Facilitative plasma membrane transporters function during ER transit

Abstract

Although biochemical studies suggested a high permeability of the endoplasmic reticulum (ER) membrane for small molecules, proteomics identified few specialized ER transporters. To test functionality of transporters during ER passage, we tested whether glucose transporters (GLUTs, SGLTs) destined for the plasma membrane are active during ER transit. HepG2 cells were characterized by low-affinity ER transport activity, suggesting that ER uptake is protein mediated. The much-reduced capacity of HEK293T cells to take up glucose across the plasma membrane correlated with low ER transport. Ectopic expression of GLUT1, -2, -4, or -9 induced GLUT isoform-specific ER transport activity in HEK293T cells. In contrast, the Na(+)-glucose cotransporter SGLT1 mediated efficient plasma membrane glucose transport but no detectable ER uptake, probably because of lack of a sufficient sodium gradient across the ER membrane. In conclusion, we demonstrate that GLUTs are sufficient for mediating ER glucose transport en route to the plasma membrane. Because of the low volume of the ER, trace amounts of these uniporters contribute to ER solute import during ER transit, while uniporters and cation-coupled transporters carry out export from the ER, together potentially explaining the low selectivity of ER transport. Expression levels and residence time of transporters in the ER, as well as their coupling mechanisms, could be key determinants of ER permeability.

Figure 1.

Sugar flux analysis in HepG2 cells with novel set of sensors. A, B) FRET analysis of cytosolic and ER glucose levels with FLIPglu-Δ13V series sensors in HepG2 cells. Glucose dynamics were measured in the cytosol (black) and ER (red) of HepG2 cells expressing FLIPglu-600μΔ13V (A) and FLIPglu-30μΔ13V (B). C, D) Analysis of the relationship dependence between extracellular concentration and cytosolic concentration (C) and the relationship between cytosolic concentration and ER concentration (D). Cytosolic and ER glucose concentrations were calculated from [C_cyt] or [C_ER] = K_d × (1 − R_t)/(R_t − R_min) at steady-state level. The concentrations were fitted by Michaelis-Menten kinetics, C_max × C_c/(C_0.5 + C_c). ER concentration was fitted using the same C_0.5 as determined across the plasma membrane. Cells were perfused with different external glucose concentrations. Bars indicate loading time of external glucose concentrations (A: 0.25, 0.75, 2, 5, 10, 25, 40 mM; B: 5, 10, 25, 50, 75, 100, 250 μM) during continuous perfusion with Hanks’ balanced buffer. Quantitative data were derived from pixel-by-pixel integration of ratiometric images. Fluorescence intensities [arbitrary units (AU)] for individual eCFP (ET470/24m) and Venus (ET535/30m) emission channels were monitored with eCFP excitation (ET430/24x), and the FRET index for Venus and eCYP was determined [y axis corresponds to sensitized fluorescence F_c, i.e., background and bleedthrough were corrected using Venus excitation (ET500/20x) and normalized to donor emission; note that peak intensities were used and that only the short wavelength emission peak of eCFP was used for donor emission]. FRET images were acquired every 5 s, and cytosolic steady-state glucose levels and accumulation and elimination rates were analyzed. Data are means ± sd (n=3–6).

Figure 2.

FRET analysis of glucose flux with FLIPglu-Δ13V series sensors in the cytosol or ER of HEK293T cells. A, B) Glucose dynamics were measured in HEK293T cells coexpressing FLIPglu-600μΔ13V (A) or FLIPglu-30μΔ13V (B) either in the cytosol (black trace) or ER (red trace). Note that cytosol and ER traces are ratios. C) Images of an individual cell using pseudocolors for the cytosolically expressed sensor FLIPglu-30μΔ13V (data shown in B). Scale at right corresponds to ratio (normalized F_c/D). D) Relationship dependence between extracellular concentration and normalized F_c/D. FRET images were acquired, and data were analyzed as in Fig. 1. Data are means ± sd (n=10–20).

Figure 3.

FRET analysis of glucose flux using FLIPglu-600μΔ13V in the cytosol or ER of HEK293T cells expressing GLUT transporters. Glucose dynamics were measured in the cytosol (black trace) or ER (red trace). A–E) Glucose dynamics in HEK293T cells coexpressing either GLUT1 (A), GLUT1-R126H (B), GLUT2 (C), GLUT4 (D), or GLUT9 (E) with FLIPglu-600μΔ13V. F–H) Analysis of relationship dependence between extracellular concentration and cytosolic concentration: GLUT1 and GLUT1-R126 (F, H); GLUT2 (G); GLUT4 (H). I–K) Analysis of relationship between cytosolic concentration and ER concentration: GLUT1 and GLUT1-R126H (I); GLUT2 (J); GLUT4 (K). Cells were perfused with different external glucose concentrations. FRET images were acquired, and data were analyzed as in Fig. 1. Data are means ± sd (n=8–36).

Figure 4.

Protein level of GLUT1 between HepG2 cells and HEK293T cells expressing GLUT1 and the localization of GLUT1-GFP fusion protein. A) Protein gel blot analysis of GLUT1 levels in HepG2 cells as compared to HEK293T cells expressing GLUT1. Microsomal membrane fractions were extracted and separated by SDS-PAGE (4–20% gradient gel). Actin levels were measured as a control. B) Localization of GLUT1-GFP in HEK293T cells. GLUT1-GFP fusion protein was expressed in HEK293T cells, and GFP-derived fluorescence was analyzed by confocal microscopy. Z sections (71 images, 10.3 μm) are marked as sections z₁ and z₂. Scale bar = 5 μm (41.5 pixels).

Figure 5.

A) FRET analysis of glucose flux using FLIPglu-600μΔ13V in the cytosol or ER of HEK293T cells expressing SGLT1. B) Relationship dependence between extracellular concentration and cytosolic concentration, analyzed as in Fig. 1B. Glucose dynamics were measured in the cytosol (black trace) or ER (red trace) in HEK293T cells coexpressing SGLT1 with FLIPglu-600μΔ13V. FRET images were acquired, and data were analyzed as in Fig. 1. Data are means ± sd (n=25–30).