Fig. 5. Domain composition and sequence alignment of Hexim proteins. (A) Sequence motifs and domain architecture of human Hexim1 (accession no. BAA36166). An N-terminal proposed self-inhibitory region of low sequence homology and variable length is preceding the central nuclear localization signal (NLS). A region of highest identity that includes a negatively charged cluster is suggested to be involved in kinase inhibition. The 12 kDa Cyclin T-binding domain (TBD) covers the C terminus. (B) Alignment of eukaryotic Hexim1-and Hexim2-TBD sequences and secondary structure of the human Hexim1-TBD as determined here. The region of highest identity ranges from D266 to L308, encompassing a flexible linker, the preceding helix a1 and the first coiled coil segment a2 of the TBD. Sequence numbering is according to human Hexim1. Letters below the sequences indicate the heptad repeat positions of the coiled coil segments assigned according to the solution structure.

Fig. 6. Interaction between helix a1 and the first coiled coil segment a2. (A) Close up view on the structure between the preceding helix a1 and the following coiled coil segment. Residues on a1 (T276, L279, Q280) interact with the hydrophobic residues I288 and L292 at positions e and b of the heptad repeat, respectively. Residues in the coiled coil interface L287 and Y291 are marked. (B) Close up view of the superposition of the twenty lowest energy structures around the interface of the preceding helix a1 and the first coiled coil segment. Side chains of residues at the coiled coil heptad positions a and d are displayed in dark and light green, respectively. Side chains of residues that mediate the hydrophobic interactions between helices a1 and a2 are shown in yellow. Structures were displayed with the programs MolMol (1) or PyMOL (DeLano Scientific).

1. Koradi R, Billeter M, Wüthrich K (1996) J Mol Graphics 14:51-55, 29-32.

Fig. 7. Backbone dynamics and structural integrity of Hexim1-TBD. (A) The T1 (Top) and T2 (Middle) relaxation times and [¹H]-¹⁵N heteronuclear NOE values are presented for each residue number. Secondary structure elements are indicated at the top. The two coiled coil segments appeared rather rigid while the linker region 314-318 increased the backbone dynamics locally. The N-and C-terminal residues showed also increased mobility. (B) ¹H/¹⁵N-HSQC overlay plot of three different Hexim1-TBD fragments. Spectra of fragments 242-359 (black), 255-359 (blue) and 255-316 (red) were superimposed, all measured in 20 mM KPi (pH 7.2) buffer at 35°C. The similarity of the peak pattern confirmed the presence of the coiled coil structure and the preceding helix within residues 255-316 for all fragments.

Fig. 8. Titration series of ¹⁵N Hexim1-TBD with unlabelled CycT1 (1-292). The ¹H/¹⁵N HSQC spectra are colored as follows: ¹⁵N-labeled Hexim1-TBD (255-359) to unlabelled CycT1 (1-292) = 1:0 black; 1:0.2 red; 1:0.5 green; 1:1 blue. The concentration of Hexim1-TBD was adjusted to 0.1 mM in all samples. Resonance signals that disappear at a molar ratio of 1:0.5 Hexim1 to CycT1 are marked.

Fig. 9. In vitro GST pull-down assay of the CycT1-Hexim1-TBD interaction. Equal amounts of wild type or mutant TBD proteins were incubated with GST-fused CycT1 (1-292) and precipitated from solution with GSH Sepharose beads. Proteins bound to beads were separated by 17% SDS/PAGE and stained with Coomassie blue. Twenty percent of the input quantity of each TBD is shown in the middle panel. The relative intensities determined from an integration of the bands of three independent experiments are shown below. Error bars give standard errors of the mean. GST-CycT1 efficiently precipitated Hexim1-TBD from solution but not GST alone. As could be seen, all mutant TBD's were precipitated from solution and the strength of the bands showed only for K289A, L292A, E302A, and 3E-A a diminished binding affinity. This indicated that none of the tested point mutations was sufficient to result in complete loss of binding to CycT1, suggesting that the individual contribution of single amino acids is rather low while the interaction covers a large surface.

Fig. 10. Functional analysis of the mutant Hexim1 proteins in cells with lower amounts of flag-tagged Hexim1. (A-C) Multiple amino acid substitutions in TBD disable inhibitory function of Hexim1 in cells. HeLa cells expressed plasmid reporter pG6TAR (0.2 mg). Proteins that were coexpressed from corresponding plasmid effectors (Gal4.CycT1, 0.5 mg; f.Hxm1, 0.5 mg) with the plasmid reporter are presented below the CAT data. The lower panels present the expression of f.Hxm1 proteins as indicated by the arrow. (D) Mutation of positively charged residues K289A or K296A had no effect on the inhibitory function of Hexim1 in cells.

Fig. 11. Far-UV CD spectroscopy of Hexim1-TBD (wt) and the mutant K289A. The mean residue ellipticity of both samples measured over a range from 190 to 260 nm wavelength appeared similar for the mutant protein as for the wild type protein, suggesting that mutation of this site did not significantly change the helical conformation of the protein domain.