MECHANISTIC STUDIES OF THE METHYLTHIOLATION REACTION CATALYZED BY THE RADICAL SAM ENZYME RIMO
Open Access
- Author:
- Landgraf, Bradley James
- Graduate Program:
- Chemistry
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 05, 2016
- Committee Members:
- Squire J. Booker, Dissertation Advisor/Co-Advisor
Squire J Booker, Committee Chair/Co-Chair
Carsten Krebs, Committee Member
Amie Kathleen Boal, Committee Member
William O Hancock, Outside Member - Keywords:
- S-adenosylmethionine
Radical SAM
Methylthiotransferase
Enzymology
5'-deoxyadenosyl radical
S-adenosylhomocysteine
Post-translational modification - Abstract:
- The S12 protein, a component of the bacterial 30S subunit of the ribosome, contains a universally conserved aspartic acid at position 89 (D89) in Escherichia coli (Ec). D89 is the target of a unique post-translational modification (PTM), methylthiolation (–SCH3), at its C3 position to form 3-methylthioaspartyl 89. This reaction is chemically challenging, requiring the activation of an unactivated sp3-hybridized carbon center for insertion of the -SCH3 group. The enzyme responsible for catalyzing this PTM is RimO (ribosomal modification O), a member of the superfamily of enzymes named radical SAM (RS). RS enzymes reductively cleave S-adenosyl-L-methonine (SAM) upon its binding to a reduced [4Fe-4S] cluster to generate methionine and a 5'-deoxyadenosyl 5'-radical (5'-dA•), a potent oxidant. This radical is used to abstract one of the prochiral hydrogen atoms from C3 of D89, thereby activating it for sulfur- or methylthio-insertion. The protein ligates an additional [4Fe-4S] cluster, known as an auxiliary cluster, in its N-terminal region. This auxiliary cluster is thought to be a sacrificial source of sulfide for the methylthiolation reaction in vitro. Radical recombination between the substrate and a µ3-sulfido ion of the auxiliary cluster would result in a thiolated intermediate of D89. In addition to its use of SAM as a precursor to a 5'-dA•, RimO also catalyzes the transfer of a methyl group from a second molecule of SAM, presumably to the sulfur atom of the thiolated intermediate of D89, or to an acceptor site on the RimO polypeptide that acts as an intermediary that then transfers the methyl group to the inserted sulfur atom and completes the reaction. Biochemical experiments described in chapter 2 determined that RimO catalyzes methyl transfer from SAM to an acceptor site on itself in the absence of a chemical reductant or an S12 peptide substrate, indicating that methyl transfer precedes radical chemistry. Radiotracing studies in which RimO was incubated with [14C-methyl]-SAM and the mixture subsequently separated by size-exclusion chromatography demonstrated that radioactivity was associated with the protein. HPLC analysis of the protein fraction using an acidic mobile phase resulted in the complete loss of radioactivity from SAM-derived breakdown products, which suggested that the radioactivity was liberated upon treatment with acid. GC-MS analysis of headspace injections taken from sealed vials containing RimO incubated with SAM or [methyl-d3]-SAM showed production of methanethiol or d3-methanethiol, respectively, indicating that the acid-labile auxiliary [4Fe-4S] cluster was the methyl acceptor site. This methylated cluster intermediate was shown to be chemically and kinetically competent, and the presence of methanethiol in reaction mixtures of RimO resulted in its incorporation in the S12 peptide substrate and also enabled the enzyme to catalyze ~ 3 turnovers. RimO from the gut bacterium Bacteroides thetaiotaomicron (Bt) was characterized and shown to be similar to RimO from T. maritima in chapter 3. One of the key differences of Bt RimO was the fact that the flavodoxin/flavodoxin reductase/NADPH (Fld/Fdx/NADPH) reducing system from Ec was a competent source of electrons required for catalysis, thereby obviating the use of the chemical reductant sodium dithionite. Use of the Fld/Fdx/NADPH reducing system decreased the amount of 5'-deoxyadenosine (5'-dAH) formed abortively—meaning uncoupled from methylthiolated product formation—but did not eliminate it. The flavodoxin semiquinone was used as a spectroscopic handle to estimate that Bt RimO uses ~ 1 electron for each methylthiolated product formed. It was determined that Bt RimO does not harbor any additional sulfide or persulfide species through the use and quantification of the fluorescent sulfur-labeling reagent, I-AEDANS. In chapter 4, chemoenzymatic synthesis of 3-pro-R and 3-pro-S deuterium-labeled aspartic acid was achieved by exploiting the enzymatic reaction catalyzed by aspartate ammonia-lyase. The resulting product identities and retention of the deuterium label were confirmed by 1H NMR after orthogonal protecting groups appropriate for solid phase peptide synthesis (SPSS) were added. The labeled and protected aspartic compounds were incorporated into synthetic peptides corresponding to residues 83-95 of the S12 protein by SPSS, and confirmation of the correct peptide and retention of the deuterium label was obtained by MALDI-TOF MS. When RimO from Bt and Tm were incubated under turnover conditions with the pro-R or pro-S labeled S12 peptide substrates, observation of deuterium incorporation into 5'-deoxyadenosine occurred only with the peptide substrate containing deuterium at the pro-S position, indicating that the enzyme stereoselectively abstracts the pro-S hydrogen atom from its target substrate. This finding also established the stereochemical course of methylthiolation to occur with inversion of configuration, and the apparent primary kinetic isotope effect for H-atom versus D-atom abstraction of ~ 1.9 indicates that this step is at least partially rate-limiting. Additionally, a large apparent secondary isotope effect of ~1.4 was observed with the pro-R labeled substrate. In chapter 5, the Tm S12 protein was overproduced in Ec, purified from inclusion bodies under denaturing conditions, and slowly refolded to yield homogenous protein after size-exclusion chromatography. S12 was shown to be a competent substrate when incubated with Tm RimO under turnover conditions, with incorporation of an -SCH3 group into S12 observed by MALDI-TOF MS. Variant proteins of Tm RimO in which one conserved amino acid found in the protein active site was substituted by site-directed mutagenesis were characterized; the specific Tm RimO variant were as follows: K12A, K12Q, Y227A, Y227F, and Q192A. Substitution of K12 with alanine or glutamine abolished 5'-dAH and methylthiolated product formation and decreased both the amount and rate of SAH formation, but did not affect the ability of the K12A variant to bind SAM. These results suggested that the lysine residue may play a minor role in methyl transfer, but is required in some unknown capacity for generation of the 5'-dA•. Both Y227A and Y227F Tm RimO variants catalyzed methyl transfer and formation of 5'-dAH, but in neither reaction was the methylthiolated product observed. Reactions containing either the Y227A or Y227F variant in ~60% D2O resulted in no deuterium enrichment into 5'-deoxyadenosine, suggesting that the 5'-dA• abstracts a hydrogen atom from a site on the S12 peptide or the RimO protein that is not solvent exchangeable. Substitution of F for Y at position 227 had little effect on the determined dissociation constant for SAM binding compared to the wild-type enzyme as determined by ITC. The Q192A variant was capable of catalyzing the full methylthiolation reaction, albeit it to lower extents and at slower rates compared to the wild-type enzyme, making the role for this conserved residue nebulous.