Bianchi L and Díez-Sampedro A (2010), A single amino acid change converts the sugar s...

XB-ART-41602

PLoS One 2010 Apr 08;54:e10241. doi: 10.1371/journal.pone.0010241.

Show Gene links Show Anatomy links

A single amino acid change converts the sugar sensor SGLT3 into a sugar transporter.

Bianchi L , Díez-Sampedro A .

???displayArticle.abstract???
Sodium-glucose cotransporter proteins (SGLT) belong to the SLC5A family, characterized by the cotransport of Na(+) with solute. SGLT1 is responsible for intestinal glucose absorption. Until recently the only role described for SGLT proteins was to transport sugar with Na(+). However, human SGLT3 (hSGLT3) does not transport sugar but causes depolarization of the plasma membrane when expressed in Xenopus oocytes. For this reason SGLT3 was suggested to be a sugar sensor rather than a transporter. Despite 70% amino acid identity between hSGLT3 and hSGLT1, their sugar transport, apparent sugar affinities, and sugar specificity differ greatly. Residue 457 is important for the function of SGLT1 and mutation at this position in hSGLT1 causes glucose-galactose malabsorption. Moreover, the crystal structure of vibrio SGLT reveals that the residue corresponding to 457 interacts directly with the sugar molecule. We thus wondered if this residue could account for some of the functional differences between SGLT1 and SGLT3. We mutated the glutamate at position 457 in hSGLT3 to glutamine, the amino acid present in all SGLT1 proteins, and characterized the mutant. Surprisingly, we found that E457Q-hSGLT3 transported sugar, had the same stoichiometry as SGLT1, and that the sugar specificity and apparent affinities for most sugars were similar to hSGLT1. We also show that SGLT3 functions as a sugar sensor in a living organism. We expressed hSGLT3 and E457Q-hSGLT3 in C. elegans sensory neurons and found that animals sensed glucose in an hSGLT3-dependent manner. In summary, we demonstrate that hSGLT3 functions as a sugar sensor in vivo and that mutating a single amino acid converts this sugar sensor into a sugar transporter similar to SGLT1.

???displayArticle.pubmedLink??? 20421923
???displayArticle.pmcLink??? PMC2857651
???displayArticle.link??? PLoS One
???displayArticle.grants??? [+]

Species referenced: Xenopus
Genes referenced: slc5a1.2

???attribute.lit??? ???displayArticles.show???

	Figure 1. Amino acid alignment of SGLTs.Four SGLT1s and 3 SGLT2s from different species (h: human, m: mouse, p: pig) have a glutamine at position 457. SGLT1s and SGLT2s proteins transport sugars. The alignment also includes 6 SGLT3s. None of the SGLT3s encodes a glutamine at position 457 (boxed). Rather a glutamate, glycine or serine is present at that position. In SGLT1, Q457 is involved in sugar recognition, binding and translocation [8], [9] and mutation at this site can cause glucose-galactose malabsorption [2], [7]. Stars, semicolons and periods denote identity and conservation respectively. Transmembrane segments (TM) 10 and part of 11 are also shown (based on alignment with vSGLT, whose crystal structure has been solved [9]).
	Figure 2. Substitution of glutamate by glutamine at position 457 in hSGLT3 changes the protein from a glucose sensor to a glucose transporter.A. We measured 50 ÂµM Î±M-glc transport in control oocytes (non-injected) and WT-hSGLT3 and E457Q-hSGLT3 expressing oocytes. Data show the mean Â± SE of the sugar uptake in individual oocytes. The number of oocytes per condition varied from 4 to 8. In oocytes expressing E457Q-hSGLT3, experiments were also carried out in the presence of 0.1 mM phlorizin (Pz), or in absence of Na+. B. Charge uptake and sugar uptake were simultaneously measured in oocytes expressing E457Q-hSGLT3 (filled circles) and Q457E-hSGLT1 (open circles). Each point corresponds to datum obtained from one oocyte clamped in the presence of Î±M-glc. The background uptake of Î±M-glc in non-injected oocytes has been subtracted. The solid line shows a fit of E457Q-hSGLT3 data with a linear regression giving a slope of 2.0 indicating that 2 net positive charges are transported per Î±M-glc molecule.
	Figure 3. Current recordings in WT and in E457Q-hSGLT3.A. Voltage/pulse protocol used in experiments shown in panels BâG. The voltage is initially clamped at â50 mV, then the voltage jumped from â150 to +50 mV with 20 mV increments for 100 ms and finally the voltage is returned to -50 mV. B. Current recordings in a E457Q-hSGLT3 expressing oocyte in presence of Na+. C. Analogous currents are shown from the same oocyte in the presence of 150 mM Î±M-glc. Note that the presteady-state currents disappear after adding sugar. D. Steady-state currents in a E457Q-hSGLT3 expressing oocyte at different voltages (from â150 to +50 mV) in Na+ alone and after adding 150 mM Î±M-glc. E, F, and G. Currents recorded with the same conditions as in A, B, and C respectively in a WT-hSGLT3 expressing oocyte. Note that the presteady-state currents in E457Q-hSGLT3 (B) are larger than in WT-hSGLT3 (E).
	Figure 4. Apparent affinities in E457Q-hSGLT3 are similar to hSGLT1 affinities.A. Depolarizations induced by different sugars in an E457Q-hSGLT3 expressing oocyte. When used at saturating concentrations, different sugars induced similar maximal depolarizations. The concentrations were 500 ÂµM for DNJ; 10 mM for 1DO-glc, glc and Î±M-glc; 50 mM for 2DO-glc, 3DO-glc, 6DO-glc; and 100 mM for galactose. Galactose showed a smaller depolarization because the concentration tested was not saturating. 2DO-glc and DNJ showed small depolarizations and are known to be poor hSGLT1 agonists. B. Depolarizations induced by glucose in one representative oocyte expressing E457Q-hSGLT3 and one hSGLT3 are shown as a function of sugar concentration. Dotted lines represent the fit of the data with equation (1) to estimate the K0.5 and maximal depolarization (ÎVmax) values. K0.5 in this example was 0.4Â±0.1 mM for E457Q-hSGLT3 and 28Â±3 mM for hSGLT3. C. Ki of phlorizin for E457Q-hSGLT3. Graphic illustrates the inhibition of 0.5 mM Î±M-glc-induced depolarization by phlorizin. The amount of Î±M-glc-induced depolarization inhibited by 0â10 ÂµM Pz is plotted against Pz concentration. The curve is a fit to equation (1) (Methods).
	Figure 5. hSGLT3 mediates glucose chemotaxis when expressed in C. elegans sensory neurons.A. Scheme of the behavioral assay we performed in C. elegans (see Methods for details). B. Transgenic animals expressing WT-hSGLT3 or E457Q-hSGLT3 in ASK chemosensory neurons were tested for chemotaxis to 10 mM glucose on pH 6 plates in which a gradient of glucose was established. Attraction index (AI) was (number of animals at glucose spot - number of animals at control spot)/(total number of animals). Thirty to forty animals were assayed in each trial. Number of trials was 6, 5, and 8 respectively. C. elegans expressing WT-hSGLT3 or E457Q-hSGLT3 in ASK sensory neurons were repulsed by glucose. C. The same strains were assayed on pH 5 agar plates. For experiments in which we used phlorizin we incubated the chunk of agar in 10 mM glucose plus 0.1 mM phlorizin, prior to establishing the gradient on the plate. C. elegans expressing WT-hSGLT3 were attracted to glucose whereas animals expressing E457Q-hSGLT3 showed no preference for glucose over the control spot. Importantly, attraction of WT-hSGLT3 expressing C. elegans to glucose was inhibited when Pz was present, confirming that it was mediated by hSGLT3. Number of trials was 10, 14, 5 and 8 for wild type C. elegans, WT-hSGLT3, WT-hSGLT3+Pz and E457Q-hSGLT3 respectively. Data are expressed as mean Â± SE. * indicates p<0.05 by comparison with wild type non-transgenic control C. elegans, by t-test.
	Figure 6. Glucose-induced currents are larger at pH 5 than pH 6 in hSGLT3 expressing oocytes.A. At pH 6, 10 mM glucose induced small currents in WT-hSGLT3 oocytes. However, at pH 5 the glucose-induced current increased approximately 4 fold. In addition, there was an inward current when the extracellular solution pH was reduced from 6 to 5. These glucose- or H+- induced currents returned to the baseline when the glucose was removed from the bath and when the pH was returned to 6. Striped bars indicate the presence of glucose in the bath. B. The same experiment was performed in oocytes expressing E457Q-hSGLT3. Sugar-induced currents were similar at both pHs and H+-induced current was absent. C. The effect of phlorizin was measured on glucose-induced currents at pH 5 in a WT-hSGLT3 expressing oocyte. Phlorizin blocked approximately two-thirds of the total induced current. D. Phlorizin also blocked approximately two-thirds of the current induced by pH 5 in the absence of sugar. These data suggest that phlorizin blocks both the H+-induced and sugar-induced components of current.

References [+] :

Accardi, Secondary active transport mediated by a prokaryotic homologue of ClC Cl- channels. 2004, Pubmed