Protein switches use the binding energy gained upon recognition of ligands to modulate the conformation and binding properties of protein segments. We explored whether the programmable nucleic acid mediated recognition might be used to design or mimic constraints that limit the conformational freedom of peptide segments. The aim was to design nucleic acid–peptide con-jugates in which the peptide portion of the conjugate would change the affinity for a protein target upon hybridization. This approach was used to control the affinity of a PNA–phosphopeptide conjugate for the signal transduction protein Src kinase, which binds the cognate phosphopeptides in a linear conformation. Peptide–nucleic acid arms were attached to known peptide binders. The chimeric molecules were studied in three modes: 1) as single strands, 2) constrained by intermolecular hybridization (duplex formation) and 3) constrained by intramolecular hybridization (hairpin formation). Of note, duplexes that were designed to accommodate bulged peptide structures (for example, in hairpins or bulges) had lower binding affinities than duplexes in which the peptide was allowed to adopt a more relaxed conformation. Greater than 90-fold differences in binding affinities were observed. It was, thus, feasible to make use of DNA hybridization to reversibly switch from no to almost complete inhibition of Src-SH2–peptide binding, and vice versa. A series of DNA and PNA-based hybridization experiments revealed the importance of charges and conformational effects. Nucleic acid mediated switching was extended to the use of RNA; this enabled a regulation of the enzymatic activity of the Src kinase. The proof-of-principle results demonstrate for the first time that PNA–peptide chimeras can transduce changes of the concentration of a given RNA molecule to changes of the activity of a signal transduction enzyme.