Ribonucleotide reductases

Ribonucleotide reductase (RNR) enzymes are required in all organisms for DNA replication and repair. This central function offers important, but under-utilized, potential as a target for therapeutic intervention, including development of new antibiotics. We have discovered and structurally characterized two new bacterial class I RNR subclasses, termed Id and Ie. Class Id enzymes use a stable high-valent Mn₂ cluster activated directly by superoxide. Class Ie enzymes completely lack transition metals in their activated form. The Ie enzymes require a separate protein activase to convert a conserved Tyr to a post-translationally modified semiquinone radical (DOPA•). Our work highlights the predominance of Mn-based RNR activity in microbes, thought to be important in escape of iron-dependent nutritional immunity. Class Ie enzymes – found in Streptococcus and Mycoplasma pathogens linked to scarlet fever, pneumonia, and flesh-eating bacterial infections – may represent the most extreme RNR-based adaptation to metal-targeting innate immunity of the host.  Future directions include understanding the structure and biosynthesis of these and other novel class I RNR catalytic cofactors.

Fe/2-OG-dependent oxygenases

We also aim to understand the structural features that dictate reaction outcome in Fe(II)/2-oxo-glutarate (Fe/2-OG) dependent oxygenases. These enzymes constitute a large and diverse family of catalysts, all members of which are believed to utilize a common Fe(IV)=O intermediate to enable a variety of transformations initiated by H-atom abstraction, typically from chemically inert sites. Most of these enzymes are hydroxylases, many with biomedical relevance. A number of other outcomes are possible including halogenation, stereoinversion, cyclization, and desaturation. The structural basis for a given reaction outcome is not fully understood. Our group has solved the first x-ray structures of substrate-bound Fe/2-OG enzymes responsible for stereoinversion, halogenation, and desaturation reactions. We now seek to elucidate the structural underpinnings of other alternative outcomes and develop methods to capture structures of reaction intermediates. We use the resulting information to engineer new reactivities into these scaffolds.

Iron-sulfur enzymes

Radical SAM (RS) enzymes share a common functional motif, a [4Fe4S] cluster that coordinates SAM, to generate (typically) a 5’-deoxyadenosine radical (5’-dA•). The 5’-dA• moiety in RS proteins initiates reactions with substrates by abstracting an H-atom, similar to Fe/2OG enzymes. We study radical SAM methylases that target sp² carbons in the nucleobases of large structured RNAs, systems that additionally face special challenges in substrate recognition and outcome control.

Nearly half of the antibiotics in current clinical use target the bacterial ribosome. A mobile genetic element, cfr, encodes a rapidly emerging mechanism of enzymatic antibiotic resistance that manifests via methylation of an inert carbon atom in a key adenine nucleotide in the large subunit of the ribosome. Cfr and a related radical SAM methyltransferase, RlmN, employ a distinctive chemical mechanism. This pathway involves (1) methyl modification of a conserved cysteine (Cys) residue, (2) 5’-dAŸ-mediated activation of the Cys-appended methyl group, (3) addition of the resulting methylene radical to C2 of A2503 to form a transient protein-RNA crosslink, and (4) resolution of the adduct via proton abstraction and thiyl radical formation. The last step is facilitated by additional conserved Cys and Met residues. In collaboration with Squire Booker’s laboratory, we have solved structures of a tRNA-bound crosslinked RlmN intermediate, the first view of a radical SAM enzyme in complex with a large macromolecular substrate.  Current work focuses on structure determination of Cfr enzymes and other iron-sulfur enzymes implicated in C-C bond formation reactions.

Heme-oxygenase-like diiron oxidases/oxygenases (HDOs)

Recently we have defined a new, ~10,000 member structural superfamily emerging from the large and well-characterized ferritin-diiron oxygenase (FDO) family. The HDOs derive their name from structural resemblance to heme oxygenase, and conserve three distorted core α-helices that bind the metallocofactor. These distorted helices render the HDO cofactor unstable, preventing straightforward interrogation of structure and mechanism. We have developed methods to investigate properties of the labile HDO cofactor and recently solved the structure of the first HDO with its cofactor intact. To date, four HDO enzymes have been well characterized experimentally and reveal two distinct reactivities: SznF and RohS function as N-oxygenases, and UndA and BesC function as desaturases. Characterized HDO enzymes function in natural product biosynthetic pathways to a chemotherapeutic (Streptozotocin, SznF), an antibiotic (azomycin, RohS), and other natural products with relevance to medicine and biotechnology.

Current projects in the lab seek to further our understanding of the largely uncharacterized HDO family using methods that include X-ray crystallography and spectroscopy. We are also exploring novel HDOs that function in bacteriocin biosynthetic pathways to expand the scope of known HDO-catalyzed reactions.