluorouracil (5-FU) treatment of rodents caused the damage of small intestine, resulting in the malabsorption, while we recently found that repeated administration of 5-FU to rats increased Na(+)-dependent glucose absorption in the small intestine. This study investigated the cause of enhanced glucose absorption. 3-O-methyl-d-glucose (3-OMG) absorption was examined using the everted intestine technique. d-Glucose uptake, phlorizin binding, Western blot analysis and membrane fluidity were examined using small intestinal brush-border membrane vesicles (BBMV). Repeated oral administration of 5-FU to rats increased Na(+)-dependent 3-OMG absorption in the small intestine, while alkaline phosphatase activity in the small intestine decreased. Na(+)/K(+)-ATPase activity of 5-FU-treated rats was about three-fold higher than that of control rats. Although the amount of Na(+)-dependent glucose co-transporter (SGLT1) in 5-FU-treated rats decreased, the overshoot magnitude of d-glucose uptake in BBMV was not altered. Maximum binding of phlorizin in 5-FU-treated rats was 1.5-fold larger than that of control rats, but not altered the maximal rate of d-glucose absorption, Michaelis constant of d-glucose and dissociation constant of phlorizin. The membrane fluidity of 5-FU-treated rats increased. The enhanced d-glucose absorption in 5-FU-treated rats seems to occur secondarily due to the activation of Na(+)/K(+)-ATPase activity in basolateral membranes (BLM). Because the amounts of SGLT1 in 5-FU-treated rats decreased, the increase of turnover rate of SGLT1 and/or an expression of unknown Na(+)-dependent glucose co-transporter with high affinity for d-glucose and phlorizin sensitivity would contribute to the enhancement of d-glucose transport in 5-FU-treated rats.     

A phosphorylated phloretin derivative. Synthesis and effect on intestinal Na(+)-dependent phosphate absorption

2'-Phosphophloretin (2'-PP), a phosphorylated derivative of the plant chalcone, was synthesized. The effect of 2'-PP, on Na(+)-dependent phosphate uptake into intestinal brush-border membrane vesicles (BBMV) isolated from rabbit and rat duodenum and jejunum was examined. 2'-PP decreased Na(+)-dependent phosphate uptake into rabbit BBMV with an IC(50) of 55 nM and into rat BBMV with an IC(50) of 58 nM. 2'-PP did not affect Na(+)-dependent glucose, Na(+)-dependent sulfate, or Na(+)-dependent alanine uptake by rabbit intestinal BBMVs. 2'-PP inhibition of rabbit intestinal BBMV Na(+)-dependent phosphate uptake was sensitive to external phosphate concentration, suggesting that 2'-PP inhibition of Na(+)-dependent phosphate uptake was competitive with respect to phosphate. Binding of [(3)H]2'-PP to rabbit intestinal BBMV was examined. Binding of [(3)H]2'-PP was Na(+)-dependent with a K(0.5) for Na(+)(Na(+) concentration for 50% 2'-PP binding) of 30 mM. The apparent K(s) for Na(+)-dependent [(3)H]2'-PP binding to rabbit BBMVs was 58 nM in agreement with the IC(50) for 2'-PP inhibition of Na(+)-dependent phosphate uptake. These results indicate that 2'-PP bound to rabbit or rat intestinal BBMV Na(+)-phosphate cotransporter and inhibited Na(+)-dependent phosphate uptake. In rats treated with 2'-PP by daily gavage, the effect of 2'-PP on serum phosphate, serum glucose, and serum calcium was examined. In a concentration-dependent manner, 2'-PP reduced serum phosphate by 45% 1 wk after starting treatment. 2'-PP did not alter serum calcium or serum glucose. The apparent IC(50) for 2'-PP in vivo was 3 microM.     

Biotin transport in rat intestinal brush-border membrane vesicles

Transport of biotin across rat intestinal brush-border membrane was examined using the brush-border membrane vesicle (BBMV) technique. Uptake of biotin by BBMV is the result of transport of the substrate into the intravesicular space with negligible binding to membrane surfaces. In the presence of a Na+ gradient (out greater than in), transport of biotin was higher with a transient 'overshoot' phenomenon. In comparison, transport of biotin in the presence of a choline gradient (out greater than in) was lower with no 'overshoot' phenomenon. In both jejunal and ileal BBMV, the transport of biotin as a function of concentration was saturable in the presence of a Na+ gradient (out greater than in) but was linear in the presence of a choline gradient (out greater than in). Vmax of the Na+-dependent transport system was 0.88 and 0.37 pmol/mg protein per s and apparent Kt was 7.57 and 7.85 microM in jejunal and ileal BBMV, respectively. Structural analogues inhibited the transport process of biotin. Unlike the electrogenic transport of D-glucose, the transport of the anionic biotin was not affected by imposing a relatively positive intravesicular potential with the use of valinomycin and an inwardly-directed K+ gradient, suggesting that biotin transport is most probably an electroneutral process. This suggestion was further supported by studies on biotin transport in the presence of anions of different lipid permeability. The results of this study demonstrate that biotin transport across rat intestinal brush-border membrane is by a carrier-mediated, Na+-dependent and electroneutral process. Furthermore, transport of biotin is higher in the jejunum than the ileum.     

Divalent metal is required for both phosphate transport and phosphate binding to phosphorin, a proteolipid isolated from brush-border membrane vesicles

The Na+-dependent phosphate transport system in the brush border of rabbit kidney exhibits a positive requirement for a divalent metal ion. Treatment of the brush-border membrane vesicles (BBMV) with a divalent metal chelator in combination with the divalent metal ionophore A23187 dramatically and selectively decreased the Na+-dependent uptake of phosphate; Na+-independent uptake of phosphate was not affected. The combination of chelator plus A23187 also inhibited uptake of phosphate in the presence of Na+ but in the absence of a gradient for sodium across the BBMV. This indicates that the inhibitor is not a result of an alteration in the Na+ gradient by chelator plus ionophore. The inhibited Na+ gradient-dependent transport of phosphate was restored by removing the chelator and adding Mn2+ to the BBMV. The phosphate-binding proteolipid (phosphorin) isolated from rabbit kidney BBMV binds inorganic phosphate with high affinity and specificity. Binding of phosphate to phosphorin is also inhibited by divalent metal chelators and can be restored by addition of a divalent metal. We conclude that a divalent metal ion is required both for the Na+-dependent phosphate transport in BBMV and for the binding of phosphate to the proteolipid phosphorin. These findings are consistent with our suggestion that phosphorin is a component of the Na+-dependent phosphate transport system in renal brush-border membranes.     

Green tea decoction improves glucose tolerance and reduces weight gain of rats fed normal and high-fat diet

Green tea containing polyphenols exerts antidiabetic and antiobesity effects, but the mechanisms involved are not fully understood. In this study, we first analyzed and compared polyphenol compounds [epigallocatechin gallate (EGCG), epigallocatechin (EGC)] in decoction of green tea leaves versus usual green tea extracts. Second, the effects of acute (30 min) or chronic (6 weeks) oral administration of green tea decoction (GTD) on intestinal glucose absorption were studied in vitro in Ussing chamber, ex vivo using isolated jejunal loops and in vivo through glucose tolerance tests. Finally, we explore in rat model fed normal or high-fat diet the effects of GTD on body weight, blood parameters and on the relative expression of glucose transporters SGLT-1, GLUT2 and GLUT4. GTD cooked for 15 min contained the highest amounts of phenolic compounds. In fasted rats, acute administration of GTD inhibited SGLT-1 activity, increased GLUT2 activity and improved glucose tolerance. Similarly to GTD, acute administration of synthetic phenolic compounds (2/3 EGCG+1/3 EGC) inhibited SGLT-1 activity. Chronic administration of GTD in rat fed high-fat diet reduced body weight gain, circulating triglycerides and cholesterol and improved glucose tolerance. GTD-treated rats for 6 weeks display significantly reduced SGLT-1 and increased GLUT2 mRNA levels in the jejunum mucosa. Moreover, adipose tissue GLUT4 mRNA levels were increased. These results indicate that GTD, a traditional beverage rich in EGCG and EGC reduces intestinal SGLT-1/GLUT2 ratio, a hallmark of regulation of glucose absorption in enterocyte, and enhances adipose GLUT4 providing new insights in its possible role in the control of glucose homeostasis.     

Segmental diversity of electrogenic glucose transport characteristics in the small intestines of weaned pigs

It has been shown in several species that the intestinal Na(+)-dependent glucose co-transporter 1 (SGLT1) is more abundant in the jejunum than in ileum. In contrast, the efficiency of intestinal glucose uptake rates in suckling piglets or weaned pigs is not clearly fitting with this segmental distribution. The aim of this study was to evaluate SGLT1 mediated glucose absorption in the jejunum and ileum of growing pigs (Sus scrofa) in more detail. In Ussing chambers, basal short-circuit currents were significantly more positive in the jejunum. It could be demonstrated that the electrogenic ileal glucose transport was significantly more pronounced in different breeds and occurred at 5 mmol?L(-1) glucose 7 times faster in the ileum, although slightly higher jejunal expression of glycosylated SGLT1 was detected by Western blotting. This expression pattern was connected to significantly lower phlorizin sensitivity in the jejunum. As the more efficient ileal glucose absorption was also observable with glucose uptake studies into isolated brush-border membrane vesicles without differences in abundance and activity of the Na(+)/K(+)-ATPase in both segments, we conclude that the segmental differences in porcine glucose transport characteristics may be based on direct or indirect modulations of SGLT1 activity.     

Intestinal transport of monosaccharides and amino acids during postnatal development of mink

Intestinal development is typically studied using omnivores. For comparative purposes, we examined an altricial carnivore, the mink (Mustela vison). In mink, intestinal dimensions increase up to 8 wk after birth and then remain constant (length) or decrease (mass) into maturity despite continuing gains in body mass. Rates of glucose and fructose transport decline after birth for intact tissues but increase for brush-border membrane vesicles (BBMV). Rates of absorption for five amino acids that are substrates for the acidic (aspartate), basic (lysine), neutral (leucine and methionine), and imino acid (proline) carriers increase between birth and 24 h for intact tissues before declining, but increase after 2 wk for BBMV. The proportion of BBMV amino acid uptake that is Na(+)-dependent increases during development but for aspartate is nearly 100% at all ages. Tracer uptake by BBMV can be inhibited by 100 mmol/l of unlabeled amino acid, except for lysine. BBMV uptake of the dipeptide glycyl-sarcosine does not differ between ages, is not Na(+) dependent, and is only partially inhibited by 100 mmol/l unlabeled dipeptide. Despite the ability to rapidly and efficiently digest high dietary loads of protein, rates of amino acid and peptide absorption are not markedly higher than those of other mammals.     

Ontogenic Changes of Villus Growth,Lactase Activity,and Intestinal Glucose Transporters in Preterm and Term Born Calves with or without Prolonged Colostrum Feeding

Oral glucose supply is important for neonatal calves to stabilize postnatal plasma glucose concentration. The objective of this study was to investigate ontogenic development of small intestinal growth, lactase activity, and glucose transporter in calves (n = 7 per group) that were born either preterm (PT; delivered by section 9 d before term) or at term (T; spontaneous vaginal delivery) or spontaneously born and fed colostrum for 4 days (TC). Tissue samples from duodenum and proximal, mid, and distal jejunum were taken to measure villus size and crypt depth, protein concentration of mucosa and brush border membrane vesicles (BBMV), total DNA and RNA concentration of mucosa, mRNA expression and activity of lactase, and mRNA expression of sodium-dependent glucose co-transporter-1 (SGLT1) and facilitative glucose transporter 2 (GLUT2) in mucosal tissue. Additionally, protein expression of SGLT1 in BBMV and GLUT2 in crude mucosal membranes and immunochemical localization of GLUT2 in the enterocytes were determined. Villus height in distal jejunum was lower in TC than in T. Crypt depth in all segments was largest and the villus height/crypt depth ratio in jejunum was smallest in TC calves. Concentration of RNA was highest in duodenal mucosa of TC calves, but neither lactase mRNA and activity nor SGLT1 and GLUT2 mRNA and protein expression differed among groups. Localization of GLUT2 in the apical membrane was greater, whereas in the basolateral membrane was lower in TC than in T and PT calves. Our study indicates maturation processes after birth for mucosal growth and trafficking of GLUT2 from the basolateral to the apical membrane. Minor differences of mucosal growth, lactase activity, and intestinal glucose transporters were seen between PT and T calves, pointing at the importance of postnatal maturation and feeding for mucosal growth and GLUT2 trafficking.     

Dicarboxylate transport in renal basolateral and brush-border membrane vesicles.

Characteristics of succinate transport were determined in basolateral and brush-border membrane vesicles (BLMV and BBMV, respectively) isolated in parallel from rabbit renal cortex. The uptake of succinate was markedly stimulated by the imposition of an inwardly directed Na+ gradient, showing an "overshoot" phenomenon in both membrane preparations. The stimulation of succinate uptake by an inwardly directed Na+ gradient was not significantly affected by pH clamp or inhibition of Na(+)-H+ exchange. The Na(+)-dependent and -independent succinate uptakes were not stimulated by an outwardly directed pH gradient. The Na dependence of succinate uptake exhibited sigmoidal kinetics, with Hill coefficients of 2.17 and 2.38 in BLMV and BBMV, respectively. The Na(+)-dependent succinate uptake by BLMV and BBMV was stimulated by a valinomycin-induced inside-negative potential. The Na(+)-dependent succinate uptake by BLMV and BBMV followed a simple Michaelis-Menten kinetics, with an apparent Km of 22.20 +/- 4.08 and 71.52 +/- 0.14 microM and a Vmax of 39.0 +/- 3.72 and 70.20 +/- 0.96 nmol/(mg.min), respectively. The substrate specificity and the inhibitor sensitivity of the succinate transport system appeared to be very similar in both membranes. These results indicate that both the renal brush-border and basolateral membranes possess the Na(+)-dependent dicarboxylate transport system with very similar properties but with different substrate affinity and transport capacity.     

Oral polyamine administration modifies the ontogeny of hexose transporter gene expression in the postnatal rat intestine

Gastrointestinal mucosal polyamines influence enterocyte proliferation and differentiation during small intestinal maturation in the rat. Studies in postnatal rats have shown that ornithine decarboxylase (ODC) protein and mRNA peak before the maximal expression of brush-border membrane (BBM) sucrase-isomaltase (SI) and the sugar transporters sodium-dependent glucose transporter 1 (SGLT1) and glucose transporter 2 (GLUT2). This study was undertaken to test the hypothesis that the oral administration of spermidine in postnatal rats upregulates the expression of ODC, thereby enhancing the expression of SI and SGLT1 in the brush-border membrane as well as basolateral membrane-facilitative GLUT2 and Na(+)-K(+)-ATPase. Northern and Western blot analyses were performed with antibodies and cDNA probes specific for SI, SGLT1, GLUT2, alpha(1)- and beta(1)-subunits of Na(+)-K(+)-ATPase, and ODC. Postnatal rats fed 6 mumol spermidine daily for 3 days from days 7 to 9 were killed either on postnatal day 10 (Sp10) or day 13 following a 3-day washout period (Sp13). Sp10 rats showed a precocious increase in the abundance of mRNAs for SI, SGLT1, and GLUT2 and Na(+)-K(+)-ATPase activity and alpha(1)- and beta(1)-isoform gene expression compared with controls. ODC activity and protein and mRNA abundance were also increased in Sp10 animals. The increased expression of these genes was not sustained in Sp13 rats, suggesting that these effects were transient. Thus, 3 days of oral polyamine administration induces the precocious maturation of glucose transporters in the postnatal rat small intestine, which may be mediated by alterations in ODC expression.     

Na-glucose and Na-neutral amino acid cotransport are uniquely regulated by constitutive nitric oxide in rabbit small intestinal villus cells

Na-nutrient cotransport processes are not only important for the assimilation of essential nutrients but also for the absorption of Na in the mammalian small intestine. The effect of constitutive nitric oxide (cNO) on Na-glucose (SGLT-1) and Na-amino acid cotransport (NAcT) in the mammalian small intestine is unknown. Inhibition of cNO synthase with N(G)-nitro-l-arginine methyl ester (L-NAME) resulted in the inhibition of Na-stimulated (3)H-O-methyl-D-glucose uptake in villus cells. However, Na-stimulated alanine uptake was not affected in these cells. The L-NAME-induced reduction in SGLT-1 in villus cells was not secondary to an alteration in basolateral membrane Na-K-ATPase activity, which provides the favorable Na gradient for this cotransport process. In fact, SGLT-1 was inhibited in villus cell brush-border membrane (BBM) vesicles prepared from animals treated with L-NAME. Kinetic studies demonstrated that the mechanism of inhibition of SGLT-1 was secondary to a decrease in the affinity for glucose without a change in the maximal rate of uptake of glucose. Northern blot studies demonstrated no change in the mRNA levels of SGLT-1. Western blot studies demonstrated no significant change in the immunoreactive protein levels of SGLT-1 in ileal villus cell BBM from L-NAME-treated rabbits. These studies indicate that inhibition of cNO production inhibits SGLT-1 but not NAcT in the rabbit small intestine. Therefore, whereas cNO promotes Na-glucose cotransport, it does not affect NAcT in the mammalian small intestine.     

Decreased D-glucose transport across renal brush-border membrane vesicles from streptozotocin-induced diabetic rats

The uptake of Na(+)-dependent D-glucose by renal brush-border membrane vesicles (BBMV) isolated from streptozotocin-induced diabetic rats was decreased as compared with controls. Since a Vmax of 4.8 nmol/mg protein per 30 s in diabetic BBMV was significantly decreased as compared with that of controls (Vmax = 7.0 nmol/mg protein per 30 s) without changing an apparent affinity for D-glucose, the decrease in the Na(+)-dependent D-glucose uptake in diabetic rats is likely to be due to the reduction in the number of the transporter. These results are also confirmed by the binding study of [3H]phlorizin to diabetic BBMV. When the blood glucose level is lowered in diabetic rats by both the treatment with insulin and starvation, the decreased Na(+)-dependent D-glucose uptake is returned to control level. These results suggest that Na(+)-dependent D-glucose reabsorption through the apical membrane in proximal tubular kidney cells is dynamically regulated by the change in blood glucose level.     

Kinetics of D-glucose transport across the intestinal brush-border membrane of the cat

1. D-glucose transport across the intestinal brush-border membrane of the cat, a carnivorous animal, was investigated using isolated brush-border membrane vesicles (BBMV). Kinetic experiments were performed under zero-trans conditions (initial [Na+]in and [Gluc]in = O) with the transmembrane electrical potential difference clamped to zero. 2. D-glucose uptake by the BBMV was strongly stimulated by an inwardly directed Na+-gradient. Uptake under Na+-free conditions seemed to occur by simple diffusion. 3. The apparent kinetic constants (Vmax, Km) of Na+-dependent D-glucose transport were computed by forcing initial uptake rates at 0.002-10.0 mmol/l D-glucose to either a Michaelis-Menten type equation with a single or with two carrier-mediated components. 4. Best fit of the experimental data was obtained with the two-component model indicating the existence of two Na+-dependent carrier-mediated mechanisms. System 1 and system 2 differ with respect to the transport velocity as well as the substrate affinity constants with Vmax being 2.5-fold and Km being 5-fold higher for system 1 compared with system 2.     

Aboral changes in D-glucose transport by human intestinal brush-border membrane vesicles.

D-Glucose transport was investigated in isolated brush-border membrane vesicles from human small intestine. Characteristics of D-glucose transport from the jejunum were compared with that in the mid and terminal ileum. Jejunal and mid-ileal D-glucose transport was Na+-dependent and electrogenic. The transient overshoot of jejunal D-glucose transport was significantly greater than corresponding values in mid-ileum. The terminal ileum did not exhibit Na+-dependent D-glucose transport, but did exhibit Na+-dependent taurocholate transport. Na+-glucose co-transport activity as measured by tracer-exchange experiments was greatest in the jejunum, and diminished aborally. We conclude that D-glucose transport in man is Na+-dependent and electrogenic in the proximal intestine and directly related to the activity of D-glucose-Na+ transporters present in the brush-border membranes. D-Glucose transport in the terminal ileum resembles colonic transport of D-glucose.     

Kinetics of D-glucose and L-leucine transport into sheep and pig intestinal brush border membrane vesicles

The kinetic parameters (Vmax, Kt) of Na+-dependent D-glucose transport into brush border membrane vesicles (BBMV) from sheep and pig jejunum were determined. Due to the fermentation of ingested carbohydrates in the rumen the small intestine of ruminants (sheep) has to absorb much less glucose than the small intestine of monogastric omnivores (pigs) or herbivores. Kinetic analysis of the concentration dependence of D-glucose transport revealed a ten-fold smaller Vmax value combined with a five times lower Kt value in sheep BBMV compared with pig BBMV. The Vmax value for L-leucine transport did not differ between the two species investigated, whereas the Kt value in the sheep exceeded that in the pig. It is concluded from these results that the mechanism for Na+-dependent D-glucose transport in ruminants is adapted to the small amounts of carbohydrates reaching the small intestine.     

Na(+)-dependent fructose transport via rNaGLT1 in rat kidney

We found a system of Na(+)-dependent uptake of fructose by rat renal brush-border membrane vesicles. It consisted of two saturable components, and was thought to involve at least two transporters. rNaGLT1, a novel glucose transporter in rat kidney, showed fructose uptake as well as alpha-methyl-D-glucopyranoside uptake by transfected HEK293 cells. The features of the lower affinity type of fructose transporter in the brush-border membranes, such as affinity and substrate recognition, were very comparable with those of rNaGLT1-transfected HEK293 cells. These results indicated that rNaGLT1 is a primary fructose transporter in rat renal brush-border membranes.     

Enhancement of sucrase-isomaltase gene expression induced by luminally administered fructose in rat jejunum

We have previously shown that feeding a diet containing sucrose to rats causes an elevation of sucrase-isomaltase (SI) mRNA level in the jejunum. In this study, we examined whether the SI mRNA level could be directly elevated by administration of one of the constituting monosaccharides (i.e., glucose and/or fructose). Gastric intubation of a sucrose solution caused increases in both sucrase activity and SI mRNA level in the jejunum. Intrajejunal intubation of fructose, but not glucose, led to an elevation of sucrase activity and SI mRNA level. To examine whether fructose directly affects the gene expression of SI at the segment where the absorption of this sugar takes place or the sugar-induced increase in the gene expression of SI is secondary to any possible changes in the level(s) of certain hormonal factor(s) in the blood stream, a solution containing either fructose or glucose was simultaneously perfused into two consecutive cannulated and irrigated loops of jejunum that were not isolated from blood circulation. Compared with the loop perfused with glucose, the loop perfused with fructose exhibited significantly greater sucrase activity and SI mRNA level as well as the elevated GLUT5 mRNA level. These results suggest that fructose is capable of directly increasing the gene expression of SI and GLUT5 in the confined segment where fructose is absorbed.     

Interactions of [14C]phosphonoformic acid with renal cortical brush-border membranes. Relationship to the Na+-phosphate co-transporter

Since phosphonoformic acid (PFA) acts as a specific competitive inhibitor of Na+-Pi co-transport across renal brush-border membrane (BBM), we employed the [14C]PFA as a probe to determine the mechanism of its interaction with rat renal BBM. The binding of [14C]PFA to BBM vesicles (BBMV), with Na+ present in extravesicular medium (Na+o), was time- and temperature-dependent. The replacement of Na+o with other monovalent cations reduced the PFA binding by -80%. Cl- was the most effective accompanying monovalent anion as NaCl for maximum PFA binding. The Na+o increased the apparent affinity of BBMV for [14C]PFA binding, but it did not change the maximum binding capacity. The maximum [14C]PFA binding was achieved at Na+o approximately equal to 50 mM. The extent of Na+-dependent [14C]PFA binding correlated (r = 0.98; p less than 0.01) with percent inhibition by an equimolar dose of PFA of the (Na+o greater than Na+i)-dependent BBMV uptake of 32Pi. Intravesicular Na+ (Na+i) decreased [14C]PFA binding, on BBMV, and this inhibition by Na+i was dependent on the presence of Na+o. The increase in Na+i, at constant [Na+]o, decreased the Vmax, but not the Km, for [14C]PFA binding on BBMV. Bound [14C]PFA was displaced from BBMV by phosphonocarboxylic acids proportionally (r = 0.99; p less than 0.05) to their ability to inhibit (Na+o greater than Na+i)-gradient-dependent Pi transport, whereas other monophosphonates, diphosphonates, L-proline, or D-glucose did not influence the [14C]PFA binding. The Na+-dependent binding of [14C]PFA and of [3H]phlorizin by BBMV was 10 times higher than binding of these ligands to renal basolateral membranes and to mitochondria. [14C]PFA probably binds onto the same locus on the luminal surface of BBM, where Pi and Na+ form a ternary complex with the Na+-Pi co-transporter.