Interrogating Key Positions of Size-Reduced TALE Repeats Reveals a Programmable Sensor of 5-Carboxylcytosine

Abstract

Transcription-activator-like effector (TALE) proteins consist of concatenated repeats that recognize consecutive canonical nucleobases of DNA via the major groove in a programmable fashion. Since this groove displays unique chemical information for the four human epigenetic cytosine nucleobases, TALE repeats with epigenetic selectivity can be engineered, with potential to establish receptors for the programmable decoding of all human nucleobases. TALE repeats recognize nucleobases via key amino acids in a structurally conserved loop whose backbone is positioned very close to the cytosine 5-carbon. This complicates the engineering of selectivities for large 5-substituents. To interrogate a more promising structural space, we engineered size-reduced repeat loops, performed saturation mutagenesis of key positions, and screened a total of 200 repeat-nucleobase interactions for new selectivities. This provided insight into the structural requirements of TALE repeats for affinity and selectivity, revealed repeats with improved or relaxed selectivity, and resulted in the first selective sensor of 5-carboxylcytosine.

Recognition of Oxidized 5‐Methylcytosine Derivatives in DNA by Natural and Engineered Protein Scaffolds

“…Repeats G* (as well as S* and T*) bound to all five nucleobases with similar affinities and thus were revealed as “universal” repeats. Modelling studies suggested that this is due to a strongly reduced RVD loop size and an associated complete removal of interactions with the five substituents (Figure E shows a model in comparison to RVD NG) . Interestingly, a fine‐tuning of sterics led to a new selectivity: RVD P* with a likely more rigid loop and protruding pyrrolidin ring turned out to be a selective negative sensor of the most bulky nucleobase caC.…”

“…Moreover, size‐reduced repeats are rare in natural TALEs and typically feature only a single deletion in the RVD . In a later study, TALE mutant libraries were constructed by saturation mutagenesis at position 12 of a single repeat combined with single or double deletions at positions 13 and 14 and screened for binding to the five cytosine nucleobases …”

“…Interestingly, a fine‐tuning of sterics led to a new selectivity: RVD P* with a likely more rigid loop and protruding pyrrolidin ring turned out to be a selective negative sensor of the most bulky nucleobase caC. Models suggested that this loop clashed with caC but not with the other four nucleobases (Figure F) . Though not conducted in the context of all five cytosine nucleobases, recent screens of alternative mutant libraries also afforded interesting selectivities.…”

“…[114] In a later study, TALE mutant libraries were constructed by saturation mutagenesis at position 12 of a single repeat combined with single or double deletions at positions 13 and 14 and screened for binding to the five cytosine nucleobases. [126] Whereas mutants with two deletions generally exhibited low binding, several mutants with single deletions exhibited interesting selectivity profiles. Repeats G* (as well as S* and T*) bound to all five nucleobases with similar affinities and thus were revealed as "universal" repeats.…”

Harnessing natural DNA modifying activities for editing of the genome and epigenome

“…The discovery of transcription activator-like effector (TALE) proteins from plant pathogens later offered a means for targeting single bases, and, like with ZFPs, concatenation of TALEs results in selective recognition of longer DNA motifs [14]. Structure-function insights into the TALE residues that contact nucleobases have permitted the development of constructs that differentiate unmodified cytosine from 5mC and ox-mCs [15–18]. Such modified TMs (TM*) have not yet been employed with different MMs, but the combination would make for a logical and anticipated advance.…”