Acinetobacter baumannii Catabolizes Ethanolamine in the Absence of a Metabolosome and Converts Cobinamide into Adenosylated Cobamides

To determine whether A. baumannii could use EA as a nitrogen source, growth was assessed in minimal medium containing M9 salts made without NH₄Cl but supplemented with ethanolamine as the sole source of nitrogen (Fig. 4A). To determine whether A. baumannii could use EA as a carbon and energy source, M9 minimal medium (containing NH₄Cl) was supplemented with ethanolamine as the sole source of carbon and energy (Fig. 4B). The growth of A. baumannii with ethanolamine depended on wild-type alleles of the eutBC genes (Fig. 4A and B, gray squares), which encode the subunits of the adenosylcobalamin (AdoCbl)-dependent EAL enzyme subunits. Chromosomal levels of eutBC expression were sufficient to support growth with ethanolamine as a nitrogen source (Fig. 4A, yellow squares). When eutBC was overexpressed in trans, the cultures reached a higher density (Fig. 4A, blue circles). In contrast, A. baumannii grew with ethanolamine as the sole source of carbon and energy in medium supplemented with cyanocobalamin (CNCbl) only when eutBC was overexpressed (Fig. 4B, blue circles versus yellow squares).

We isolated spontaneous mutant strains of A. baumannii that abolished the extended lag time of the wild-type strain growing on ethanolamine as a carbon and energy source (Fig. 4B). The putative mutants were obtained after ~100 h of incubation in liquid medium at 37°C. To test the stability of the causative mutation(s), we passaged the mutant strain several times on solid and liquid rich medium devoid of ethanolamine. This strain maintained the ability to rapidly grow with ethanolamine as a carbon and energy source. These results suggested that the phenotype was the result of one or more stable mutations. To determine the nature of the mutation(s), we sequenced the genome of the mutant strain, and the analysis of the sequence revealed an insertion of the ISAba1 transposable element upstream of the eutBC genes, between the 3′ end of the eutH gene (which encodes a putative ethanolamine permease) and the start of eutB (which encodes the putative large subunit of ethanolamine ammonia-lyase) (Fig. 5; see also Fig. S5 in the supplemental material). This element was inserted immediately after the final nucleotide of eutH, leaving the 16-bp length upstream of eutB undisrupted.

Previous reports have shown that the ISAba1 element contains a σ⁷⁰ promoter at the 3′ end of the gene and that this promoter is used to increase the expression of downstream genes (22). This element is frequently associated with the increased expression of antibiotic resistance genes in a clinical setting (22–26). The predicted σ⁷⁰ sites found in ISAba1 are illustrated in Fig. S5.

We used the strain with the ISAba1 element driving the synthesis of AbEutBC to determine the optimal concentrations of ethanolamine as a sole carbon source and CNCbl during growth in minimal medium (Fig. 6A and B). Strong growth was observed at 10 nM CNCbl and 75 mM ethanolamine as the sole source of carbon and energy (Fig. 6A, blue squares, and Fig. 6B, blue circles). As little as 0.5 nM CNCbl and 10 mM EA promoted robust growth when EA was used as the sole source of nitrogen by the A. baumannii parental strain (Fig. 6C, yellow squares, and Fig. 6D, blue circles).

We used well-characterized eut mutant strains of S. Typhimurium to assess the level of activity of the AbEutBC enzyme in different genetic backgrounds, namely, a ΔeutBC strain, a ΔeutABC strain, or a strain carrying a deletion of the entire eut operon (Δeut) (Fig. 7). AbEutBC supported the growth of an S. Typhimurium ΔeutBC strain with EA as the sole source of nitrogen as well as SeEutBC did (Fig. 7A, orange squares versus green circles). Notably, the AbEutBC enzyme remained active in the absence of the SeEutA reactivase (Fig. 7B, green circles), while, as expected, the SeEutBC enzyme did not support growth in the absence of SeEutA (Fig. 7B, orange squares). Most intriguingly, AbEutBC alone supported ethanolamine utilization in an S. Typhimurium strain in which the entire eut operon was deleted. Consistent with the need for SeEutA and other eut operon gene products, SeEutBC supported poor growth of the Δeut strain (Fig. 7C, green circles versus orange squares). SeEutBC components (49.4 and 32.1 kDa, respectively) were detected at higher levels in the crude lysate (Fig. 7E, lanes 2 to 4) than AbEutBC (50.6 and 30.1 kDa, respectively) (Fig. 7E, lanes 5 to 8). This was likely due to a stronger reaction to the anti-EAL antibody, as this antibody was generated against SeEutBC. The reduced signal observed for AbEutC was likely due to protein instability under the tested conditions, as the theoretical pI of AbEutC is 6.96. The loading buffer and gel stacking layer in this experiment were both pH 6.8.

To test whether the putative eut genes were cotranscribed, primers were generated to amplify the intergenic regions between ald1 and eutH, eutH and eutB, and eutB and eutC (Fig. 8A). As previously shown in Pseudomonas aeruginosa, all four genes were transcribed as a polycistronic mRNA and expressed from a promoter upstream of ald1 in wild-type A. baumannii (Fig. 8B) (7). We used quantitative reverse transcription PCR (RT-qPCR) to determine that the evolved strain (ES) overexpresses eutBC (Fig. 8C). In the absence of EA, wild-type A. baumannii produced extremely low levels of eutBC transcripts (Fig. 8C, yellow striped bars). The evolved strain carrying an IsAba1 insertion upstream of eutBC grown with ethanolamine as a sole carbon source showed significant increases in eutBC expression (100-fold and 64-fold, respectively) compared to the parental strain, while eutH levels were very low in both strains and under both conditions tested. We tested the expression of the upstream ald1 gene to determine whether this might act as an acetaldehyde dehydrogenase gene and aid in the management of reactive intermediates during EA catabolism. Interestingly, ald1 was expressed at high levels in wild-type A. baumannii even in the absence of EA (Fig. 8D); therefore, fluctuations in gene expression in response to EA may not be detectable under the conditions used. This suggested that under some conditions, ald1 and eutHBC were cotranscribed, but the expression of ald1 may be subject to other regulatory mechanisms.

We used transmission electron microscopy (TEM) and immunogold labeling to localize the AbEutBC proteins within the A. baumannii cell. Previous reports have shown that SeEutBC was associated with the metabolosome (27), which can be visualized by TEM (6). No similar structures were visible in A. baumannii cells in the presence of EA as a sole carbon source (Fig. 9A; Fig. S6A and B). In addition, AbEutBC localized primarily to the periphery of cells, with some protein apparently being interspersed throughout the cytosol. We further investigated localization within the cell by overexpressing AbEutBC in an S. Typhimurium ΔeutBC strain (Fig. 9B; Fig. S6C and D). It is important to note that the genome of this strain of S. Typhimurium harbors the genes for the metabolosome shell. When expressed in the presence of the eut metabolosome, AbEutBC did not associate with the metabolosome structure; strikingly, it was associated exclusively with the cell membrane (Fig. 9B). No EAL was detected in control samples treated with a secondary antibody only (Fig. S7).

The SeEutBC enzyme requires AdoCbl as its cofactor (28). Our bioinformatics analysis of the A. baumannii genome indicated that this bacterium lacked the functions needed for the de novo synthesis of the corrin ring of cobamides but appeared to encode the function needed to adenosylate the corrin ring, i.e., to attach the adenosyl (Ado) moiety from ATP to the Co ion of the ring, and the functions required for the assembly of the nucleotide loop (29).

The formation of a complete coenzyme from the precursor cobinamide requires the nucleotide loop assembly (NLA) pathway (Fig. S4). Under the conditions tested, A. baumannii salvaged the precursor cobinamide and other nitrogenous bases to form a functional coenzyme and facilitate growth with EA, demonstrating that NLA pathway enzymes were functional (Fig. 10A). We also tested the activity of NLA enzymes using heterologous expression in a MetH-dependent S. Typhimurium strain. The A. baumannii NLA enzymes CobU, CobS, and CobT compensated for the absence of the cobU, cobS, and cobT genes of S. Typhimurium (Fig. 10B and C). Furthermore, the A. baumannii pduO-type ACAT supported EA metabolism in an S. Typhimurium ACAT-null strain. The NLA enzymes CobU and CobS require adenosylated substrates (30, 31); thus, salvaging of the precursor cobinamide depended on the function of the A. baumannii ACAT under these conditions (Fig. 10B and C). These results demonstrated that A. baumannii NLA and ACAT homologues were functional.

Prokaryotes can utilize numerous nucleobases to yield a functional coenzyme (32). As shown in Fig. 10A, various bases can be used by A. baumannii to form a functional coenzyme. Salvaging was performed in the absence of exogenous nucleobases; however, this led to diminished growth (Fig. 10A, purple circles). To determine the identity of this cobamide produced by A. baumannii in the absence of exogenous nucleobases, we extracted cobamides and separated them using reverse-phase high-performance liquid chromatography (RP-HPLC) as described in Materials and Methods. A compound eluted off the column 33 min after sample injection. The retention time was equal to the retention time of commercially available CNCbl (Fig. S8A and B). This peak was collected, cleaned, and analyzed using matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) mass spectrometry, confirming that cobalamin, the cobamide containing 5,6-dimethylbenzimidazole (DMB) as its nucleobase, was the cobamide synthesized by A. baumannii in the absence of exogenous nucleobases (Fig. S8C).

Based on the data presented here, we surmised that the wild-type strain of A. baumannii used in this work synthesizes insufficient AbEutBC to support growth with EA as the sole carbon and energy source. It is possible, however, that in its natural environment, A. baumannii uses EA mainly as a source of nitrogen, and for that purpose, enough EAL is produced from chromosomal expression levels. Our data show that A. baumannii can use the ISAba1 element to increase the expression of eutBC if selective pressure for the use of ethanolamine as a carbon and energy source arises in its environment. Previous studies have identified ISAba1 upstream of antibiotic resistance genes in patient samples, suggesting that this is a clinically relevant adaptive mechanism used by this organism during human infection (23). The insertion of an ISAba1 element does not affect the expression of the upstream eutH and ald1 genes. The low level of eutH, even in the presence of EA, suggests that, under the conditions tested, active transport is not required and that sufficient EA diffuses across the membrane into the cell. We have shown that the requirements for both cobamide and EA concentrations are lower in A. baumannii than in S. Typhimurium. This reduced need may reflect the faster access of enzymes to these molecules due to the lack of restriction by the metabolosome.

We did not identify genes encoding homologues of metabolosome shell proteins in the A. baumannii genome, suggesting to us that A. baumannii may be able to perform this metabolism without containment in a microcompartment. The N terminus of SeEutC contains a signal sequence involved in localization to the microcompartment (33). Consistent with the lack of any genes encoding microcompartment shell proteins, the AbEutC sequence also lacks any predicted N-terminal signal sequence (see Fig. S1D in the supplemental material). We showed here that AbEAL does not require a reactivase; furthermore, AbEAL alone can support EA utilization in the absence of the entire S. Typhimurium eut operon. This suggests that either the AbEAL enzyme directly quenches the acetaldehyde generated from EA or A. baumannii reduces cellular damage caused by acetaldehyde using an alternative mechanism. The mechanism by which AbEAL functions in the absence of a physical metabolosome is intriguing. Using TEM, we showed that A. baumannii does not form metabolosomes and that the enzyme does not associate with metabolosomes when expressed in S. Typhimurium (Fig. 9). Based on these data, we surmise that the AbEAL enzyme acts independently of other Eut components in S. Typhimurium. The membrane association of AbEAL was more pronounced when expressed in S. Typhimurium. It is possible that this is due to the association of AbEAL with other B₁₂-related enzymes not found in A. baumannii. We hypothesize that in A. baumannii, other enzymes, such as the cotranscribed aldehyde dehydrogenase, associate with AbEAL to facilitate the swift quenching of reactive molecules and prevent cellular damage. To our knowledge, this is the first report of ethanolamine metabolism occurring in an organism that does not form a metabolosome.

A. baumannii carries the genes for nucleotide loop assembly as well as a functional PduO-type ACAT. Due to the apparent absence of other genes related to propanediol metabolism, we instead refer to this protein as a general ACAT (AcaT). The unique organization of these genes in A. baumannii is notable, as the cobUST genes are typically organized immediately adjacent to each other. We have shown experimental evidence that the cobU, cobT, and cobS gene products facilitate the use of various nucleobases to form the nucleotide loop and that the resulting cobamides function as coenzymes for AbEAL. This is consistent with a previous report in which SeEAL was shown to also use pseudocobalamin (34). Structural studies would provide insight into the ability of this enzyme complex to utilize multiple cobamide coenzymes.

Understanding the metabolic capabilities of A. baumannii is of great interest as we attempt to combat antimicrobial resistance. EA has been shown to play a role in the environmental sensing and pathogenesis of S. Typhimurium and Escherichia coli (3, 35, 36). Although A. baumannii is not generally associated with gastrointestinal disease or part of a normal gut microbiome, it has been found to inhabit the digestive tract of long-term-ICU (intensive care unit) patients (37), indicating that the intestine may be a reservoir for this nosocomial pathogen. EA breakdown may be an important metabolic capability for A. baumannii during the course of colonization or infection, and if so, the enzymes studied here could serve as effective therapeutic targets.

The S. Typhimurium and A. baumannii strains used in this study are listed in Table S1 in the supplemental material. All A. baumannii strains used were derivatives of A. baumannii ATCC 17978. All S. Typhimurium strains used were derivatives of S. enterica subsp. enterica serovar Typhimurium strain LT2. E. coli DH5α was used for plasmid construction. All cultures were grown at 37°C. A. baumannii was grown in lysogeny broth (LB) (38, 39) or M9 minimal medium (40). To test EA as a carbon source, M9 minimal medium was supplemented with MgSO₄ (1 mM), CaCl₂ (10 μM), EA-HCl (90 mM, unless otherwise noted), LB (1%, vol/vol), and cyanocobalamin (CNCbl) (100 nM, unless otherwise noted). To test EA as a nitrogen source, A. baumannii was grown in nitrogen-free M9 salts supplemented with MgSO₄ (1 mM), CaCl₂ (10 μM), sodium pyruvate (20 mM), CNCbl (10 nM, unless otherwise noted), and EA-HCl (30 mM, unless otherwise noted). S. Typhimurium strains were grown on nutrient broth (NB; Difco) or no-carbon essential (NCE) minimal medium (41); E. coli was grown in LB. When antibiotics were used, their concentrations were indicated for each experiment. Unless otherwise noted, isopropyl β-d-1-thiogalactopyranoside (IPTG) was used at 1 mM, and l-(+)-arabinose was used as an inducer of gene expression at 0.5 mM. Unless otherwise specified, all chemicals were purchased from Sigma-Aldrich Chemical Company.

Deletions were constructed as described previously (42, 43). A. baumannii harboring plasmid pRecABtet was transformed with a fragment containing flanking regions of the gene of interest fused to the kanamycin (Km) resistance gene from plasmid pKD4 by electroporation. Cells were recovered using superoptimal broth with catabolite (SOC) repression medium (tryptone [20 g/L, wt/vol], yeast extract [5 g/L], NaCl [0.5 g/L], glucose [20 mM]) at 37°C with shaking at 220 rpm for 2 to 4 h. Following incubation, cells were plated onto LB (Difco) agar plates containing Km (25 μg/mL) and IPTG (2 mM). Kanamycin-resistant (Km^r) transformants were patched onto plates containing tetracycline (Tc) (10 μg/mL), and tetracycline-sensitive (Tc^s) Km^r colonies were analyzed further. To resolve the kanamycin cassette, plasmid pFLPtet was moved into the strains by electroporation. Cells were recovered with SOC medium (2 mL) for 2 to 4 h at 37°C with shaking at 220 rpm and plated onto LB agar containing tetracycline (10 μg/mL). Cells were streaked for isolation onto plates containing tetracycline (10 μg/mL) and IPTG (2 mM) to induce the expression of flp. Isolated colonies were streaked onto LB and then patched onto LB agar containing kanamycin (30 μg/mL) or tetracycline (10 μg/mL). We isolated genomic DNA (gDNA) from Tc^s Km^s colonies as described previously (42), followed by PCR amplification and DNA sequencing to verify the disruption of the gene of interest and the loss of the Km resistance cassette.

Plasmids used in this study are listed in Table S1. Primers used for the construction of strains, the construction of complementation plasmids, and PCR analysis are listed in Table S2. To test complementation in S. Typhimurium, each gene of interest from A. baumannii was cloned into cloning vector pCV1 (44) using BspQI restriction sites and electroporated into E. coli DH5α. Plasmids were miniprepped using a Wizard Plus SV Minipreps DNA purification system (VWR) and sequenced to verify each insert. Positive plasmids were electroporated into an S. Typhimurium strain carrying a deletion of the homologous gene. Cells containing the plasmid were selected for by the addition of ampicillin (100 μg/mL) to the growth medium.

Complementation studies in A. baumannii were performed as described previously (45). A. baumannii eutBC was inserted into pMMB67EH using EcoRI and SalI (Thermo Fisher) restriction digests, and ligation was then performed using T4 DNA ligase (Thermo Fisher) according to the manufacturer’s instructions. Chemically competent E. coli DH5α was transformed and plated onto Tc (10 μg/mL). Selected colonies were miniprepped and sequenced to verify the insertion. Verified plasmids were used to transform electrocompetent A. baumannii ΔeutBC as described above, and cells were recovered in SOC medium for 2 to 4 h and plated onto LB agar containing tetracycline.

A. baumannii strains were grown in LB overnight and subcultured (5%, vol/vol) into M9 minimal medium or nitrogen-free M9 minimal medium. Growth analyses were performed using a 96-well microtiter dish (Falcon) with 200 μL per well, cultures were incubated at 37°C with shaking, and increases in cell density were monitored at 630 nm for 24 h. Each growth analysis consisted of three biological replicates in technical triplicate. Concentrations of CNCbl or ethanolamine were changed where indicated to assess the minimal requirements of A. baumannii for these nutrients.

The heterologous expression of A. baumannii genes was performed in S. Typhimurium to test EAL, NLA, and ACAT functions. NB was inoculated with a single colony, and the colony was grown overnight at 37°C. The following day, 1% (vol/vol) of overnight culture (~16 h old) was transferred into minimal medium in a 96-well microtiter plate. S. Typhimurium growth experiments were performed in NCE minimal medium (41) containing Wolfe’s trace minerals (46), MgSO₄ (1 mM), glycerol (22 mM), and ampicillin (100 μg/mL). Cobamide precursor salvaging was performed by assessing growth with the addition of dicyanocobinamide [(CN)₂Cbi] and 5,6-dimethylbenzimidazole (DMB) (150 μM). Gene expression was induced with l-(+)-arabinose (0.5 μM). The growth of cultures in microtiter plates was monitored every hour at 630 nm during incubation with shaking at 37°C in a BioTek Elx808 microplate reader. To test the functionality of the A. baumannii eutBC gene products, minimal medium containing no-carbon nonessential NCE medium (47) with ethanolamine (30 mM) provided as the sole source of nitrogen was used. In each case, growth analysis was performed in a BioTek Elx808 microplate reader with three biological replicates in technical triplicate.

Strains of A. baumannii able to grow on ethanolamine as the sole source of carbon and energy were grown overnight in LB (10 mL) at 37°C with shaking at 180 rpm. Cells were harvested by centrifugation at 3,000 × g using a tabletop Eppendorf 5810R centrifuge for 30 min at room temperature. Genomic DNA was isolated using a blood and cell culture DNA midi kit and a genomic DNA buffer set (Qiagen) according to the manufacturer’s protocol for the preparation of Gram-negative samples. The quality of genomic DNA was measured using an Invitrogen Qubit model 4 fluorometer, and only samples containing >10% (wt/vol) double-stranded DNA were used for whole-genome sequencing (WGS). DNA-end repair was performed using NEBNext FFPE (formalin-fixed, paraffin-embedded) DNA repair mix and NEBNext Ultra II end repair (New England BioLabs [NEB]) according to the manufacturer’s instructions. The repair reaction was allowed to proceed for 30 min at 20°C and 30 min at 65°C. DNA was isolated by mixing with AMPure XP beads, washing with 70% (vol/vol) ethanol, and eluting off beads with Ambion nuclease-free water. The eluate containing end-repaired DNA was quantified using a Qubit fluorometer. Barcodes were attached using NEB blunt/TA ligase and master mix (New England BioLabs) according to the manufacturer’s instructions, and the mixture was incubated at 25°C for 15 min. DNA was purified from the reaction mix using AMPure beads according to the manufacturer’s instructions and eluted in 25 μL nuclease-free water. Barcoded DNA was quantified using a Qubit fluorometer. Equimolar amounts of barcoded samples were pooled to a total of 1,050 ng in 65 μL. Adapters were attached to barcoded DNA using adapter mix, NEBNext ligation buffer, and NEBNext quick T4 DNA ligase (New England BioLabs), and the mixture was incubated at 20°C for 15 min. Following incubation, DNA was isolated from the reaction mixture using AMPure XP beads as described above. Beads were washed with long-fragment buffer (Nanopore) twice before elution with elution buffer (Nanopore). The final concentration of the library was quantified using a Qubit fluorometer. Libraries were loaded into an R10.3 flow cell on a MinION Mk1C system (Nanopore Technologies).

Purified S. Typhimurium EAL was provided to Envigo for the production of rabbit polyclonal antibodies. Antiserum was precleared against JE8094 (ΔeutBC). Cells were grown at 37°C with shaking at 180 rpm in minimal NCE medium containing ribose (20 mM), EA (30 mM), and CNCbl (200 nM) to an optical density at 600 nm (OD₆₀₀) of ~1.0. Cells were incubated on ice for 15 min and then harvested by centrifugation at 2,000 × g for 15 min in a 5810 R centrifuge (Eppendorf). Cells were resuspended in 100 μL Tris-HCl buffer (50 mM [pH 7.5] at 4°C) containing SDS (2%, wt/vol). To lyse cells, the suspension was boiled for 5 min and then incubated on ice for 10 min. The solution was brought to a final volume of 1 mL with Tris-HCl buffer (50 mM [pH 7.5] at 4°C). Antiserum was added at a 1:100 dilution, and the mixture was incubated at 4°C for 36 h. Precleared antibody was obtained by centrifugation at 16,000 × g for 10 min and removal of the supernatant. Precleared antibody was flash-frozen in liquid N₂ and stored at −80°C until use.

S. Typhimurium strains were grown in NCE or no-carbon or no-nitrogen (NCN) medium containing EA (30 mM) as the sole source of nitrogen. Each medium was supplemented with Wolfe’s trace minerals, MgSO₄ (1 mM), glycerol (22 mM), CNCbl (150 nM), ampicillin (100 μg/mL), and l-(+)-arabinose (0.5 μM). Cultures were inoculated with LB (2%, wt/vol) and grown overnight with shaking at 180 rpm at 37°C to an OD₆₀₀ of 0.45 to 0.65. Cells were harvested using a 5810 R centrifuge (Eppendorf) at 2,220 × g for 15 min and stored at −80°C. Cells were resuspended in 1 mL sterile phosphate-buffered saline (PBS) (10 mM sodium phosphate [pH 7.4], 0.9% [wt/vol] NaCl) and centrifuged at 4,000 × g for 10 min in a 5415 D centrifuge (Eppendorf) at 4°C. Cell pellets were resuspended in bacterial protein extraction reagent (B-PER) (300 μL; Thermo Fisher) and vortexed for 1 min to lyse cells. The protein concentration in the crude lysate was quantified using Bradford protein assay dye (Bio-Rad Laboratories) according to the manufacturer’s instructions. Bovine gamma globulin (Thermo Fisher) was used to generate a standard curve. Cell lysates were diluted to 500 ng/μL in PBS and incubated with loading buffer (Tris [pH 6.8] [300 mM], glycerol [60%, vol/vol], EDTA [12 mM], 2-mercaptoethanol [0.85 M], and bromophenol blue [0.05%, wt/vol]) for 5 min at 95°C. The crude lysate was loaded in duplicate onto a 15% (wt/vol) SDS-PAGE gel (48). Gels were loaded with Precision-Plus protein standards (Bio-Rad Laboratories) or a SuperSignal molecular weight protein ladder (Thermo Fisher). Gels intended for Coomassie staining were loaded with 5 μg of the crude lysate; gels intended for Western blotting were loaded with 5 μg for strains complemented with AbEutBC or vector-only control (VOC) and 1.25 μg for strains complemented with SeEutBC. A constant voltage of 200 V was applied to gels for approximately 35 min. Gels loaded with Precision-Plus protein standards were stained with brilliant blue R dye to visualize proteins. Gels loaded with the SuperSignal molecular weight protein ladder were transferred to a polyvinylidene difluoride (PVDF) membrane using a Trans-Blot Turbo system (Bio-Rad Laboratories). Following transfer, the membrane was incubated in blocking buffer composed of 0.5% (wt/vol) nonfat milk in PBST (phosphate-buffered saline [pH 7.4] containing 0.1% [vol/vol] Tween 20 [Millipore Sigma]) for 1 h at room temperature on an advanced digital shaker (VWR). The primary antibody generated against S. Typhimurium EAL was diluted 1:2,000 in blocking buffer, and the membrane was incubated in this solution at 4°C overnight on a tabletop shaker. The following morning, the membrane was washed three times with PBST. The membrane was incubated in PBST containing a goat anti-rabbit IgG–horseradish peroxidase conjugate (1:20,000; Invitrogen) for 1 h at room temperature on a tabletop shaker. The membrane was washed three times with PBST, incubated with SuperSignal West Pico Plus chemiluminescent substrate (Thermo Fisher) for 10 min at room temperature, and imaged using a UVP ChemStudio system (Analytik Jena) with an exposure time of 5 s.

RNA was isolated as described previously (49). A culture grown overnight in LB medium inoculated with a single colony was subcultured (2.5%, vol/vol) in 10 mL minimal medium. Cultures grown for RT-qPCR analyses were grown in M9 minimal medium containing either ethanolamine (90 mM) or pyruvate (20 mM) as a sole carbon and energy source. For cultures used in experiments aimed at determining whether the A. baumannii putative eut genes were cotranscribed, wild-type A. baumannii (JE25551) was grown in no-nitrogen M9 minimal medium supplemented with 30 mM ethanolamine and 20 mM pyruvate. In all cases, cultures were incubated with shaking (200 rpm) at 37°C to an OD₆₀₀ of 0.4 with 0.5 and harvested using a 5810 R centrifuge (Eppendorf) at 3,220 × g for 10 min. The supernatant was removed, and cells were stored at −80°C until they were processed. Cell pellets were lysed by resuspending the cells in boil solution (EDTA [18 mM], SDS [0.025%, vol/vol], 2-mercaptoethanol [1%, vol/vol], formamide [95%, vol/vol], RNase-free water brought to a final volume of 1 ml and heating the suspension at 95°C for 7 min. The cell suspension was centrifuged at 16,000 × g for 5 min at room temperature, and 100 μL of the supernatant was removed into a new tube. RNase-free water (400 μL) was added to dilute the culture, and sodium acetate was added to a final concentration of 0.3 M. Ice-cold ethanol (100%) was added (1,650 μL), and samples were vortexed briefly to mix and then incubated at −80°C overnight. Samples were centrifuged for 30 min at 16,000 × g at 4°C, and the supernatant was discarded. The remaining pellet was washed with 300 μL of ice-cold 70% ethanol and centrifuged at 8,000 × g for 5 min at 4°C. The supernatant was discarded, and the pellet was allowed to air dry. RNA pellets were resuspended in 100 μL RNase-free water and centrifuged for 1 min at 16,000 × g to remove any insoluble contaminants, and 90 μL was transferred into a new tube. A DNA digest was performed using the rigorous DNase treatment for Turbo DNase according to the manufacturer’s instructions (Thermo Fisher). Following DNase treatment and inactivation, samples were centrifuged at 10,000 × g for 1.5 min, and 90 μL was transferred into a new tube. A second sodium acetate-ethanol precipitation step was performed as described above. RNA yield and quality were analyzed using a Qubit 4 fluorometer using the RNA broad range (BR) assay kit and the RNA IQ assay kit, respectively (Invitrogen). Samples with an RNA IQ value of 6 or higher were used for RT-qPCR experiments.

RNA to be used as the cDNA template was diluted to a final concentration of 12.5 ng/μL, and cDNA was synthesized using an iScript cDNA synthesis kit (Bio-Rad) according to the manufacturer’s protocol. cDNA was diluted to 2.5 ng/μL when used for RT-qPCR experiments. RT-qPCRs were performed in a 96-well plate. In each well, 20 μL of reaction mix containing 10 mL Fast SYBR green master mix (Life Technologies), gene-specific primers (500 nM), and cDNA (10 ng) was added. Gene-specific primers were designed using Primer3 software. The A. baumannii 16S rRNA gene was used as an internal control. RT-qPCRs were run using a 7500 Fast real-time PCR system (Applied Biosystems). Three biological replicates each in technical triplicate were analyzed for each strain. The cycle threshold (C_T) was normalized against 16S values, and these values were transformed using the formula 2(e − ΔC_T)/10⁻⁶ and reported as arbitrary expression units (EU). The mean EU from each biological replicate was used to calculate the standard error in Prism software v9 (GraphPad). Statistical significance was determined using Welch’s t test in Prism software v9 (GraphPad).

Intergenic primers were generated to amplify the regions between genes within the putative operon. Five microliters of the iScript reaction product with reverse transcriptase (+RT) or without reverse transcriptase (−RT) was added to each reaction mixture. gDNA was diluted to 12.5 ng/μL, and 5 μL was added to each control reaction mixture. PCR was performed using GoTaq green master mix (Promega) according to the manufacturer’s instructions. Products were analyzed on a 1% (wt/vol) agarose DNA gel.

A. baumannii was grown in M9 minimal medium containing EA as a sole carbon source. S. Typhimurium was grown in NCN medium supplemented with glycerol (22 mM) and EA (30 mM). In both cases, cultures were harvested at an OD₆₀₀ of ~0.6 in a 5810 R centrifuge (Eppendorf) at 3,220 × g for 10 min. Cell pellets were resuspended in Sorensen’s phosphate buffer (0.1 M sodium phosphate [pH 7.4]) containing 4% paraformaldehyde and 0.2% glutaraldehyde and incubated at room temperature for 15 min. Fixed cells were centrifuged at 6,000 × g for 2 min. The supernatant was removed, and cells were washed twice in Sorensen’s phosphate buffer. Cells were enrobed in melted, 60°C Noble agar (4% in Sorensen’s buffer [pH 7.4]) and cut into 2-mm blocks for dehydration.

Samples were dehydrated by treatment with the following steps: 50% ethanol for 15 min, 70% ethanol for 15 min, and 80% ethanol for 10 min. The sample at each step was incubated on a slow rotator set at a 45° angle at room temperature. Cells were incubated with a mixture of LR white resin (London Resin Company, Ltd.) and 80% ethanol (2:1, vol/vol) for 1 h at room temperature. Cells were then incubated with LR white resin (100%) three times for 1 h each. Each incubation was performed at room temperature on a slow rotator set at a 45° angle. Cells were embedded in LR white resin in gelatin capsules and incubated at 50°C for 14 to 16 h. Sections of 30 to 50 nm were cut using a DiaTOME Ultra diamond knife mounted on an RMC-MT-X microtome. Sections were placed onto gold grids coated with Formvar and carbon until immunogold labeling was performed.

Grids were incubated top-down on a 25-μL drop of blocking buffer containing 20 mM Tris (pH 7.4), 150 mM NaCl, and 1% (wt/vol) bovine serum albumin (BSA) for 10 min and then washed by repeated dipping in a beaker containing 10 mL blocking buffer for 1 min. Grids were incubated top-down on a 25-μL drop of blocking buffer containing the primary antibody (rabbit polyclonal antibody generated against S. Typhimurium EAL) diluted to 1:50 overnight at room temperature. Grids were washed by dipping in a beaker containing 10 mL blocking buffer for 1 min and then incubated top-down on a 25-μL drop of blocking buffer for 15 min. Grids were then transferred top-down to a 25-μL drop of goat anti-rabbit IgG-gold antibody (Millipore Sigma) and incubated for 1 h at room temperature. A final wash was performed by dipping repeatedly in a beaker containing 10 mL blocking buffer for 0.5 min and then dipping repeatedly in a beaker containing 10 mL water for 0.5 min. Where indicated, grids were poststained by incubating them on a droplet of filtered 4% aqueous uranyl acetate for 20 min, washed thoroughly by dipping them into a beaker of deionized (DI) H₂O 30 times, and then incubated on a drop of filtered DI H₂O for 1 min. Negative controls were treated exclusively with secondary antibody. Imaging was performed using a JEOL 1011 transmission electron microscope (JEOL USA, Peabody, MA).

The isolation and identification of cobamides were performed as described previously (32). Briefly, 100 mL of M9 minimal medium containing ethanolamine (90 mM) as the sole source of carbon and energy and cobinamide (100 nM) was inoculated with 2.5% (vol/vol) of an LB liquid culture, and the culture was grown overnight at 37°C with shaking at 180 rpm. The following morning, cultures were harvested using a 5810 R centrifuge (Eppendorf) at 3,220 × g for 20 min. The cell pellet was resuspended in 10 mL buffer A (50 mM KH₂PO₄ [pH 6.5], 5 mM KCN) with 20 mL methanol and then incubated at 65°C for 2 h with shaking (220 rpm). Cell debris was removed by centrifugation at 30,000 × g in an Avanti JP20-XVI centrifuge equipped with a JA-25.50 rotor for 1 h. A SepPak C₁₈ Plus cartridge (360 mg sorbent per cartridge, 55- to 105-μm particle size; Waters) was activated with methanol (10 mL) and water (10 mL). The supernatant was diluted 1:5 in water and passed over the activated SepPak cartridge. Cobamides bound to the SepPak cartridge were washed with water (20 mL) and eluted using methanol (100% [vol/vol]; 1.5 mL). Cobamides were dried overnight in a VacFuge Plus instrument (Eppendorf) and resuspended in buffer A (200 mL). KCN was added to a final concentration of 25 mM, and the mixture was incubated for 10 min in a sand bath kept at 90°C. Cobamides were then passed over a Spin-X centrifuge tube filter (Corning) for 5 min at 13,000 × g. Cleaned reaction mixtures were separated by RP-HPLC on a Shimadzu Prominence Ultra Fast Liquid Chromatography (UFLC) SPD-M30A instrument equipped with a Synergi 4μ Hydro-RP 80Å 150- by 4.6-mm column. The following method was used to separate cobamides. The column was equilibrated with 97% mobile phase A (MP A) (KH₂PO₄ [0.1 M] [pH 6.5], KCN [10 mM])–3% mobile phase B (MP B) (KH₂PO₄ [0.1 M] [pH 8], KCN [10 mM]-acetonitrile in a 1:1 ratio). Sample injection was followed by a 5-min linear gradient to 75% MP A–25% MP B, a 15-min linear gradient to 65% MP A–35% MP B, a 5-min isocratic step at 65% MP A–35% MP B, and a 2.5-min linear gradient to 100% MP B, concluding with a 5-min isocratic step at 100% MP B. Cobamides were identified by measuring the absorbance at 367 nm. Fractions containing cobamides were collected, desalted using a SepPak C₁₈ cartridge, and dried overnight in a VacFuge Plus instrument. Cobamide fractions were resuspended in sterile water and sent for MALDI-TOF mass spectrometry analysis.