R [m2kql5edki [v9eel12lskii [n16yhl19eneiii [v23arl26kkliv v30g

and reveals four so-called heptad repeats typical of all coiled coils. Each of these repeating heptads consists of seven amino acid residues, corresponding to two helical turns (see below), and is generically represented by [abcdefg]., where i stands for the heptad number (heptads I-IV in the present case). This is conveniently illustrated in a helical-wheel diagram (Fig. 1), where all residues along the four heptads of the GCN4 coiled coil are projected onto a plane perpendicular to the supercoil axis.

Positions a and d (denoted by lower-case residue numbers in the above sequence) play a key role as they constitute the dimerization interface, that is, the hydrophobic core that is shielded from the aqueous environment upon coiled-coil assembly. These positions are usually occupied by hydrophobic, often branched amino acid residues, such as leucine, isoleucine, and valine (McLachlan and Stewart 1975; Conway and Parry 1990, 1991). In the case of the GCN 4 coiled coil, valine is the predominant residue found at the a positions, whereas all d

1-29

1-29

Fig. 1 Helical-wheel diagram of the homodimeric GCN4 leucine-zipper coiled coil, corresponding to residues 249-279 of the full-length GCN4 protein. View from N-terminus. Residues next to the viewer are surrounded by circles. Single-letter code for amino acids is used. Crossed arrows in the center denote interactions in the hydrophobic core, whereas dashed arrows represent inter- and intrahelical salt bridges (copyright Wiley-VCH Verlag GmbH & Co. KGaA, reproduced with permission from Portwich et al. 2007)

Fig. 1 Helical-wheel diagram of the homodimeric GCN4 leucine-zipper coiled coil, corresponding to residues 249-279 of the full-length GCN4 protein. View from N-terminus. Residues next to the viewer are surrounded by circles. Single-letter code for amino acids is used. Crossed arrows in the center denote interactions in the hydrophobic core, whereas dashed arrows represent inter- and intrahelical salt bridges (copyright Wiley-VCH Verlag GmbH & Co. KGaA, reproduced with permission from Portwich et al. 2007)

positions carry leucine residues. Like many similar dimeric transcription factors, GCN4 displays an asparagine residue (N16) in the a position of the third heptad (Hurst 1994, 1995), which is crucial for dimerization. Substitution of this conserved asparagine by other amino acid residues, even glutamine, gives rise to a mixture of dimers and trimers (Harbury et al. 1993; Potekhin et al. 1994; Knappenberger et al. 2002) or destabilizes the dimeric coiled coil (Gonzalez et al. 1996). More generally, the residues in the hydrophobic core positions a and d act as a switch for the transition from the native dimeric to trimeric coiled-coil structures, and even very subtle modifications, such as supposedly conservative amino acid substitutions may exert a tremendous effect on coiled-coil stoichiometry (Portwich et al. 2007). Positions b, c, e, f, and g face the aqueous outside of the assembly and are mostly occupied by hydrophilic residues. In particular, the positions neighboring the hydrophobic core, that is, e and g, are rich in glutamic acid and lysine residues.

The X-ray structure of the parallel GCN4 homodimer (O'Shea et al. 1991; Fig. 2) confirmed the model of knobs-into-holes packing suggested by Crick (1953), in which each of the hydrophobic side chains at positions a and d of either helix protrudes into a cavity molded by the side chains of four residues of the opposing helix. Importantly, a slight distortion of both constituent helices as compared with a typical monomeric or globular a-helix causes the helical pitch to amount to 3.5 residues rather than the usual 3.6 residues. Therefore, each heptad corresponds to exactly two helical turns, allowing all a and d side chains along all four heptads to contact the hydrophobic helix-helix interface.

The glutamic acid and lysine residues at positions e and g stabilize the assembly in a twofold manner. On one hand, the methylene groups of their side chains shield the hydrophobic core from water (Alber 1992; Hodges et al. 1994). On the other hand, oppositely charged residues in these positions form inter- and intramolecular salt bridges (O'Shea et al. 1991). In the GCN4 leucine zipper, intrahelical salt bridges exist between residues K8 (g) and E11 (c) as well as E22 (g) and R25 (c), which most likely competes with an interhelical salt bridge between E22 (g) and K27 (e'). Two other interhelical salt bridges are established between E20 (e) and K15 (g') as well as K27 (e) and E22 (g'). Since these intermolecular salt bridges are not entirely surrounded by water, but partially exposed to the hydrophobic core, ion-pair formation contributes considerably to coiled-coil stability (Krylov et al. 1994; Lavigne et al. 1996; Lumb and Kim 1995). Not unexpectedly, introduction of like charges into juxtaposing residues in positions g and e' or g' and e drastically destabilizes bZip coiled coils (O'Shea et al. 1992).

Close inspection of the crystal structure (Fig. 2) and, more obviously, of the pattern of salt bridges highlighted in the helical-wheel diagram (Fig. 1) reveals that the homodimeric GCN4 coiled coil is not strictly symmetric with respect to the supercoil axis. Also, the core-flanking positions e and g are not equivalent, as might naively be deduced from the helical-wheel diagram. This asymmetry manifests itself in the observation that position g is significantly more sensitive towards amino acid substitutions than position e (Portwich et al. 2007).

Fig. 2 The 1.8-A X-ray structure of the homodimeric GCN4 leucine-zipper coiled coil, corresponding to residues 249-279 of the full-length GCN4 protein (PDB accession code 2ZTA; O'Shea et al. 1991). (a) Side view, N-terminus at top. (b) Axial view from N-terminus. Side chains of residues at the hydrophobic core positions a and d are shown in black, side chains of residues at the core-flanking positions e and g in light gray

Fig. 2 The 1.8-A X-ray structure of the homodimeric GCN4 leucine-zipper coiled coil, corresponding to residues 249-279 of the full-length GCN4 protein (PDB accession code 2ZTA; O'Shea et al. 1991). (a) Side view, N-terminus at top. (b) Axial view from N-terminus. Side chains of residues at the hydrophobic core positions a and d are shown in black, side chains of residues at the core-flanking positions e and g in light gray

Fig. 2 (continued)

0 0

Post a comment