Structural basis for diverse metallocofactor usage in class I ribonucleotide reductases from bacterial pathogens
Open Access
- Author:
- Palowitch, Gavin
- Graduate Program:
- Biochemistry, Microbiology, and Molecular Biology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- December 17, 2020
- Committee Members:
- Amie Kathleen Boal, Dissertation Advisor/Co-Advisor
Amie Kathleen Boal, Committee Chair/Co-Chair
Joseph M Bollinger, Jr., Committee Member
Carsten Krebs, Committee Member
Manuel Llinas, Outside Member
Squire J Booker, Committee Member
Wendy Hanna-Rose, Program Head/Chair - Keywords:
- ribonucleotide reductase
RNR
metallocofactor
bacterial pathogens
biochemistry
bioinorganic chemistry
class I
iron
manganese
tyrosyl radical
radical chemistry
PCET - Abstract:
- Ribonucleotide reductases (RNR) are essential enzymes that catalyze the first committed step in the biosynthesis of 2’-deoxyribonucleotides (dNTP). RNRs catalyze the two-electron reduction of ribonucleotides to dNTPs through a conserved radical-mediated mechanism. Class I RNRs represent the primary source of de novo dNTPs in aerobic organisms. Class I RNRs operate as a multi-subunit (generally alpha2beta2) holoenzyme complex in which substrate reduction is initiated by formation of a cysteinyl radical (Cys•) in the catalytic alpha subunit of the enzyme. The beta subunit of the holoenzyme assembles, stabilizes, and reversibly deploys the Cys oxidant, either an oxidized transition metal complex or a tyrosine (or tyrosine-derived) radical (Tyr•). The transient generation of the Cys• in alpha is achieved by a long-range, intersubunit, radical-translocation (RT) process mediated by a chain of conserved aromatic residues that spans the intersubunit interface. RT is bidirectional, occurring in the first and last step of each turnover to generate and quench the 3'-H-cleaving Cys•. Class I RNRs can further be subdivided based on the identity of the alpha-Cys oxidant. Class I RNRs are further subdivided into subclasses a-e on the basis of four distinct features of the cofactor in the beta subunit and its assembly pathway. These include (i) the number and type of metal ions required in the beta subunit for radical or metal cofactor generation, (ii) the oxidant used to generate the active cofactor (O2 or superoxide), (iii) the requirement for an additional activase protein, and (iv) the identity of the stable oxidant that forms the Cys• in alpha. Class Ia RNRs, which are found in all eukaryotic organisms and many prokaryotes, use an Fe2(III/III)-Y• cofactor to initiate nucleotide reduction. Class Ib RNRs, which are the primary source of deoxynucleotides in many bacterial pathogens, are Mn-dependent and assemble the active Mn2(III/III)-Tyr• state via a mechanism analogous to that of the class Ia enzymes, yet require a flavoprotein activase, NrdI, for cofactor assembly. Class Ic RNRs use an Mn(IV)Fe(III) cofactor to initiate nucleotide reduction. Class Ic RNRs utilize the oxidized metal cluster itself in place of the Tyr• for Cys• formation. Class Id RNRs use an oxidized Mn(III)Mn(IV) cluster to initiate radical translocation, similar to the Ic system. Class Ie RNRs use a Tyr-derived 3,4-dihydroxyphenylalaninyl radical (DOPA•) to initiate radical translocation and are the first metal-free example of the enzyme. The recent discovery of novel cofactors among class I RNRs highlights the diverse strategies that organisms use to replicate their DNA in specialized environmental niches, perhaps forming a link to pathogenic virulence. The goal of the research described in this thesis is to understand new ways that nature has adapted microbial class I RNRs, while also elucidating the chemistry behind the assembly processes of these novel cofactors. In Chapter 1, the consensus mechanism of ribonucleotide reduction is discussed and the details of metallocofactor assembly in subclasses a-e are detailed. The chapter emphasizes the structural features of the beta subunits that enable new mechanisms of cofactor assembly and use of novel oxidants. The historical context of subclass discovery is presented and techniques and inhibitors used in previous studies are discussed. Finally, we predict the identity of novel cofactors in uncharacterized class I RNR beta sequences based on bioinformatics approaches, genomic context, and biological rationales. In Chapter 2, the discovery of the class Ie RNR is detailed. The goal of this research was to identify the mechanism by which a subset of unique beta sequences were able to perform successful ribonucleotide reduction, as shown in a previous study. These sequences, natively encoded in pathogenic and commensal bacteria, lacked three conserved metal-binding Glu residues from subclasses a-d. Through the use of x-ray crystallography, mass spectrometry, and advanced electron paramagnetic resonance (EPR) methods, the beta subunits of this subclass were determined to harbor a Tyr-derived 3,4-dihydroxyphenylalaninyl radical (DOPA•) that is generated in a NrdI- and O2-dependent reaction. This DOPA•-containing form of the protein was active in the absence of bound transition metals. The cofactor assembly reaction was able to be reproducibly achieved via co-expression of beta and NrdI in E. coli, but the reaction was not able to be replicated in vitro. The DOPA• was confirmed to be the alpha-Cys oxidant via a mechanism-based inhibition experiment using 2’-azido-2’deoxycytidine 5’-diphosphate (N3-CDP). While the mechanism of DOPA• cofactor assembly is still unknown, class Ie RNRs are proposed to represent the most potent countermeasure against transition-metal-directed innate immunity. In Chapter 3, more recent studies probing the class Ie RNR DOPA• cofactor assembly reaction are discussed. Although it appeared in first reports to have escaped the usual requirement for transition metals, we show here that maintenance of the DOPA• in the Ie enzyme by the flavoprotein activase, NrdI, is inhibited by inclusion of neocuproine, a Cu(I)-specific cell-permeant chelator. X-ray crystallographic characterization of the Ie activation complex reveals a binding site for Cu(I) and a second interaction surface on NrdI involving the C-terminal tail of the Ie RNR beta subunit, implicated as essential in RT during turnover. Cu(I) binding in the Ie activation complex also perturbs the reactivity of NrdI, leading to a more biased stabilization of the anionic form of the semiquinone state (NrdIsq) of the flavoprotein. Mass spectrometry analysis of in vitro activation attempts revealed some modification of the Tyr to DOPA (two-electron oxidation) as a result of O2 exposure to the activation complex in the fully reduced form (beta•NrdIhq). The full transformation from Tyr to the active DOPA• has yet to be achieved in vitro. The use of a multi-step mechanism of cofactor assembly is preferred by the majority of experimental evidence. Copper may play a role in one or several steps of the process, or may only be required for maintenance of the DOPA• in vivo. These findings further expand the known biological inorganic chemistry represented among these ubiquitous and essential enzymes, providing new opportunities for their inhibition in pathogens. In Chapter 4, the biochemical characterization of the sole class I RNR from the bacterial pathogen Francisella tularensis (Ft) is detailed. Ft RNR is an iron-dependent enzyme that stabilizes a class Ia-like Fe2(III/III)-Tyr• cofactor. Through the use of N3-CDP, the Tyr• does not undergo loss of spin when forward PCET is initiated. Thus, Ft beta does not use the Tyr• catalytically, allowing us to designate Ft RNR as the first member of a new subclass, termed Ig. The resting oxidized form of Ft beta is a class Ia-like, but we propose that interaction with the alpha subunit induces conformational changes that release the site 1 Fe, leaving behind a mononuclear form that is observed in crystal structures of the protein with Fe bound. A g = 4.3 EPR signal, indicative of a mononuclear Fe(III) ion, is perturbed upon incubation with alpha, which is not observed in other subclasses. Ft beta appears to initiate Cys oxidation in alpha with an Fe(III)-derived cofactor by an unknown mechanism. Ft RNR may therefore represent a new way that pathogens have evolved to diminish dependence on intracellular iron. In Chapter 5, biochemical characterization of two class Id RNRs is presented. Class Id RNRs use a simple mechanism of cofactor activation involving oxidation of a Mn(II)2 cluster by free superoxide to yield a metal-based Mn(III)Mn(IV) oxidant. This cofactor assembly pathway suggests that class Id RNRs may be representative of the evolutionary precursors to more complex class Ia-c enzymes. Identification of a second class Id RNR system, from an opportunistic human pathogen, confirms that class Id RNRs are activated in a similar manner, but that the exact electronic environments of their active cofactors vary significantly. X-ray crystal structures of the class Id catalytic alpha subunit reveal that the enzyme is distinctly small, lacking common N-terminal ATP cone allosteric motifs that regulate overall activity in class Ia-c RNRs. The findings suggest an evolutionary relationship between bacterial class Id and eukaryotic class Ia alpha subunits, and indicate as yet undefined mechanisms of overall activity regulation that may exist in simple RNRs with truncated or missing allosteric motifs. Chapter 6 discusses several recent discoveries in the ribonucleotide reductase field that have expanded upon the decades of work done to characterize the complex radical translocation process during turnover, discover new subclasses, and identify their active metallocofactors. Studies from the past few years are discussed and their key details are explored in context. Avenues for future research, both into characterization of the active holoenzyme complex and novel subclass discovery, are also detailed. Biochemically uncharacterized beta subunit sequences of interest are explored later in the chapter as promising new candidates for novel subclass discovery.