Design of single-stranded nucleic acid binding peptides based on nucleocapsid CCHC-box zinc-binding domains

Abstract

The solution structures of nucleocapsid (NC)-like CCHC zinc-binding domains bound to nucleic acid targets have revealed that these domains bind guanosine residues within single-stranded nucleic acids. Here, we have performed initial studies examining the potential use of NC-like CCHC zinc-binding domains as modules to construct single-stranded nucleic acid binding peptides. The affinity for guanosine-containing single-stranded deoxyribooligonucleotides increases with the number of CCHC domains in the peptide. The length of the linker between domains affects the spacing of guanosine residues in oligonucleotides that are preferentially bound. These studies provide a proof of principle that NC-like CCHC zinc-binding domains can be utilized as a basis for designing peptides that bind specific single-stranded nucleic acid sequences.

Figure 1

A representation of the structure typically formed by CCHC-box domains. The structure is centered around a zinc ion coordinated to the cysteinate sulfur atom of three cysteinate residues and the ε-nitrogen from a histidine. Three peptide backbone NH to S hydrogen bonds (shown in as dotted lines) stabilize the structure.

Figure 2

The structure of the HIV-1 nucleocapsid protein bound to a 20-base RNA target. Each of the two CCHC-box domains interacts primarily with a guanosine with the guanine base lying in a pocket in the CCHC-box domain, making hydrogen bonds (shown as dotted lines) with the CCHC-box domain backbone. Note that the CCHC-box domains have been reoriented for clarity.

Figure 3

Plots of the changes in fluorescence anisotropy associated with oligonucleotide binding by peptides with one (A), two (B), or three (3) CCHC-box domains joined by linkers derived from UMSBP. Data are shown for oligonucleotides containing appropriately spaced guanosine residue (red) and the corresponding oligonucleotide containing only thymidine (blue) with fit curves in black.

Figure 4

Binding studies with the peptide with two CCHC-box domains joined by the linker from the HIV-nucleocapsid protein with d(TTTGTTGTTT)-F. (A) The increase in fluorescence anisotropy (red) and the decrease in fluorescence intensity. (B) The binding curve fit to the fraction bound determined by the changes in anisotropy corrected for the changes in fluorescence intensity.

Figure 5

The effects of differences in spacing between guanosine residues on the binding of (A) UBPCON2U and (B) UBPCON2N determined by competition experiments. UBPCON2U preferentially bound the oligonucleotide with the guanosines separated by two thymidines whereas UBPCON2N preferentially bound the oligonucleotide with the guanosines separated by a single thymidine.