Fig. 1. (A) Ribbon representation of the C-terminal domain of S.cerevisiae Tup1 looking down the central axis of the propeller onto the top surface. (B) View rotated 90° about the horizontal axis in (A), looking at the propeller from the side. The blades as well as the ‘top’ and ‘bottom’ surfaces are labeled according to convention and the N- and C-termini are marked. The broken chain between blades 1 and 2 (**) and between strands 5C and 5D in blade 5 (*) represents disordered regions. Side chains important for the Tup1–Matα2 interaction are shown in green as ball-and-stick models. (A) and (B) were generated with RIBBONS (Carson, 1997). (C) Secondary structure elements mapped to the sequence of the C-terminal domain of Tup1. Residue numbers are with reference to the full-length Tup1 protein and secondary structural elements, which were assigned by PROCHECK (Laskowski et al., 1993), are depicted with a cylinder for a 3₁₀ helix and ‘→’ for a β strand. The positions where mutations affect the Tup1–Matα2 interaction (Komachi and Johnson, 1997) are shown in blue; the positions of the structural mutations (Carrico and Zitomer, 1998) are shown in orange; and the residues with poor density are shown in green. The double underline is the site of the three residue insertion linking amino acids 388 and 432. (D) Alignment of the seven WD40 repeats of Gβ indicating residues that contact phosducin (Gaudet et al., 1996) (in blue), Gα (Sondek et al., 1996) (in red), or both phosducin and Gα (orange).

Fig. 2. Packing of the N-terminal fragment against the propeller. (A) Extended β sheet formed by blade 6 and strands n1 and n2 from the N-terminal fragment. Hydrogen bonds are indicated with dashed lines. (B) Stereoview of contacts between the N-terminal fragment (283–331) and the propeller. Side chain and backbone atoms are superimposed on the C_α trace. Some side chain labels were omitted for clarity. The approximate boundaries of each blade are shown in different colors (blade 5, light green; blade 6, blue; blade 7, red; N-terminal fragment, gray). This figure was generated with RIBBONS (Carson, 1997).

Fig. 3. (A) Superposition of all seven blades of Tup1. The C_α backbone for each of the seven blades was aligned with respect to strands 3B and 3C. The four conserved side chains involved in the intrablade hydrogen-bonding network, referred to as the structural tetrad, are shown for blade 3 as ball-and-stick models colored according to atom type and hydrogen bonds are indicated with dashed lines. A representative blade from Gβ, blade 4, is shown for reference. Blade 1, light green; blade 2, red; blade 3, green; blade 4, purple; blade 5, pink; blade 6, cyan; blade 7, orange; Gβ, brown. (B) Superposition of the WD40 propellers. The C_α backbones for Gβ (blue; 1TBG) and Tup1 (yellow), which align with an r.m.s.d. of 1.3 Å, are shown in the same view as Figure 1A and labeled. Least-squares alignments for (A) and (B) were done in O (Jones et al., 1991) and figures generated with RIBBONS (Carson, 1997).

Fig. 4. Sequence identity of seven fungal Tup1 proteins mapped onto the molecular surface and C_α backbone of the C-terminal domain of the Tup1 structure. (A) View looking onto the top surface of the molecule, similar to the view in Figures 1A and 3B. (B) View rotated 180° about the vertical axis in (A), looking onto the bottom surface of the molecule. The color gradient is a continuum from white to purple to pink, representing residues with <50% identity, residues with 75% identity and residues with 100% identity, respectively. Strand 4D displays significant sequence identity when compared with all of the other outer strands of the propeller. Tup1 sequences (P16649, P56094, P56093, O14432, Q09715, P78706 and O76734) were aligned using ClustalW 1.7 and displayed in GRASP (Nicholls et al., 1991).