The Escherichia coli Radical SAM Enzyme YhcC Substitutes for the FAD-Dependent Oxidase Activity of MnmC in 5-Methylaminomethyl-2-Thiouridine tRNA Modification Under Anaerobic Conditions

Abstract

tRNA wobble base uridines are heavily modified to influence anticodon:codon base pairing and tune the structure of the anticodon stem loop for efficient and accurate translation. Both Gram-positive and negative bacteria, as well as certain archaea, modify wobble uridines to contain either a 5-carboxymethylaminomethyl (cmnm⁵) or a 5-methylaminomethyl (mnm⁵) moiety, as well as in some instances a 2-thio (s²) moiety. Bacteria utilize the conserved MnmEG complex to produce cmnm⁵U, which is further modified in some tRNAs to mnm⁵U. The steps to synthesize the latter are catalyzed by non-orthologous enzymes in distantly related bacteria. Escherichia coli utilizes a single bifunctional enzyme, MnmC, to both demodify cmnm⁵U to nm⁵U and subsequently methylate nm⁵U to mnm⁵U, while Bacillus subtilis relies on the radical SAM (rSAM) enzyme MnmL, followed by the stand-alone MnmM methylase for mnm⁵U production. It was previously noted that E. coli and related bacteria that encode MnmC can also contain homologs of MnmL. As an E. coli mnmC mutant accumulates cmnm⁵U, the function of the MnmL homolog in this organism, YhcC, was unknown. Here, we find YhcC is necessary for cmnm⁵s²U demodification in vivo under anaerobic growth, and that the same MnmC-mediated demodification activity requires O₂ and only occurs under aerobic growth. In vitro reaction experiments demonstrate that purified YhcC, reconstituted to its [4Fe-4S] form, is able to bind tRNA and catalyze the nm⁵s²U-tRNA synthesis from cmnm⁵s²U-tRNA. Together, these results define the heretofore unknown biochemistry of the E. coli rSAM enzyme YhcC and show this enzyme replaces MnmC under anaerobic conditions to carry out the synthesis of nm⁵s²U. These parallel tRNA modification pathways highlight how E. coli has adapted to maintain biosynthesis of a critical wobble base modification under both aerobic and anaerobic growth.

General audience summary

Transfer RNA (tRNA) molecules serve as critical components in protein synthesis through their direct interaction with the ribosome and messenger RNA (mRNA). During protein synthesis, tRNA molecules utilize an anticodon composed of three bases that recognize a complementary mRNA codon. At the 3'CCA end of tRNA the amino acid corresponding to the anticodon sequence is incorporated into the growing polypeptide chain. While canonical Watson-Crick base pairing, pairing between U:A and G:C, occurs between the second and third bases of the anticodon and the corresponding bases of the codon, there is increased flexibility in the ribosome at the first position of the anticodon. This results in non-canonical base pairing and allows a single tRNA molecule to recognize multiple codons. Modification at this site restricts or enhances this “wobble” base pairing. Bacteria often install mnm⁵s²U34 to various tRNAs to facilitate proper and efficient translation. In E. coli, three proteins are responsible for this hypermodification pathway: MnmE and MnmG convert s²U to cmnm⁵s²U, while MnmC subsequently removes the carboxymethyl group to generate nm⁵s²U and methylates this intermediate into the ultimate product, mnm⁵s²U. We found that an additional enzyme, YhcC, is required for mnm⁵s²U synthesis in anaerobic conditions, specifically at the cmnm⁵s²U demodification step. Previous studies showed that MnmC-catalyzed demodification is dependent on FAD for the initial oxidation of tRNA substrate, generating FADH₂. We propose that FADH₂ recycling is dependent on O₂ to regenerate FAD, allowing multiple catalytic turnovers. We show that, in vitro, O₂ is required for cmnm⁵s²U demodification by MnmC, supporting a new model of mnm⁵s²U synthesis with alternative aerobic and anaerobic routes. Conservation of multiple enzymes that perform similar chemistry, albeit under differing environmental conditions, highlights the importance of maintaining wobble base modifications as well as the ability of bacteria to adapt to their surroundings.

Introduction

Transfer RNAs (tRNAs) are essential for protein synthesis across all domains of life, bridging messenger RNA (mRNA) codons with their corresponding amino acids. To ensure proper folding, interaction with protein partners, and accurate decoding during translation, tRNAs undergo extensive post-transcriptional modification (1, 2). In particular, position 34, known as the wobble base, which pairs with the third nucleotide of the codon, is heavily modified with diverse chemical nucleobase modifications (3, 4). These wobble base modifications modulate decoding efficiency as well as accuracy by expanding or restricting codon-anticodon base pairing in the ribosome (5). This is especially important for decoding of two-codon family boxes, where the identity of the last base of the codon as either a pyrimidine or purine dictates the corresponding amino acid. Wobble base uridines (U34) specifically are nearly invariably modified across all organisms, in part due to the inherent base pairing flexibility of uridine (6). Most wobble uridine modifications are derivatives of 5-methyluridine (xm⁵) and also often contain a thio group at the C2 position (s²). In bacteria such as E. coli and B. subtilis, tRNAs that decode NNA/NNG codons typically contain a wobble uridine with either a 5-carboxymethylaminomethyl (cmnm⁵) or a 5-methylaminomethyl (mnm⁵) group, in addition to a 2-thio (s²) modification (7, 8). RNA modifications are important for various stress responses and adaptations. Recently, it was shown that mnmE mutants lacking cmnm⁵ exhibit a strong growth defect under acid stress as well as decreased cell motility (9). The importance of wobble uridine modifications is further highlighted by synthetic lethality in both E. coli and S. enterica LT2 when both mnmE and mnmA are deleted, resulting in completely unmodified U34 in various tRNAs. (10, 11).

The biosynthesis of (c)mnm⁵(s²)U34 modifications in bacteria begins with the MnmEG complex, which utilizes methylenetetrahydrofolate and either glycine (‘glycine route’) or ammonium (‘ammonium route’) to install a cmnm⁵ or nm⁵ (5-aminomethyluridine) moiety, respectively (12, 13) (Fig. 1). While MnmEG activity with ammonium was first shown in vitro, it was also found that MnmEG seems to prefer this activity over the glycine route in stationary phase in vivo (13, 14). In contrast, the glycine route was preferred in exponential phase (14). In E. coli, a single bifunctional enzyme, MnmC, then converts cmnm⁵(s²)U to nm⁵(s²)U and finally to mnm⁵(s²)U (15). The C-terminal FAD-dependent MnmC1 oxidoreductase domain transforms cmnm⁵(s²)U to nm⁵(s²)U and subsequently the N-terminal SAM-dependent MnmC2 methyltransferase domain methylates nm⁵(s²)U to mnm⁵(s²)U (16, 17). However, bifunctional MnmC is mainly restricted to the Pseudomonadota, whereas Gram-positive bacteria, such as Bacillus subtilis, instead rely on the recently discovered MnmL and MnmM enzymes to perform similar reactions (18–20). MnmL is a member of the radical SAM (rSAM) enzyme superfamily, a large and versatile group of enzymes that catalyze difficult chemical transformations often involving sp³ or sp² hybridized carbon centers via radical intermediates (21, 22). Interestingly, E. coli and related bacteria that encode MnmC also encode a homolog of MnmL, which is named YhcC in E. coli. As the deletion of mnmC in E. coli results in cmnm⁵(s²)U accumulation with no nm⁵(s²)U production under standard growth conditions, the function of these MnmL homologs in E. coli and related bacteria remained unclear (19).

Figure 1. An updated overview of the mnm⁵s²U biosynthesis pathway in E. coli.

MnmEG catalyzes one of two reactions: the ‘glycine route’ forms cmnm⁵s²U34 from U34 with glycine, tetrahydrofolate, and additional cofactors GTP, FAD, and NADH, and the postulated ‘ammonium route’ directly forms nm⁵s²U34 with ammonium in place of glycine. However, our data contests that the ‘ammonium route’ occurs in vivo (dotted line). MnmC contains two active sites: the MnmC1 active site is an FAD-dependent oxidoreductase that demodifies cmnm⁵s²U34 to nm⁵s²U34. We show that the MnmC1-catalyzed formation of nm⁵s²U is oxygen-dependent and only occurs under aerobic conditions (teal), and that the rSAM enzyme MnmL (formerly named YhcC) replaces MnmC1 under anaerobic growth conditions (purple). Finally, the MnmC2 methyltransferase domain of MnmC methylates nm⁵s²U34 to form mnm⁵s²U34 under both aerobic and anaerobic conditions.

Here, we determine the function of the previously uncharacterized E. coli rSAM enzyme YhcC by a combination of bioinformatics, genetics, and biochemistry. We show that YhcC is required for the conversion of cmnm⁵s²U to nm⁵s²U under anaerobic growth conditions in E. coli, and that the analogous reaction carried out by MnmC1 is likely inactive under anaerobiosis due to an O₂ requirement for FADH₂ oxidation during the FAD-dependent MnmC1 catalytic cycle. Additionally, we investigate the biochemistry of YhcC and show in vitro tRNA binding and conversion of cmnm⁵s²U to nm⁵s²U. Finally, we revisit the postulated growth-phase dependent activities of MnmEG in light of the presence of an anaerobically-regulated, YhcC-dependent route for nm⁵s²U synthesis and show that YhcC is in fact responsible for the reported ‘ammonium route’ activity of MnmEG during stationary phase. These studies redefine the tRNA wobble base mnm⁵s²U modification biosynthesis and its regulation under aerobic and anaerobic growth in E. coli.

Results

Gene neighborhood analysis, transcriptional regulation, and RNA-Seq data link YhcC to anaerobiosis.

To gather functional knowledge on the E. coli yhcC gene, we surveyed the literature and available transcriptomic data. It was reported that yhcC is positively regulated by the FNR (fumarate and nitrate reductase regulation) transcription factor in anaerobic growth by ChIP-Seq and RNA-Seq experiments (23, 24). FNR regulates the switch from aerobic to anaerobic metabolism in E. coli and related facultative anaerobic bacteria and is required for the expression of genes involved in anaerobic respiration (25, 26). Besides those involved in anaerobic respiration, additional genes are either activated or repressed by FNR and in total over 100 genes are regulated by FNR in E. coli. YhcC has also been confirmed to bind a [4Fe-4S] cluster and is capable of cleaving SAM to methionine and 5’-deoxyadenosine in the presence of a chemical reductant, confirming that it is a rSAM enzyme (27). However, the reaction catalyzed by the enzyme was unknown.

Genes involved in similar processes often are co-localized on bacterial chromosomes (28). Analysis of the gene neighborhood of yhcC in E. coli and other gammaproteobacteria shows that yhcC is often found adjacent to arcB, which encodes the sensor protein of the two-component Arc (aerobic respiration control) system (Fig. S1). The Arc system mainly represses aerobic respiration genes under anaerobic conditions (29). This association is in contrast to mnmL found in Gram positive bacteria, which is often adjacent to or fused with mnmM (19). Collectively, these analyses suggest YhcC may be active under anaerobic metabolic conditions as opposed to aerobic conditions.

Sequence and structural analysis do not differentiate E. coli YhcC from B. subtilis MnmL.

We gathered YhcC/MnmL sequences (IPR005911) from organisms that, like E. coli, encode mnmC and from organisms that do not, like B. subtilis, and constructed both a small and large multiple sequence alignment. The former allowed us to search for motifs that could separate proteins found in organisms that either lack or encode MnmC proteins. We failed to identify any residues that could differentiate the two groups (Fig. S2). We then used AlphaFold3 to predict the structure of E. coli YhcC and B. subtilis MnmL and the backbones aligned with an RSMD of 3.84 angstroms (Fig. S3B) (30). Subsequently, we used the large multiple sequence alignment to identify conserved residues. This enabled us to locate the approximate location of the YhcC active site based on the presence of highly conserved cysteine residues, which are used by rSAM enzymes to coordinate a [4Fe-4S] cluster. We found that the predicted structure of EcYhcC contains a cleft lined with positively charged residues adjacent to the potential active site and hypothesized that these residues may be involved in nucleic acid binding (Fig. S3). Two of these residues, K21 and R279 are highly conserved and are located at the opening of this cleft (Fig. S3).

Given the apparent association between yhcC and anaerobic metabolism and its close resemblance to recently experimentally validated MnmL enzymes, we hypothesized that YhcC in E. coli and related organisms may be a true MnmL ortholog that is active under anaerobiosis. Since previous work on E. coli ΔmnmC strains in relation to the biosynthesis of tRNA modifications only analyzed cells grown aerobically, the role of yhcC may have been overlooked.

E. coli YhcC is essential for mnm⁵s²U synthesis under anaerobic conditions.

Previous studies had shown that deletion of mnmC in E. coli results in the absence of mnm⁵s²U and the accumulation of cmnm⁵s²U in tRNA during exponential phase, or accumulation of nm⁵s²U in tRNAs extracted during stationary phase (14, 19). We confirmed these observations by analyzing modification profiles using LC-MS on bulk tRNAs extracted from E. coli cells grown aerobically to either exponential phase (OD₆₀₀ ≈ 0.6) (Fig. 2A) or stationary phase (OD₆₀₀ ≈ 3.0) (Fig. 2C). Furthermore, we showed that deletion of yhcC does not affect the ability of E. coli to produce mnm⁵s²U in tRNAs extracted from cells in both exponential and stationary phases under aerobic conditions (Fig. 2A and 2C).

Figure 2. Accumulation of nmn5s2U tRNA intermediates under aerobic and anaerobic conditions.

A-C (top): cmnm⁵s²U, nm⁵s²U, and mnm⁵s²U abundances, normalized by the amount of dihydrouridine (D) in the same sample, in tRNAs purified from E. coli at different stages of growth in either aerobic or anaerobic conditions. A: E. coli grown aerobically and harvested at mid-log phase, B: E. coli grown anaerobically and harvested at mid-log phase, and C: E. coli grown aerobically and harvested at stationary phase. D and E: In vitro MnmC activity assays under aerobic and anaerobic conditions, monitoring disappearance of substrate cmnm⁵s²U (D) and appearance of product nm⁵s²U (E) in purified tRNA pools obtained from ΔmnmC E. coli cells with recombinant MnmC enzyme.

Next, we tested the hypothesis that yhcC is required for cmnm⁵s²U to nm⁵s²U conversion under anaerobic conditions by repeating the above analysis with cells grown under anoxic conditions (Fig. 2B). We found that deleting mnmC or yhcC led to the absence of mnm⁵s²U in bulk tRNA. However, nm⁵s²U accumulated in the ΔmnmC strain but not in the ΔyhcC strain under anaerobic growth. The nm⁵s²U intermediate was not detected when both mnmC and yhcC were deleted (ΔmnmC/yhcC). Additionally, cmnm⁵s²U accumulated in tRNAs extracted from the ΔyhcC mutant grown under anoxic conditions despite the presence of mnmC. To confirm that the loss of mnm⁵s²U in anaerobically-grown E. coli ΔyhcC was not due to a polar effect, we complemented the loss of yhcC by expressing the wild-type yhcC gene using pBAD24 (Fig. 2B). The complemented E. coli ΔyhcC strain was found to only produce mnm⁵s²U, in contrast to both the knockout and the wild-type strain, which had detectable levels of cmnm⁵s²U. Finally, we substituted one of the cysteine residues (C33) likely involved in coordinating the rSAM [4Fe-4S] cluster to alanine. Expression of this variant did not recover the ΔyhcC phenotype, as it accumulated only cmnm⁵s²U, showing that the conversion of cmnn⁵s²U to nm⁵s²U relies on YhcC catalysis.

The apparent lack of YhcC activity during aerobic growth could either be due to a lack of yhcC transcription, or possibly due to the instability of the iron-sulfur cluster in the presence of molecular oxygen. To investigate these two scenarios, we expressed yhcC in pBAD24 containing the arabinose-inducible P_bad promoter under aerobic conditions in the double ΔmnmC/ΔyhcC background. We also overexpressed mnmL from Bacillus subtilis, which has previously been reported to be active when expressed in E. coli under aerobic conditions (19). We found that while the strains overexpressing either yhcC or mnmL had decreased levels of cmnm⁵s²U relative to the wild-type strain transformed with the empty vector, we surprisingly could not detect nm⁵s²U or mnm⁵s²U in either strain (Fig. S4). However, the strain overexpressing the C33A yhcC mutant reversed the observed decreased cmnm⁵s²U level chemotype, and instead, the tRNA modification profile of this strain appeared closer to the wild-type control. While we were unable to detect nm⁵s²U in the yhcC-overexpression strain, this supports that YhcC is likely active under aerobic conditions. The nm⁵s²U-containing tRNA could be unstable in this growth condition, or potentially, nm⁵s²U is converted to a modified nucleoside that we were unable to identify. We also were unable to detect the non-thiolated versions of the modified nucleosides.

Our epitranscriptomic tRNA analysis replicated an earlier finding that nm⁵s²U accumulates in tRNAs extracted from a ΔmnmC mutant grown to stationary phase (Fig. 2C) (31). However, we found that this modification is not detected in the double mutant, ΔmnmC/yhcC, and instead found an increase of the cmnm⁵s²U precursor. As stationary phase is the only growth condition in which the ammonium route of MnmEG has been proposed to be active, our findings challenge the idea that the nm⁵ moiety can be inserted directly by MnmEG in vivo (14). These results demonstrated that the formation of mnm⁵s²U in the absence of oxygen requires both the rSAM enzyme YhcC for the synthesis of the nm⁵s²U intermediate and the methyltransferase activity of MmnC2.

The FAD-dependent activity of MnmC1 was detected only under conditions where oxygen is present. We hypothesized this could be due to a requirement for O₂ to recycle the FADH₂ cofactor. Since previously reported E. coli MnmC activity assays had not tested the oxygen dependence of the MnmC1 oxidoreductase we tested the oxygen requirement of the MnmC1-catalyzed cmnm⁵s²U to nm⁵s²U conversion (Fig. 2) (17). Using hypomodified tRNA isolated from E. coli ΔmnmC overexpressing tRNA^Glu_UUC, which accumulates cmnm⁵s²U, we carried out in vitro activity assays with recombinant MnmC under both aerobic and anaerobic (< 1 ppm O₂) conditions, monitoring conversion of substrate cmnm⁵s²U into product nm⁵s²U with UHPLC-MS. Consistent with our in vivo data, we find that MnmC1 activity is significantly reduced under anaerobic conditions, suggesting that O₂ is indeed required for FADH₂ recycling and robust cmnm⁵s²U to nm⁵s²U tRNA modification activity by MnmC.

E. coli YhcC in vitro tRNA binding and catalytic activity.

As our genetic studies in E. coli showed a clear role for EcYhcC in nm⁵s²U-tRNA synthesis, we next set out to show directly that YhcC can bind tRNA and catalyze the conversion of cmnm⁵s²U to nm⁵s²U in vitro. For these experiments, we first overexpressed and purified EcYhcC from E. coli cells and chemically reconstituted its [4Fe-4S] cluster under anaerobic conditions (Fig. S5). To measure YhcC-tRNA interactions, we used electrophoretic mobility shift assays (EMSAs) with purified YhcC and unmodified E. coli tRNA^Lys_UUU prepared by in vitro transcription and found that YhcC binds tRNA with a K_D of approximately 5 μM (Fig. S6B). Next, we carried out tRNA modification assays using a substrate cmnm⁵s²U34-enriched tRNA pool and our purified, reconstituted YhcC under anaerobic conditions, monitoring cmnm⁵s²U to nm⁵s²U conversion by UHPLC-MS (Fig. 3). These reactions showed a clear YhcC- and SAM-dependent consumption of substrate cmnm⁵s²U over time, and the concomitant production of product nm⁵s²U, providing direct in vitro evidence that YhcC catalyzes the cmnm⁵s²U to nm⁵s²U reaction on tRNA.

Figure 3. YhcC in vitro assays catalyzing the conversion of cmnm⁵s²U to nm⁵s²U tRNA.

Activity assays were performed with purified, reconstituted YhcC (1 μM) and cmnm⁵s²U-enriched tRNA substrate (3.5 μM total tRNA concentration) obtained from E. coli ΔmnmC cells overexpressing tRNA^Glu carried out under anaerobic conditions. Reactions are quenched at different times with EDTA and tRNA is digested to single nucleosides for UHPLC-MS analysis. YhcC catalyzes the disappearance of substrate cmnm⁵s²U modifications (A) and production of nm⁵s²U modifications (B) in an enzyme- and SAM-dependent manner.

Discussion

Previous studies had implicated MnmL (previously called YtqA) from Bacillus subtilis, Streptococcus mutans, and Streptococcus pneumoniae, in mnm⁵s²U synthesis (19). However, while genetic studies show that MnmL from Bacillota is required for the synthesis of nm⁵s²U from cmnm⁵s²U, this had not been confirmed biochemically (19). We show here that the yhcC gene in E. coli encodes an mnmL ortholog that is required for nm⁵s²U formation specifically under anaerobic conditions, even in the presence of mnmC. Therefore, yhcC should be renamed to mnmL in E. coli and in related facultative anaerobes. Indeed, both our in vivo and in vitro data show that the FAD-dependent oxidoreductase (or MnmC1) activity of MnmC is not present under anaerobic conditions. Notably, the crystal structure of MnmC revealed that the MnmC1 active site is highly similar to that of Bacillus subtilis glycine oxidase, which utilizes molecular oxygen to oxidize reduced flavin cofactor (32, 33). In the absence of oxygen, assays with glycine oxidase from Marinomonas mediterranae with several alternative electron acceptors failed to detect activity (34). However, facultative anaerobes must adapt to conditions lacking oxygen, and under anaerobic metabolism. Some FAD-dependent oxidases can utilize alternative electron acceptors, such as fumarate in the case of L-aspartate oxidase (35). However, the lack of MnmC1-dependent cmnm⁵s²U demodification activity in the ΔyhcC strain under anaerobic conditions suggests a stringent dependency on O₂ for the oxidoreductase activity of MnmC. Indeed, in vitro kinetic measurements of MnmC in the presence or absence of oxygen show that O₂ greatly enhances activity. Although the enzyme was still capable of multiple turnovers under anaerobic conditions, it is likely due to residual oxygen in solution. Interestingly, despite the lack of MnmC1-dependent activity under anaerobic conditions, MnmC2-dependent activity is required to synthesize mnm⁵s²U from the nm⁵s²U-tRNA generated by YhcC, most likely due to an absence of an alternative nm⁵(s²)U methyltransferase (Fig. 2B). This conclusion is consistent with previous studies suggesting that the MnmC1 and MnmC2 domains act independently of each other (14, 32).

It is interesting that E. coli relies on YhcC for cmnm⁵s²U demodification when MnmC1 is inactive instead of utilizing the “ammonium route” activity of MnmEG, which produces nm⁵s²U directly. The dual MnmEG activities have been detected in vitro, and previously, it was reported that E. coli utilizes the ammonium route under stationary phase growth, and the glycine route under exponential phase (14). Indeed, stationary phase seems to be the sole growth condition in which the ammonium route has been reported to be active in vivo for any bacterial species. In the case of E. coli, it is possible that high levels of glycine in LB medium favors the glycine route over the ammonium route. Previously it was found that E. coli ΔmnmC produces cmnm⁵s²U but not mnm⁵s²U or nm⁵s²U in minimal medium containing free ammonia, however yeast extract containing glycine was also included in this experiment (31). In other cases, the ability of MnmEG to directly produce nm⁵s²U has been supported through indirect evidence. For instance, there are some bacteria that produce mnm⁵(s²)U despite lacking an enzyme capable of cmnm⁵(s²)U demodification (MnmC, standalone MnmC1 or MnmL). In other cases, such as in S. mutans or S. pneumoniae, deleting the gene encoding MnmL does not affect mnm⁵s²U production (19). While our genetic experiments were able to replicate nm⁵s²U production in stationary phase, we show this is in fact due to YhcC activity, and not MnmEG directly producing this intermediate. Therefore, direct evidence for the production of nm⁵s²U by MnmEG in vivo is lacking in this model organism despite in vitro evidence of this activity.

E. coli and other bacteria routinely encode parallel enzymes that carry out identical transformations under aerobic versus anaerobic conditions. rSAM enzymes in particular have been observed to be favored by anaerobes and facultative anaerobes to catalyze reactions otherwise catalyzed by oxygen-dependent enzymes in tetrapyrrole synthesis (36). For example, E. coli encodes both the oxygen-dependent oxidase HemF and the oxygen-independent dehydrogenase rSAM enzyme HemN, which each catalyze the conversion of coproporphyrinogen III to protoporphyronogen IX in heme biosynthesis (37, 38). Another key example is the ribonucleotide reductases (RNRs), which catalyze the formation of deoxynucleoside triphosphates for DNA synthesis, where class Ia tyrosine-cysteine radical RNRs utilize O₂ to generate a reactive complex under aerobic conditions, but class III glycyl radical RNRs (encoded by nrdD in E. coli) use an O₂-sensitive [4Fe-4S] cluster and rSAM chemistry to generate a reactive complex under anaerobic conditions (39). Like yhcC, nrdD transcription is activated by FNR under anaerobic growth conditions (40). In another example, parallel tRNA hydroxylation pathways and enzymes that utilize divergent O₂-dependent or independent chemistry ensure efficient biosynthesis of the 5-hydroxyuridine modification and its derivatives in E. coli under both aerobic and anaerobic metabolism (41). The work presented here on yhcC provides a new example of a dual pathway that allows E. coli to efficiently install mnm⁵s²U wobble base modifications that promote accurate translation under both aerobic and anaerobic growth conditions, by replacing the O₂-dependent MnmC1 reaction with the O₂-independent YhcC reaction. Such parallel pathways are key adaptations that enable bacteria to catalyze essential biochemical transformations under dramatically different environmental and growth conditions by leveraging divergent enzyme chemistries.

Materials and Methods

Bioinformatics.

The structure of E. coli YhcC and B. subtilis MnmL proteins was modeled using the Alphafold3 web server (alphafoldserver.com) and predicted structures were visualized in ChimeraX. To construct the model that included the [4Fe-4S] cluster, 5’dAdo, and methionine, EcYhcC was aligned to Homo sapiens Elp3 solved with these cofactors (PDB: 8PTY; ref 33). Multiple sequence alignments were constructed using MAFFT under default settings and visualized in UGENE (43, 44). The large multiple sequence alignment used to detect highly conserved residues was constructed with the sequences belonging to the repnode 55 sequence similarity network constructed as described previously, and proteins smaller than 225 and larger than 400 amino acids were discarded (19). Gene neighborhood analysis of YhcC and homologs from related bacteria (gamma and betaproteobacteria) was conducted using fast.genomics and gene neighborhoods were visualized using Gene Graphics (45, 46).

Bacterial strain construction and growth conditions.

A list of all strains utilized in this study can be found in Table S1. To ensure all mutants were in an isogenic background, E. coli knockout strains were constructed by P1 transduction into Escherichia coli str. K12 substr. BW25113 using corresponding mutants obtained from the Keio collection (47). To generate the double ΔyhcC ΔmnmC::kan knockout, the kanamycin resistance cassette was first evicted from BW25113 ΔyhcC::kan using the FLP recombinase expression plasmid pCP20 (48). For the complementation experiment, the E. coli yhcC gene was cloned into pBAD24 by Gibson assembly (New England Biolabs, HiFi assembly kit) utilizing the NcoI restriction site. The pBAD24 vectors containing mnmLM and mnmL from Bacillus subtilis were constructed previously by the Dos Santos lab (19). Site-directed mutants were constructed using the KLD kit (New England Biolabs). Gene expression was induced by the addition of arabinose to 0.02% at the time of inoculation. All strains were grown in LB medium (10 g L⁻¹ tryptone, 5 g L⁻¹ yeast extract, 10 g L⁻¹ NaCl) in either aerobic or anaerobic conditions at 37 °C. Under aerobic conditions, the strains were grown in 50 mL LB from overnight cultures with shaking in 500 mL Erlenmeyer flasks until either mid-log phase (OD₆₀₀ ~0.5–0.7), late log phase (OD₆₀₀ ~1.0), or stationary phase (OD₆₀₀ ~3.0). A Bactron anaerobic chamber (Sheldon Manufacturing) containing a 37 °C incubator in an atmosphere of 90% N₂, 5% H₂, 5% CO₂ was utilized for anaerobic growth. Plasticware and liquids were brought into the box two days before each growth experiment, and liquids were stirred overnight to allow for gas exchange. Strains were grown on solid LB medium aerobically, then brought into the anaerobic chamber to start the standing overnight cultures. The overnight cultures were subsequently subcultured into 50 mL of LB in Falcon tubes. Kanamycin sulfate (10 μg/mL) was added to agar and overnight cultures when growing the knockout strains, but omitted in the subcultures, and ampicillin (100 μg/mL) was added to select for pBAD24 derived plasmids.

Cloning and overexpression of YhcC

The E. coli YhcC gene was obtained as a codon-optimized gBlock from Integrated DNA Technologies and cloned into a pET28a bacterial expression vector with an N-terminal 6xHis affinity tag and TEV cleavage site using Gibson assembly (New England Biolabs’ HiFi DNA Assembly). The final vector sequence was confirmed using whole plasmid DNA sequencing (Plasmidsaurus). The YhcC pET28a and a pDB1282 plasmid containing the isc operon were co-transformed into E. coli BL21 (DE3) cells (49, 50). Bacteria were grown in LB media at 37 °C with shaking until reaching an OD600 between 0.6–0.8 before inducing YhcC expression with 10 μM IPTG. Induced cultures were grown at 18 °C overnight with shaking until cells were harvested at 6,000 × g for 40 min at 4 °C, flash frozen with liquid nitrogen, and stored at −80 °C.

Purification of YhcC

Purification of WT YhcC was carried out in an anaerobic chamber (MBraun) at an oxygen concentration <1 ppm, while YhcC used for binding studies were purified aerobically. Buffers used for purification are as follows: lysis/wash buffer: 50 mM Bis Tris, 500 mM KCl, 2 mM imidazole, 10 mM BME, pH 7.0; wash buffer: 50 mM Bis Tris, 300 mM KCl, 10 mM imidazole, 10 mM BME, pH 7.0; elution buffer: 50 mM, 300 mM KCl, 250 mM imidazole, 10% glycerol, 10 mM BME, pH 7.0; sizing buffer: 50 mM Bis Tris, 500 mM KCl, 5 mM dithiothreitol, pH 7.0. For buffers with 2-mercaptoethanol (BME), 10 mM BME was added immediately before use. All buffers were sparged with argon for 1 hour, moved into the glovebox, and stirred overnight in the anaerobic chamber. YhcC cell pellet was thawed in the glovebox, resuspended in 35 mL of lysis/wash buffer, then lysed in an air-tight sonicator adaptor on ice (Branson Sonifier 450, 4 × 2 min cycles at setting 6, 50 % duty cycle). After sonication, lysate was transferred to a centrifuge tube then centrifuged for 45 min at 14,500 × g and 4 °C. Centrifuge tubes were then brought back into the anaerobic chamber and mixed with 5 mL Ni-NTA resin for 1 hour; Ni-NTA resin was pre-equilibrated with lysis/wash buffer and sparged with argon gas. The Ni-NTA slurry was added into a gravity column and washed with 50 mL of lysis/wash buffer, 50 mL wash 2 buffer, and 50 mL wash 3 buffer. His-YhcC was eluted with 15 mL elution buffer and collected as 1 mL fractions. YhcC was identified by gel electrophoresis (10% SDS-PAGE gel), and YhcC-containing fractions were pooled and concentrated to 2.5 mL using Amicon Ultra 0.5 mL spin columns (10 kDa cutoff) and then purified with size exclusion chromatography on an AKTA Pure 25M with HiLoad 16/600 Superdex 75pg column, pre-equilibrated with 3 CV of anaerobic sizing buffer. SEC fractions were collected inside the anaerobic chamber.

Chemical Reconstitution of YhcC

SDS-PAGE was used to determine the sizing fractions containing YhcC. Fractions were concentrated and buffer exchanged into reconstitution buffer (50 mM Bis Tris, 500 mM KCl, 1% glycerol, 5 mM MgCl₂, pH 7.0). Protein was diluted to 50 μM with 1 M DTT to reach a final concentration of 5 mM DTT and allowed to incubate at 4 °C for 1 hour. Iron (III) chloride hexahydrate and sodium sulfide nonahydrate were used as Fe and S salts, respectively. A total of 10 equivalents Fe salt was added in 6 separate aliquots with 5 min between additions. After final addition of Fe salt, protein rested for 10 min before 10 equivalent S²⁻ was added in 6 separate aliquots with 15 min between additions. Reconstitution was allowed to continue for 2 hours before being spun down for 10 min at 14000 rpm and the supernatant was buffer exchanged into fresh reconstitution buffer using Amicon Ultra 0.5 mL spin concentrators (10 kDa cutoff) to remove excess Fe/S salts. Presence of the Fe-S cluster was verified by UV-Vis absorbance at 420 nm (Figure S5B). The yellow-brown supernatant was concentrated, flash frozen, and stored at −70 °C.

Purification of tRNA

tRNA was purified using the phenol-chloroform method utilizing TRIzol (Invitrogen) and the NucleoBond RNA/DNA 80 kit (Macherey-Nagel). Cultures reaching the desired growth phase were immediately placed on ice and centrifuged at 3,700 × g at 4 °C. Cell pellets were then resuspended in 1 mL TRIzol and placed on a rocker for 15 minutes. 200 μL chloroform was added, and the cell extracts were then centrifuged at 15,000 × g for 15 minutes to separate the organic and aqueous phases. Aqueous phases were transferred to new tubes and to this one volume of cold isopropanol was added. The samples were then placed at −20 °C for 30–60 minutes or 4 °C overnight. Samples were centrifuged at 15,000 × g for 20 minutes and the supernatants were discarded. In the NucleoBond purification, the RNA pellets were resuspended in 6 mL of buffer R0 (100 mM Tris/acetate, 15% ethanol, pH = 6.3) and applied to Nucleobond columns. The tRNA was then purified according to the manual, except tRNA was eluted with 1.1 mL buffer R2 (100 mM Tris/acetate, 15% ethanol, 900 mM KCl, pH = 6.3) diluted to 0.75 M KCl with buffer R0. The purified tRNA was precipitated by addition of 916 μL cold isopropanol followed by incubating at 4 °C overnight. Precipitated tRNA was centrifuged at 15,000 × g for 20 minutes, and the pellets were washed twice with 80% ethanol, dried, and resuspended in pure water.

To prepare cmnm⁵s²U-enriched tRNA for the YhcC and MnmC1 activity assays describe below, tRNA was purified from E. coli T7 Express ΔmnmC + pSGAT2-tRNA^Glu-II strain, constructed previously by the Carell lab, which provided the strain to us (17). The strain was grown aerobically with shaking in 250 mL LB medium to an OD₆₀₀ of 1.34, and tRNAs were purified as described above, except large RNAs were precipitated by the addition of NaCl to 1 M prior to the NucleoBond purification step.

Quantification of tRNA modified nucleosides

Relative levels of nucleoside modifications from tRNA purified from E. coli cultures were analyzed by chromatography-coupled mass spectrometry methods (LC-MS) as previously described (19). For each analysis, 40 ug of denatured tRNA were digested with nuclease P1 and alkaline phosphatase prior to LC-MS detection. Dihydrouridine standard (Dalton) was run as control for normalizing data as it is present in lower abundance, similar to the modifications analyzed in this study. The relative levels of each modification were determined by the area of mass abundance of each modification by mass abundance of dihydrouridine in the same sample. The desired analytes D ([M+H]⁺:247.0925), nm⁵s²U ([M+H]⁺: 290.0805), cmnm⁵s²U ([M+H]⁺: 348.086), mnm⁵s²U ([M+H]⁺:304.096), where the masses were determined within 5 ppm accuracy.

In vitro transcription of unmodified E. coli tRNA

The E. coli tRNA^Lys_UUU sequence (5’-GGGTCGTTAGCTCAGTTGGTAGAGCAGTTGACTTTTAATCAATTGGTCGCAGGTTCGAATCCTGCACGACCCACCA-3’) was obtained from the tRNADB-CE tRNA sequence database (https://trna.ie.niigata-u.ac.jp/cgi-bin/trnadb/whole_detail.cgi?SID=21833864&). dsDNA template with added 5’ T7 promoter sequence for this tRNA sequence was obtained from Integrated DNA Technologies; template dsDNA sequence = 5’-TAATACGACTCATATAGGGTCGTTAGCTCAGTTGGTAGAGCAGTTGACTTTTAATCAATTGGTCGCAGGTTCGAATCCTGCACGACCCACCA-3’). The tRNA sequence was in vitro transcribed using T7 polymerase in a 500 μL reaction: 1 μM DNA template, 5 mM rNTPs, 5 mM DTT, 0.05% v/v Triton X-100, 6 U/mL thermostable inorganic pyrophosphatase, and 0.15 mg/mL T7 RNA polymerase in 1 X T7 buffer (New England Biolabs). The reaction was incubated overnight at 37 °C before DNA template was digested with TURBO DNase (Thermo Fisher Scientific) for 1 hour at 37 °C. After this, tRNA was purified using denaturing urea PAGE: tRNA product band was gel extracted (RNA extraction buffer: 0.6 M NaOAC, 1 mM EDTA, pH 6.0, 0.01% SDS) and precipitated with 1/10 volume 3 M NaOAc and 3x sample volume 100% ethanol at −20 °C. Precipitated tRNA was pelleted by centrifugation, redissolved in water, and annealed (heated at 80 °C for 2 min, 60 °C for 2 min, then 100 mM MgCl₂ was added to a final concentration of 10 mM, and cooled on ice for 30 min). Annealed tRNA was aliquoted, flash frozen with liquid nitrogen, and stored at −80 °C until ready for use.

Electrophoretic Mobility Shift Assays (EMSAs)

Triplicate YhcC-tRNA binding reactions were prepared with an unmodified tRNA^Lys_UUU concentration of 200 nM and YhcC concentrations ranging from 50 – 0.26 μM protein (1.5X dilution series) in YhcC reconstitution buffer containing 10% glycerol (50 mM Bis Tris, 500 mM KCl, 10% glycerol, 5 mM MgCl₂, pH 7.0). Binding reactions were prepared by combining 2X protein with 2X tRNA to a final volume of 10 μL, incubated at 4 °C for 1.5 hours, and then 7 μL was loaded into each well of a 5 % MP TBE gel (Bio-Rad). EMSA gels were run in 0.5X TBE at 4 °C and 8 mA. Staining of tRNA was carried out with SYBR Gold in 0.5X TBE buffer for 30 min with gentle shaking. Gels were imaged on a Fluorochem Q imager (EtBr setting). Gel images were then analyzed using ImageJ to quantify bound and unbound tRNA species. Binding curves were fit to a Hill coefficient plot using Prism.

MnmC1 tRNA modification assay

Purified MnmC1 was obtained from the Carell lab and MnmC1 activity assays were carried out similar to as previously described (17). but under aerobic (benchtop) or anaerobic (inside glovebox with degassed buffers) conditions. Equal volumes of 2X tRNA purified from E. coli ΔmnmC cells as described above in water (7 μM) and 2X MnmC1 enzyme (0.1 μM) in 2X reaction buffer (120 mM Tris, pH 8.0, 40 mM NH₄Cl, 5 mM KCl, and 1.3 mM MgCl₂) were mixed at 37 °C to initiate the reaction. Final tRNA and enzyme concentrations in the reaction were 3.5 μM and 0.05 μM, respectively, in 1X reaction buffer. Reaction timepoints were quenched by 1:1 (by volume) addition of methanol : reaction aliquot. Quenched reaction samples were cleaned up using the Zymo Research Clean and Concentrate RNA kit, eluting isolated tRNA in 15 μL water prior to using an NEB nucleoside digestion mix (0.5 μL enzyme with 1.72 μL 10X digestion buffer). Digestion occurred overnight at 37 °C prior to being transferred to UHPLC-MS block for analysis by Agilent Bio-Inert 1260 Infinity II UHPLC system with Infinity Lab LC/MSD. The UHPLC was equipped with the Agilent Zorbax SB-Aq Rapid Resolution HD using a mobile phase with 0.1% formic acid and 100% LCMS-grade acetonitrile (A), 0.1% formic acid (B) and detected using positive ionization mode. The gradient was as follows for detection of cmnm⁵s²U (m/z = 348.1, 1.1min) and nm⁵s²U (m/z = 290.1, 0.607 min) at 0.4 mL/min: 0–2.50 min, 100% A; 2.50–6.00 min, 20% A/80%B; 6.00–10.5 min, 100% A. Concentrations of substrate cmnm⁵s²U and product nm⁵s²U nucleosides in quenched reaction samples were calculated relative to calibration curves made with synthetic nucleoside standards, which were synthesized as previously described (51).

YhcC tRNA modification assay

YhcC activity assays were carried out under anaerobic conditions (<1 ppm oxygen) with final reaction mix concentrations of: 1 μM reconstituted YhcC, 25 μM S-adenosyl methionine, 0.5 mM dithionite, and 3.5 μM tRNA purified from ΔmnmC E. coli cells, similar to as described above for MnmC1 assays. 2X YhcC + cofactors solutions were incubated at 37 °C for 10 min and reactions were initiated by adding 2X tRNA. All activity assays were run in triplicate with time points collected by quenching 15 μL of reaction mix with 1:1 methanol. Timepoints were collected at 0, 90, 300, 600, 1800, 5400 s. Quenched samples were removed from the anaerobic chamber and cleaned up using Zymo Research Clean and Concentrate RNA kit, eluting isolated tRNA in 15 μL water. NEB nucleoside digestion mix was added, 0.5 μL enzyme mix with 1.72 μL 10X digestion buffer and allowed to digest overnight at 37 °C. Reactions were transferred to UHPLC-MS block for analysis by Agilent Infinity series and concentrations of substrate cmnm⁵s²U and product nm⁵s²U nucleosides were calculated as detailed above for MnmC1 assays.