…

The significance of BpyAla-mediated nucleic acid cleavage becomes clearer when placed in the broader context of how natural nucleases operate. Cellular organisms rely on metal-dependent catalysis to process genetic material, with enzymes such as restriction endonucleases, ribonucleases, and DNA repair proteins all exploiting divalent cations to activate water molecules or stabilize transition states during phosphodiester bond hydrolysis. The two-metal-ion mechanism, in particular, has been characterized extensively in polymerases and nucleases, where one metal ion positions the substrate and a second activates the nucleophile. What the BpyAla approach achieves is a deliberate simplification of this architecture: rather than assembling an entire catalytic domain with multiple coordinating residues, a single genetically encoded residue bearing a bipyridine chelator can recruit a metal ion and generate a catalytic center at a chosen position. This reductionist strategy means that the catalytic power of a metallonuclease can be grafted onto proteins that already possess useful binding properties but lack any cleavage activity, which is precisely what was demonstrated with the CAP and p19 systems described in the source work.

The choice of protein scaffolds in the proof-of-concept studies is instructive for understanding the design logic. CAP is a well-characterized bacterial transcription factor whose DNA-binding specificity has been mapped in detail, and its dimeric structure positions residues near the major groove of its consensus sequence. By substituting a lysine at position 26 with BpyAla, researchers placed the metal-chelating side chain within reach of the phosphodiester backbone, allowing copper- or iron-dependent cleavage to occur selectively at the bound site. Similarly, p19 is a viral protein that binds double-stranded RNA duplexes with size-dependent affinity, sequestering small interfering RNAs and microRNAs as part of a plant viral immune evasion strategy. Converting p19 into a cleavage-active construct by installing BpyAla at position 111 meant that the protein could not only bind but also destroy its RNA targets, including both siRNA and miRNA species. In both cases, the underlying binding protein dictated target selection while the engineered metallocenter supplied the chemical reactivity, illustrating the modular division of labor that makes the approach programmable.

From a mechanistic standpoint, the metal ion coordinated by the bipyridine moiety is thought to function as a Lewis acid, polarizing the phosphodiester bond and, in many proposed mechanisms, activating a water molecule or hydroxide for nucleophilic attack. The precise pathway may vary depending on the metal employed, the geometry of the coordination sphere, and the local protein environment. Copper and iron, the cofactors used in the early CAP experiments, are redox-active metals capable of participating in oxidative cleavage chemistry as well as hydrolytic mechanisms, which raises interesting questions about the nature of the cleavage products and the reversibility of the reaction. Hydrolytic cleavage of RNA is generally more facile than DNA hydrolysis because the 2′-hydroxyl group of ribose can act as an internal nucleophile, which may partly explain why RNA-targeting systems like the engineered p19 achieved detectable cleavage activity relatively readily. DNA cleavage, lacking this internal nucleophile, typically demands more aggressive catalytic strategies, whether through metal-hydroxide mechanisms or through oxidative pathways that generate strand scission via radical intermediates.

The proposal to develop next-generation BpyAla analogues deserves particular attention because the bipyridine scaffold offers numerous points of chemical modification. Substituents at the 4- and 5-positions of the pyridine rings can tune the electron density at the coordinating nitrogens, which in turn affects metal-binding affinity, the redox potential of the resulting complex, and the pKa of any metal-bound water ligands. Electron-withdrawing groups would be expected to stabilize lower oxidation states and potentially shift the optimal metal identity, while electron-donating substituents could strengthen coordination to certain divalent cations. Beyond simple substitution, replacing the bipyridine with phenanthroline would pre-organize the two nitrogen donors into a more rigid, coplanar arrangement, potentially increasing binding affinity and enforcing a defined geometry around the metal center. Terpyridyl analogues would add a third coordinating nitrogen, enabling octahedral coordination environments when combined with additional ligands from the protein or solvent, and opening the door to metals with higher coordination numbers or to bimetallic assemblies. Each of these structural variations carries implications for catalytic mechanism, metal selectivity, and compatibility with the folding constraints of the host protein.

The exploration of alternative metal cofactors is another dimension with substantial untapped potential. Zinc, the most common metal in natural nuclease active sites, is redox-inert and operates through purely Lewis-acid mechanisms, which could offer cleaner hydrolytic chemistry and reduced off-target oxidative damage to biomolecules. Magnesium and calcium, abundant in biological systems, are harder Lewis acids that prefer oxygen donors, and their compatibility with a nitrogen-donor chelator like bipyridine is questionable, suggesting that analogues with mixed donor sets might be needed to exploit these biologically benign metals. Manganese, cobalt, and nickel occupy intermediate territory, with varying preferences for coordination geometry and redox behavior. Lanthanide ions, with their large radii and high coordination numbers, are exceptionally effective at RNA cleavage and could inspire chelator designs that accommodate them. The ability to swap metal cofactors within the same protein scaffold, simply by changing the metal supplied in solution, would constitute a form of post-translational tunability that no canonical amino acid substitution could achieve.

Cooperative and multi-residue design strategies represent a natural extension of the single-BpyAla paradigm. Natural metallonucleases frequently employ two or more metal ions in their active sites, with the second metal assisting in leaving-group stabilization or nucleophile activation. Installing two BpyAla residues at carefully chosen positions within a protein could generate a dimetallic center, provided the spacing and orientation of the two residues permit simultaneous metal coordination. The genetic code expansion toolkit makes this feasible, since multiple amber codons can be decoded by the same engineered tRNA-synthetase pair, though efficiency considerations and the need for both residues to be incorporated correctly add practical complexity. Alternatively, BpyAla could be combined with canonical residues that contribute secondary coordination sphere interactions, such as arginine or lysine residues that position the phosphate, or histidines that serve as additional metal ligands. Such hybrid active sites, blending genetically encoded unnatural chemistry with conventional catalytic residues, mirror the strategies that directed evolution has refined in artificial metalloenzyme research more broadly.

The question of delivery and biological context looms over any discussion of therapeutic applications. For BpyAla-engineered systems to function inside cells or in vivo, the host organism must possess the machinery for genetic code expansion, which in practice means either engineered bacterial strains, transfected mammalian cells carrying the orthogonal tRNA-synthetase pair, or cell-free expression systems. Cell-free protein synthesis offers a particularly attractive middle ground, allowing production of BpyAla-containing proteins under controlled conditions before deployment, and several studies have already demonstrated efficient UAA incorporation in cell-free platforms. The metal cofactor must also be supplied and maintained in an oxidation state compatible with catalysis, and the intracellular milieu presents competing chelators, reducing agents, and abundant nucleic acids that could sequester the metal or challenge target specificity. These constraints do not negate the therapeutic vision but they shape the realistic near-term applications, which are more likely to involve ex vivo protein production, research reagents for targeted nucleic acid degradation, and components of synthetic biology circuits rather than immediate clinical deployment.

Specificity considerations also merit careful thought. A DNA-binding protein converted into a metallonuclease inherits the sequence specificity of its parent scaffold, but the cleavage event itself is mediated by a small metal center whose reach is limited to the immediate vicinity of the BpyAla residue. This means cleavage sites are determined by the geometry of the protein-nucleic acid complex, and off-target cleavage could occur at sites where the protein binds adventitiously. Natural restriction enzymes solve this problem through extensive substrate contacts and conformational proofreading, and engineered systems may need analogous safeguards, whether through multivalent binding, allosteric activation, or pairing with guide nucleic acids that confer programmable targeting in the manner of CRISPR systems. The conceptual overlap with CRISPR-associated nucleases is worth noting: both approaches aim to direct cleavage activity to chosen sequences, but the BpyAla strategy achieves this through protein engineering and genetic code expansion rather than through RNA-guided colocalization, and the two paradigms could prove complementary rather than competing.

The expansion of compatible protein scaffolds and nucleic acid substrates rounds out the forward-looking agenda. Beyond DNA-binding transcription factors and RNA-sequestering viral proteins, candidate scaffolds include zinc-finger domains, which offer modular DNA recognition; pentatricopeptide repeat proteins, which bind single-stranded RNA with programmable specificity; and antibody fragments selected against nucleic acid targets. Each scaffold class presents distinct geometric constraints that influence where BpyAla can be productively installed, and computational tools for predicting protein-nucleic acid interfaces will increasingly guide site selection. On the substrate side, extending cleavage activity to structured RNAs, modified nucleotides, or even chemically modified synthetic oligonucleotides would broaden the utility of these systems in research and therapeutic contexts. The convergence of an expanding UAA repertoire, refined chelator chemistry, diverse metal cofactors, and ever-better structural prediction tools positions BpyAla-engineered metallonucleases as a genuinely programmable platform, one whose full scope will be defined by the design choices and experimental creativity of the coming years.

Subject of Research: Towards programmable metallonucleases via genetic code expansion

Article Title: Towards programmable metallonucleases via genetic code expansion

Article References: Lundrigan, E., O’Neill, M., & Pezacki, J. P. (2026). Towards programmable metallonucleases via genetic code expansion. Frontiers in Chemical Biology, 5, Article 1914162. https://doi.org/10.3389/fchbi.2026.1914162

Image Credits: AI Generated

DOI: 10.3389/fchbi.2026.1914162

Keywords: Towards, programmable, metallonucleases, genetic, code, expansion, scientific research