| ECCC-3 Contents Page | THEOCHEM Home Page | Elsevier Chemistry Home Page |
Table 1 | Figure 1 | Figure 2 | Figure 3 | Figure 4 | Figure 5
zinc, imidazole, hydrogen sulfide, zinc finger, quantum mechanics
We have performed ab initio computations examining the structure and relative acidity for a model (Zn(II)-(H2S)2(Imidazole)2) for the zinc finger binding site. All possible deprotonated forms were studied. The lowest energy complex was a mono-deprotonated form (Zn(II)-(H2S)(HS-)(Imidazole)2). In addition, we found that a triply deprotonated form (Zn(II)-(H2S)-(HS-)(deprotonated imidazole)2) rearranges to unique, but not unexpected forms.
Many proteins, such as the transcription factors, will recognize and bind to relatively long stretches of DNA. There are a variety of proteins motifs for binding to a nucleic acid, including: zinc fingers, helix-turn-helix, homeodomains and leucine zippers. The zinc finger is the most prevalent of the various DNA-binding motifs [1-5]. The zinc finger domain is relatively small (consensus sequence shown in Figure 1) and will recognize only short segments of DNA [1,5]. The DNA-sequence selectivity is changed by mutating a minimal number of residues in the zinc finger domain [1,6,7]. Thus zinc finger domains represent building modules with easily modified DNA-sequence specificity. DNA-binding proteins will often contain a linear series of zinc fingers allowing for the recognition of a long DNA sequence (e.g., ZIF268 contains three [8] and transcription factor IIIA (TFIIIA) contains nine zinc finger domains [9]). The structure and function of zinc fingers can be summarized as:
Zinc fingers, as the name implies, are Zn(II) ion binding domains. In the absence of Zn(II) ions, the protein is largely unfolded [10]. Addition of Zn(II) ions results in the protein folding to a relatively compact domain [1,5]. In the original Zinc Fingers, the Zn(II) ion is tetrahedrally bound to four ligands from the protein [2]: the side chains from two cysteines and two histidines (Figure 2). The conserved hydrophobic residues (Red in Figure 1 and 2) form a core of the protein that is essential for proper folding [10]. Residues 13, 16 and 19 are responsible for DNA recognition (Figure 1 and 2). The zinc finger domain contains an unusually large number of secondary structure elements, thus many positions are intolerant to proline substitution [1].
At pH 5.5 and below, but in the presence of Zn(II) ions, the zinc finger is in a slow equilibrium between the folded form and an unfolded form [9,11,12]. Recent pH titration studies have shown that the unfolding is a two proton process that ultimately results in the protonation of both imidazole rings. The NMR studies on a consensus sequence zinc finger [11] reveals that there is an intermediate structure in which one imidazole is protonated and no longer complexed with the Zn(II) ion leaving only three protein-supplied ligands about the Zn(II) ion. The coordination number of the resulting Zn(II) ion was not reported.
We have a long standing interest in models for metal ion binding sites [13-19] and now turn to models for the zinc finger metal ion site. Previous researchers performing computational studies on similar systems (e.g., alcohol dehydrogenase) have chosen to doubly deprotonate (on sulfur) the complex to provide an overall neutral Zn(II) ion site [20-23]. We have interest in examining whether this deprotonation model is appropriate for zinc fingers.
Ab initio quantum mechanical calculations were performed using the Gaussian94 program [24] on the Cray C916/16512 at the Pittsburgh Supercomputing Center and the T90 at the North Carolina Supercomputing Center. Optimizations employed the eigenvalue-following methods [25,26] with the initial force constants calculated analytically. The internal 3-21g* basis set was employed. All resulting energy minimized structures were shown to be energy minima by confirming that all harmonic frequencies were real. The relative energies in the tables are reported in kcal/mole as both the computed difference in energy (comp) and corrected for zero point energy and to 298° (corr) using the thermochemistry code in the Gaussian94 program [24]. The energies reported in the figures are corrected. In addition, single point computations were performed using a 6-311g* basis set at the 3-21g* optimized geometry.
The results for this study are summarized in Table 1 and Figure 3. The lowest energy complex (C) is deprotonated on only one hydrogen sulfide. That is, the neutral complex (doubly deprotonated) is not the lowest energy. The lowest energy complex (C) is substantially more stable (by 72.4 kcal/mol) than tetraaquozinc ( Zn(II)(H2O)4 ). It is interesting that the protonation and deprotonation energies for the lowest energy complex (C) are very similar. Triple deprotonation (both imidazoles and one H2S) of the parent complex resulted in rearrangement with the H2S no longer in a contact ion pair with the Zn(II) ion (Figure 4, Figure 5).
The Zn-S distance is more susceptible to deprotonation than is Zn-N(Imidazole) distance (d(Zn-S)=0.254 A, 10.6%; d(Zn-N)=0.057 A, 3.0%). The N-Zn-S angle was sensitive to the protonation state of the sulfur. This is clearly seen in the lowest energy complex (C), for which the N-Zn-SH2 angle is 102.2° and the Imd-Zn-SH- is 114.3°. However, the N-Zn-N(Imd or Im-) angle showed only slight sensitivity to the protonation state. Furthermore, deprotonation on sulfur is preferred by an average of 38.6 kcal/mol as compared to deprotonation of an imidazole.
Several comments can be made about the pH titration of zinc fingers. During the protonation and hydrolysis of zinc fingers, this study predicts that the lowest energy complex (C) is initially protonated to give the parent complex (B). Further protonation on imidazole results in a partially aquated form (L) and then finally the fully aquated Zn(II) ion (M). A species has been observed in NMR studies that is consistent with either complex A or the partially aquated form (L). These computations strongly suggest that the partially aquated form (L) is the observed species and a reasonable intermediate at low pH in the hydrolysis of zinc fingers.
The observed rearrangement of the triply deprotonated complexes was not unexpected since we have observed this phenomenon previously in similar model systems (e.g., Zn(II)(SH-)2(OH-)*(H2S) ) [19]. Due to the relatively high energy of these complexes, these rearranged forms are probably only computational curiosities.
We would like to thank the Pittsburgh Supercomputing Center for time on the Cray C916/16512 and the North Carolina Supercomputing Center for time on the Cray T90. This work was supported in part by a grant from the Pittsburgh Supercomputing Center through the NIH - National Center for Research Resources cooperative agreement P41 RR06009 and through individual NIH grants HL-27995 (LGP) and HL-26309 (LGP). LGP especially thanks NIEHS for providing summer support during the course of this work.