Structural conservation in the major facilitator superfamily as revealed by comparative modeling

Abstract

The structures of membrane transporters are still mostly unsolved. Only recently, the first two high-resolution structures of transporters of the major facilitator superfamily (MFS) were published. Despite the low sequence similarity of the two proteins involved, lactose permease and glycerol-3-phosphate transporter, the reported structures are highly similar. This leads to the hypothesis that all members of the MFS share a similar structure, regardless of their low sequence identity. To test this hypothesis, we generated models of two other members of the MFS, the Tn10-encoded metal-tetracycline/H(+) antiporter (TetAB) and the rat vesicular monoamine transporter (rVMAT2). The models are based on the two MFS structures and on experimental data. The models for both proteins are in good agreement with the data available and support the notion of a shared fold for all MFS proteins.

Figure 1.

Flow chart of the modeling procedure. All sequence adjustments were done without insertions of gaps to TM regions (predicted or structural). Model examination was done by testing experimental data on the model. Cycles of manual optimization of the sequence alignment were continued until no further improvement could be achieved.

Figure 2.

(A) Superimposition of GlpT and LacY structures. (B) Superimposition of LacY structure (bright) and model (according to structural alignment) by GlpT (dark). Shown are TM1 (left) and TM5 (right). (C) Superimposition of GlpT structure (bright) and model by LacY (dark). Shown are TM1 (left) and TM5 (right). Models in B and C are based on structural alignment. (D) Superimposition of two LacY models created according to structural alignment with GlpT (bright) or by manually optimized multiple sequence alignment (dark). Shown are TM1 (left) and TM5 (right). A phase shift between the modeled helices is observed in TM5.

Figure 3.

(A) Cytoplasmic view of NEM-accessible residues (mauve) in TetAB model (LacY as template). (B) Helix projection of TetAB model: cytoplasmic and periplasmic view. (C) Electrostatic surface representation of TetAB model. The cytoplasmic view reveals a negatively charged central cavity. (D) Superimposition of two TetAB models based on the two templates, LacY (white) and GlpT (black). Shown are TM2 (left) and TM11 (right). A phase shift of one to two residues is visible at the cytoplasmic half of TM11. That phase shift is corrected after the third turn, and the residues in the rest of the TM overlap. In TM2, a turn shift is observed. Each residue in TM2 of TetAB by GlpT is one helix turn closer to the periplasmic side of the membrane than the corresponding residue in TetAB by LacY.

Figure 4.

Blue areas represent NEM-accessible residues. (A) NEM accessibility to periplasmic half of TM4 (yellow) is prevented by TM1 and TM2. For TM10 (cyan), accessibility is prevented by TM7 and TM8. (B) NEM accessibility to cytoplasmic half of TM1 (red) is prevented by TM4 and TM5. For TM7 (green), accessibility is prevented by TM10 and TM11. (C) Two-dimensional representation of NEM accessibility profile in TetAB (Tamura et al. 2001).

Figure 5.

(A) Membrane-embedded charged residues in rVMAT2 model. (B) Suggested salt bridge between Asp 427 in TM11 and Lys 139 in TM2 of VMAT as it is shown in the model. (C) Multiple sequence alignment of VMATs and VAChTs show that Tyr in TM8 of VMATs is replaced by His in VAChTs. (D) Shown are TM8 and TM10, Tyr 342 in TM8 (which corresponds to His in VAChT) is close to Asp 400 in TM10.