Figure 1.

(a) Robetta model (Model 1) of residues 1-340 of the metalloprotease domain of MecR1 (cartoon). The left half maps each of the five transmembrane helices predicted from sequence (Table 1), colored from blue (N-terminal) to red (C-terminal). The right half depicts this model aligned to a membrane-relaxed structure of the Ste24p membrane metalloprotease (see Fig. S2). (b) Flat scheme showing the overall topology of the metalloprotease domain of MecR1 according to Model 1; light and dark blue for the whole homology model, dark blue for the portion backed up by coevolution data. (c) Flat scheme showing the overall topology of full-length MecR1 according to Model 2, which contemplates the results from the experimental mapping of the transmembrane topology. The gluzincin core of MecR1 (MecR1^GLZ, residues 147–304) is shown in dark blue with red border. In (b,c), the numbers between brackets indicate the amino acid residue where each membrane-embedded helix is proposed to start or finish; the pink sphere approximates the position of the zinc ion; the green arrows labelled F1, F2N/2N, F2C/2C, F3, F4N/4N, F4C/4C, F5/5b and F6 indicate the positions where either eGFP was fused or a TEV recognition site was inserted for the experimental topology mapping.