SUPPLEMENTARY INFORMATION

Transcription

1 SUPPLEMENTARY INFORMATION doi: /nature11524 Supplementary discussion Functional analysis of the sugar porter family (SP) signature motifs. As seen in Fig. 5c, single point mutation of the conserved SP motif residues of XylE, G83A, E153A, R160A, G340A, R341A, and E397A, led to complete loss of function in both cell-based uptake and proteoliposome-based counterflow assays. Gly 83 and Gly 340, the highly conserved residues in the TM2-3 and TM8-9 motifs, are located on the tight turns connecting TMs 2-3 and TMs 8-9, respectively (Supplementary Figs. 1b & 2). Change of either Gly to Ala may cause disturbance of the local structural integrity, thereby leading to complete loss of transport activity. The charged residues, Glu 153 and Arg 160 on the TM4-5 motif, Arg 341 on the TM 8-9 motif, and Glu 397 on the TM10-11 motif, are involved in the inter-domain contacts (Supplementary Fig. 6). The observation that single point mutation of any of these residues led to complete abrogation of transport activity supports our speculation that the interactions between the intracellular domain and the TMs are functionally critical for XylE. Mutation of Arg404 resulted in more than 75% loss of the transport activity in both assays. Structural examination shows that Arg 404 contributes to the local, but not the inter-domain, H-bond network in the present conformation of XylE (Supplementary Fig. 6). Mutation of the residues in the TM12C motif, E465A or K467A, was largely tolerated, probably because these two residues are barely involved in the interaction network (Supplementary Fig. 6). Notably, the XylE variant containing E222A exhibited little change in transport activities. Supporting 1

2 RESEARCH SUPPLEMENTARY INFORMATION this functional observation, Glu222 is located at one side of the structure and not involved in any H-bond formation in the present conformation. The biochemical characterizations suggested that the SP motifs as well as the intracellular helix domain play an essential role for the function of XylE. Given that extensive H-bonds are formed by these structural elements, some of the residues may be involved in proton-symport. On the other hand, single missense mutation of some residues, such as Glu153, Arg160, Arg341, and Glu397, led to complete loss of activity even in the counterflow assay, suggesting that these residues may play an important role in the structural changes required for the completion of a transport cycle. Elucidation of the functional mechanism of the SP motifs awaits further biochemical, biophysical, and simulation analyses. 2

3 SUPPLEMENTARY INFORMATION RESEARCH Supplementary Figures and Legends 3

4 RESEARCH SUPPLEMENTARY INFORMATION Supplementary Figure 1 XylE is a bacterial homologue of GLUT1-4. a, Phylogenetic tree of GLUT1-4 and their homologues. Multiple sequence alignments were performed with ClustalW and the result was presented with PHYLIP. Sequences of 84 members in the sugar porter family 2.A.1.1 in the Transporter Classification Database (TCDB), excluding 6 redundant sequences for the same proteins, were used for the alignment. b, Sequence alignment of XylE with human GLUT1-4. Secondary structural elements of XylE are indicated above the sequence alignment. Invariant and highly conserved amino acids are shaded yellow and grey, respectively. The conserved SP family signature motifs are underscored with red lines or stars. The residues that are H-bonded to D-xylose are shaded red. The aromatic residues that surround the bound ligands in the crystal structures are shaded orange. Gln 175 and Gly 388 in XylE as well as their corresponding amino acids in GLUT1-4, which are involved in binding to D-glucose, but not D-xylose, are shaded blue. c, Sequence conservation between the E. coli XylE and the human proteins GLUT1-4. The pair-wise sequence comparison was performed with Blastp (NCBI. Basic Local Alignment Search Tool). 4

5 SUPPLEMENTARY INFORMATION RESEARCH Supplementary Figure 2 Sequence alignment of the E. coli XylE with homologs from other organisms. Secondary structural elements of the E. coli XylE are indicated above the sequence alignment. Conserved amino acids are shaded orange, yellow, and grey with decreasing degree of conservations. The conserved SP family signature motifs are underscored with red lines or dots. The listed XylE homologs, from top to bottom, are from Escherichia coli O157:H7 str. EDL933, GI: ; Cyanothece sp. PCC 7822, GI: ; Saccharomyces cerevisiae EC1118, GI: ; Plasmodium knowlesi strain H, GI: ; Arabidopsis thaliana, GI: ; and GLUT1 of Homo sapiens, GI: The sequences were aligned with ClustalW. 5

6 RESEARCH SUPPLEMENTARY INFORMATION Supplementary Figure 3 Recombinant XylE binds to and transports D-xylose. a, ITC titration for WT XylE with 10 mm D-xylose at ph 6.5. The titration signals were fitted by Origin 7. b, XylE transports D-xylose in a ph-dependent manner in the proteoliposome-based counterflow assay. The proteoliposomes were preloaded with 20 mm cold D-xylose at indicated ph. At time point zero, 2 μl concentrated proteoliposomes were diluted into 100 μl KPM buffer (same ph as the proteoliposomes were prepared with) containing 0.83 μm [ 3 H]- D-Xylose. Reaction was stopped at indicated time points by immediately filtering the solution through 0.22 μm filter membranes and washed with ice-cold buffer. The filter was then taken for liquid scintillation counting. All Error bars represent the standard deviation of three independent experiments. c, Transport activity of XylE was measured using a liposome-based counterflow assay at ph 6.5. Details of the experiments can be found in Methods. 6

7 SUPPLEMENTARY INFORMATION RESEARCH Supplementary Figure 4 Structural determination of XylE bound to D-xylose. a, Up to four Hg atoms were found. The anomalous signal for Hg, shown as magenta mesh, was contoured at 5 σ. The refined positions of Hg atoms are represented by grey spheres. b, A stereo view of the 2Fo-Fc electron density for one representative slab is contoured at 1.2 σ. 7

8 RESEARCH SUPPLEMENTARY INFORMATION Supplementary Figure 5 The structure of D-xylose-bound XylE is outward-facing and partly occluded. a, The N- and C-domains of XylE share a similar fold. The two distinct domains can be superimposed with a root-mean-square deviation (rmsd) of 3.11 Å over 153 Cα atoms. b, TM7 and TM10 are bent helices. c, D-xylose resides in the center of the structure, occluded from intracellular side, but solvent accessible from extracellular side of the membrane. A halfway, cut-open view of the surface electrostatic potential is shown here. The van der Waals surface of XylE was calculated with the program HOLE and shown in the right panel, which revealed that the bound D-xylose is accessible to solvent on the extracellular side, but insulated from cytoplasm. The radii of the potential water path are tabulated. The surface electrostatic potential was calculated with PyMol. 8

9 SUPPLEMENTARY INFORMATION RESEARCH Supplementary Figure 6 The polar interactions between the intracellular helices and the TMs of XylE. The residues that mediate the polar interactions between the intracellular helices and the TMs are shown in sticks. The invariant residues between XylE and GLUT1-4 are colored and labeled green. H-bonds are represented by red, dashed lines. A stereoview is shown. IC: intracellular α-helix. 9

10 RESEARCH SUPPLEMENTARY INFORMATION Supplementary Figure 7 D-xylose is coordinated by residues mainly from the C-domain. a, The N- and C-domains are colored green and cyan, respectively. TM7, which is a discontinuous helix, is highlighted in blue. Right panel: D-xylose stands 10

11 SUPPLEMENTARY INFORMATION RESEARCH against the C-domain. D-xylose is shown as white ball-and-sticks. b, D-xylose is bound to XylE by both polar and aromatic residues. A stereoview is shown. Otherwise, the panel is the same as Fig. 2b. c, Tyr298 and Gln415 of XylE contribute to D-xylose coordination through water-mediated H-bonds. The direct H-bonds discussed in the main text are shown as black, dashed lines. Water-mediated H-bonds are highlighted in red. The water molecules are shown as red spheres. A stereoview is shown. Supplementary Figure 8 XylE selectively binds to D-glucose. a, The chemical structures of the sugars examined in the competition assay. b, The binding affinity between D-glucose and XylE was measured with ITC at ph 6.5. The details of the experiment can be found in online Methods. 11

12 RESEARCH SUPPLEMENTARY INFORMATION Supplementary Figure 9 Electron densities for the ligands bound in the three structures. a. The 2Fo-Fc electron density of 6-bromo-6-deoxy-D-glucose (6-Br-D-glucose), shown in cyan mesh, is contoured at 1σ. The anomalous signal for bromide, shown in magenta mesh, is contoured at 5σ. b-d, The omit electron densities, observed in the structures of XylE obtained in the presence of (b) D-xylose, (c) D-glucose, and (d) 6-Br-D-glucose, are contoured at 3σ and shown in blue meshes. The ligands and water molecules, which were not built when calculating the omit densities, are displayed to better illustrate the positions and orientations of the ligands. 12

13 SUPPLEMENTARY INFORMATION RESEARCH Supplementary Figure 10 D-xylose and D-glucose bind to XylE in a similar manner. a, Coordination of D-glucose by XylE. A stereoview is shown. Otherwise, the panel is the same as Fig. 3b. b, D-xylose (black) and D-glucose (silver), which exhibit similar configurations, are located at the same position within XylE. Two perpendicular views are shown. c-d, Comparison of the coordination of D-xylose and D-glucose by XylE. Water-mediated H-bonds are represented by red, dashed lines, and the direct H-bonds discussed in the main text are colored black. The water molecules are shown as red spheres. 13

14 RESEARCH SUPPLEMENTARY INFORMATION Supplementary Table 1. Statistics of data collection and refinement Data D-xylose-bound XylE D-glucose- Hg-SAD Pt-SAD Native bound XylE 6-BrGlcbound XylE Space Group P P P P P Unit Cell (Å) 95.1, , 96.1, , 95.3, , 95.0, , 95.2, Unit Cell ( ) 90, 90, 90 90, 90, 90 90, 90, 90 90, 90, 90 90, 90, 90 Wavelength (Å) Resolution (Å) ( ) ( ) ( ) ( ) ( ) R merge (%) 8.6 (52.7) 6.3 (59.9) 6.5 (61.9) 8.2 (69.0) 8.4 (71.6) I/σ 19.2 (2.6) 21.8 (2.40) 27 (2.5) 24.4 (2.2) 29.3 (3.8) Completeness (%) 98.7 (89.7) (96.8) 98.1 (97.9) 99.6 (97.9) (100.0) Number of 80,336 91, , , ,030 measured reflections Number of unique 19,816 28,438 19,996 17,572 24,981 reflections Redundancy 4.1 (3.2) 3.2 (3.2) 5.4 (4.0) 6.1 (6.0) 7.8 (8.1) Wilson B factor (Å 2 ) R work / R free (%) 22.75/ / /24.64 Number of atoms Overall Protein Ligand Water Other entities Average B value (Å 2 ) Overall Protein Ligand Water Other entities R.m.s. deviations Bonds (Å) Angle ( ) Ramachandran plot statistics (%) Most favourable Additionally allowed Generously allowed Disallowed Values in parentheses are for the highest resolution shell. R merge =Σ h Σ i I h,i -I h /Σ h Σ i I h,i, where I h is the mean intensity of the i observations of symmetry related reflections of h. R=Σ F obs -F calc /ΣF obs, where F calc is the calculated protein structure factor from the atomic model (R free was calculated with 5% of the reflections selected). 14