Evidence that a Metabolic Microcompartment Contains and Recycles Private Cofactor Pools

Douglas L Huseby; John R Roth

doi:10.1128/JB.02179-12

. 2013 Jun;195(12):2864–2879. doi: 10.1128/JB.02179-12

Evidence that a Metabolic Microcompartment Contains and Recycles Private Cofactor Pools

Douglas L Huseby ^1,^*, John R Roth ^1,^✉

PMCID: PMC3697265 PMID: 23585538

Abstract

Microcompartments are loose protein cages that encapsulate enzymes for particular bacterial metabolic pathways. These structures are thought to retain and perhaps concentrate pools of small, uncharged intermediates that would otherwise diffuse from the cell. In Salmonella enterica, a microcompartment encloses enzymes for ethanolamine catabolism. The cage has been thought to retain the volatile intermediate acetaldehyde but allow diffusion of the much larger cofactors NAD and coenzyme A (CoA). Genetic tests support an alternative idea that the microcompartment contains and recycles private pools of the large cofactors NAD and CoA. Two central enzymes convert ethanolamine to acetaldehyde (EutBC) and then to acetyl-CoA (EutE). Two seemingly peripheral redundant enzymes encoded by the eut operon proved to be essential for ethanolamine utilization, when subjected to sufficiently stringent tests. These are EutD (acetyl-CoA to acetyl phosphate) and EutG (acetaldehyde to ethanol). Obligatory recycling of cofactors couples the three reactions and drives acetaldehyde consumption. Loss and toxic effects of acetaldehyde are minimized by accelerating its consumption. In a eutD mutant, acetyl-CoA cannot escape the compartment but is released by mutations that disrupt the structure. The model predicts that EutBC (ethanolamine-ammonia lyase) lies outside the compartment, using external coenzyme B₁₂ and injecting its product, acetaldehyde, into the lumen, where it is degraded by the EutE, EutD, and EutG enzymes using private pools of CoA and NAD. The compartment appears to allow free diffusion of the intermediates ethanol and acetyl-PO₄ but (to our great surprise) restricts diffusion of acetaldehyde.

INTRODUCTION

The lipid membrane of cells retains compounds that are either large (e.g., proteins or cofactors) or charged (e.g., phosphorylated or carboxylated metabolites). Small, uncharged molecules (e.g., acetaldehyde, CO₂, oxygen) diffuse freely across lipid membranes, making it difficult for a cell to either exclude or accumulate these compounds. We propose that microcompartments facilitate use of these small nonconfinable compounds by accelerating the metabolic pathways that consume them. Microcompartments consist of a protein shell that encapsulates enzymes of a single pathway. Three well-described metabolic pathways proceed in microcompartments: carbon fixation by ribulose bisphosphate carboxylase/oxygenase (RuBisCo) in photosynthetic bacteria (among others) and catabolism of 1,2-propanediol and ethanolamine (1–3) in enteric bacteria. These pathways include a small, uncharged intermediate or substrate: CO₂ for RuBisCo and propionaldehyde or acetaldehyde for degradation of 1,2-propanediol or ethanolamine (4, 5). The compartment is assumed to concentrate or conserve the small molecule but may also contribute by exposing that molecule to a high local concentration of relevant enzymes.

Some current models for these pathways propose that microcompartments physically retain the small, uncharged molecule, possibly at a high concentration, and restrict its exchange with the rest of the cell (4, 5). This retention model poses a mechanistic difficulty in the case of the ethanolamine and 1,2-propanediol pathways. These compartments would have to restrict passage of a small aldehyde but allow passage of the large cofactors coenzyme B₁₂, coenzyme A (CoA), and NAD⁺/NADH used by enzymes in the pathway. This discrimination would require a degree of complexity that has not been observed for these structures. We provide evidence that the large cofactors NAD and CoA do not pass freely but are retained and recycled within the compartment, where they exist as private cofactor pools used by enzymes of the ethanolamine (or 1,2-propanediol) pathways.

Ethanolamine is a source of energy, carbon, and nitrogen for the enteric bacterium Salmonella enterica (6). Enzymes to support this catabolism are encoded by the eut operon, and the functions of most of these proteins are known (2) (Fig. 1). The first reaction converts ethanolamine to acetaldehyde and ammonia and is catalyzed by the B₁₂-dependent ethanolamine-ammonia lyase EutBC (7). In the second reaction, the EutE enzyme catalyzes oxidation of acetaldehyde to acetyl-CoA, thereby consuming CoA and reducing NAD⁺ to NADH (7, 8). At this point, acetyl-CoA could be shunted directly into the tricarboxylic acid (TCA) cycle to serve as a carbon and energy source (see the horizontal reaction sequence in Fig. 1). However, the two other enzymes encoded by the eut operon (vertical arrows in Fig. 1) suggest that the ethanolamine pathway may be more complicated (2).

Fig 1 — Operon and pathway for ethanolamine catabolism. (A) Genetic map of the 17-gene *eut* operon. Black boxes, genes for catabolic enzymes; gray boxes, genes for microcompartment shell proteins. (B) The conventional pathway is the horizontal reaction sequence leading to acetyl-CoA. The vertical EutG and EutD reactions have been considered auxiliary, serving perhaps to support fermentative growth by balancing redox and providing a source of ATP when the housekeeping enzyme acetate kinase (AckA) converts acetyl-PO₄ to acetate plus ATP.

The EutG enzyme is an alcohol dehydrogenase that reduces acetaldehyde to ethanol while oxidizing NADH to NAD⁺ (4). This reaction makes no contribution to the conventional pathway but could recycle the electron carrier NADH during fermentation. The enzyme EutD is a phosphotransacetylase that catalyzes conversion of acetyl-CoA to acetyl-phosphate (9, 10), releasing free CoA. This reaction also makes no contribution to the standard pathway (horizontal pathway in Fig. 1) but could support fermentative growth by providing a source of ATP when acetyl-PO₄ is converted to acetate by the housekeeping enzyme AckA. Thus, the two peripheral activities might be thought of as supporting anaerobic fermentative growth, but thus far, fermentation has not been reported for ethanolamine as a sole source of carbon and energy (11).

Both EutG (alcohol dehydrogenase) and EutD (phosphotransacetylase) are redundant, in that each reaction is also catalyzed by a housekeeping enzyme activity: alcohol dehydrogenases (AdhE or AdhP) and phosphotransacetylase (Pta). The Pta activity contributes to assimilation of exogenous acetate and provides (with AckA) a source of ATP during acetogenic fermentations. It has been suggested that EutD phosphotransacylase is necessary because the Pta activity might be insufficient to provide the flux of carbon required for growth on ethanolamine (12). Despite their seeming redundancy, the EutD and EutG enzymes are required for both aerobic and anaerobic growth on ethanolamine. The inability of EutD and EutG mutants to grow on ethanolamine conflicts with some previous reports and will be explained below (2, 12).

Five proteins encoded in the eut operon, EutS, EutM, EutN, EutL, and EutK, form the shell of the microcompartment (2, 4, 13). The shell is thought to modulate the flow of metabolites to and from the enzymes contained in the compartment. Shell proteins may also play a direct role in packaging specific enzymes during microcompartment assembly (14, 15). On the basis of structural studies of the homologous 1,2-propanediol compartments, EutBC and EutE likely seem to be associated with the ethanolamine microcompartment, but the position of EutG or EutD remains uncertain. A lack of one or more microcompartment shell proteins disrupts microcompartment formation both in the eut system and in the analogous pdu system (13, 16).

This paper describes and supports a model for the ethanolamine pathway suggesting that the microcompartment includes at least three enzymes and retains private pools of their large cofactors, CoA and NAD. Recycling of cofactors within the compartment drives ethanolamine catabolism and couples the three reactions that process acetaldehyde. Small metabolites (e.g., ethanol and acetyl-PO₄) are proposed to diffuse freely out of the compartment, while acetaldehyde loss is impeded. Aspects of this model have been proposed previously, and evidence for NAD recycling within the propanediol microcompartment was described recently (15).

MATERIALS AND METHODS

Strains and crosses.

The bacterial strains used are listed in Table 1. All strains are derivatives of Salmonella enterica serovar Typhimurium strain LT2, referred to here as Salmonella. Transductional crosses were performed as previously described (17) utilizing the high-frequency transducing phage P22 (HT105 int) (18). The mutants with single eut gene deletion mutations used in this study were described previously (19). Mutants with deletions of multiple adjacent genes were constructed by linear transformation (20, 21) in which a small (80-bp) single-stranded primer displaced a counterselectable SacB-KanR cassette (22). In all constructed deletion mutants, the translational reading frame was maintained so as to avoid polar effects on transcription.

Table 1.

Strains used

Strain	Genotype	Source or reference
TR10000	Wild type	Laboratory collection
TT19189	eutDM302	Kofoid et al. (2)
TT22522	eutS367Δ	Penrod and Roth (4)
TT22523	eutM365Δ	Penrod and Roth (4)
TT22569	eutN366Δ::FRT^c	Penrod and Roth (4)
TT22570	eutL364Δ::FRT	Penrod and Roth (4)
TT22571	eutK363Δ::FRT	Penrod and Roth (4)
TT22815	eutE356Δ::FRT	Penrod and Roth (4)
TT23036	eutG361Δ	Penrod and Roth (4)
TT24801	eutP369Δ::FRT	Penrod and Roth (4)
TT24802	eutQ370Δ::FRT	Penrod and Roth (4)
TT24803	eutT371Δ::FRT	Penrod and Roth (4)
TT24804	eutD372Δ::FRT	Penrod and Roth (4)
TT24805	eutA373Δ::FRT	Penrod and Roth (4)
TT24806	eutB374Δ::FRT	Penrod and Roth (4)
TT24807	eutC375Δ::FRT	Penrod and Roth (4)
TT26679	eutMN384Δ eutLK385Δ	This work
TT26680	eutS367Δ eutMN384Δ eutLK385Δ	This work
TT26685	eutD372Δ::FRT eutQ386 (P32T)^a	This work
TT26686	eutQTDMN387Δ	This work
TT26687	eutQTDMNE388Δ	This work
TT26688	eutD372Δ::FRT eutM389 (G5R)^a	This work
TT26689	eutD372Δ::FRT eutM390 (T9P)^a	This work
TT26690	eutD372Δ::FRT eutM391 (T9N)^a	This work
TT26691	eutD372Δ::FRT eutM392 (deletion of bp 80–94)^b	This work
TT26692	eutD372Δ::FRT eutM393 (deletion of bp 81–92)^b	This work
TT26693	eutD372Δ::FRT eutM394 (duplicate bp 117–121)^b	This work
TT26694	eutD372Δ::FRT eutM395 (deletion of bp 137–144)^b	This work
TT26695	eutD372Δ::FRT eutM396 (D48Y)^a	This work
TT26696	eutD372Δ::FRT eutM397 (A51V)^a	This work
TT26697	eutD372Δ::FRT eutM398 (deletion of bp 176–183)^b	This work
TT26698	eutD372Δ::FRT eutM399 (Δ191, G192C, C193A, T198C)^b	This work
TT26699	eutD372Δ::FRT eutM400 (G250T)^b	This work
TT26700	eutD372Δ::FRT eutE401 (E7K)^a	This work
TT26701	eutD372Δ::FRT eutE402 (L13Q)^a	This work
TT26702	eutS367Δ eutD372Δ::FRT	This work
TT26703	eutP369Δ::FRT eutD372Δ::FRT	This work
TT26704	eutQ370Δ::FRT eutD372Δ::FRT	This work
TT26705	eutG361Δ eutD372Δ::FRT	This work
TT26706	eutJ362Δ eutD372Δ::FRT	This work
TT26707	eutL364Δ::FRT eutD372Δ::FRT	This work
TT26708	eutK363Δ::FRT eutD372Δ::FRT	This work
TT26709	pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)^d	This work
TT26710	eutS367Δ eutMN384Δ eutLK385Δ pta331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26711	eutMN384Δ eutLK385Δ pta331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26712	eutS367Δ pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26713	eutM365Δ pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26714	eutN366Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26715	eutL364Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26716	eutK363Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26717	eutA373Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26718	eutB374Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26719	eutC375Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26720	eutDΔ pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26721	eutE356Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26722	eutG361Δ pta-331Δ::Tet(sw) acs-103ΔΔ::Cam(sw)	This work
TT26723	eutP369Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26724	eutQ370Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26725	eutT371Δ::FRT pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26726	eutJ362Δ pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26727	eutDM302Δ pta-331Δ::Tet(sw) acs-103Δ::Cam(sw)	This work
TT26728	eutJ362Δ	Penrod and Roth (4)
TT26729	eutM365Δ eutG361Δ	This work
TT26730	eutMNE403Δ	This work
TT26731	eutM365Δ pta-331Δ::Tet(sw)	This work
TT26737	eutM365Δ eutG361Δ adhP1::Cam(sw)	This work
TT26738	eutS367Δ eutMN384Δ eutLK385Δ adhP1::Cam(sw)	This work
TT26869	eut-38::MudA	Roof and Roth (25)
TT26870	eutD372Δ::FRT eut-38::MudA	This work
TT26871	eutD372 Δ::FRT eutQ386 eut-38::MudA	This work
TT26872	eutD372Δ::FRT eutE401 eut-38::MudA	This work
TT26873	eutD372Δ::FRT eutM389 eut-38::MudA	This work
TT26874	eutD372Δ::FRT eutM395 eut-38::MudA	This work
TT26875	eutD372Δ::FRT eutM399 eut-38::MudA	This work
TT26876	eutD372Δ::FRT eutM400 eut-38::MudA	This work
TT26877	eutQTDMNE388 eut-38::MudA	This work

Open in a new tab

Protein sequence change.

Nucleotide sequence change.

FRT, FLP recombination target.

(sw), swap, indicating that a drug resistance cassette replaces coding sequence of target gene.

Multiply mutant strains were made by transduction crosses that created combinations of nonpolar deletion mutations in the chromosomal eut operon. Placement of all mutations in the chromosome complicates the process of strain construction but conserves the ratio of proteins that contribute to the multiprotein microcompartment. Individual genes cloned in plasmids were found to give anomalous phenotypes compared to those of strains with all genes in the operon. In constructing the original deletions and in the transduction crosses used to construct multiply mutant strains, no selection for growth on ethanolamine was imposed. This was done to avoid unintentional selection of suppressor mutations, which are extremely frequent.

Chemicals and growth media.

The rich medium used in this study was LB (lysogeny broth, or Luria-Bertani medium). Unless otherwise noted, the minimal medium used was no-carbon E medium (NCE) (17, 23) supplemented with 150 nM cyanocobalamin (CN-B₁₂; Sigma), 4 mM Tricine (USB), 50 μM FeSO₄ (Sigma), and trace minerals, as previously described (11). Unless otherwise noted, glycerol (Mallinckrodt) was used at 0.2% (wt/vol), ethanolamine hydrochloride (Sigma) was used at 0.3%, sodium acetate (Sigma) was used at 0.5%, ethanol (Omnipur grade; EMD) was used at 0.25%, and ethylenediamine (Sigma) was used at 10 mM. For anaerobic cultures, medium was also supplemented with 10 mM sodium bicarbonate (Sigma), and the anaerobic electron acceptor sodium tetrathionate (Sigma) was used at 20 mM. The growth medium used for the selection of suppressors was that of Neidhardt et al. (24); for the pH 8.0 version of this medium, the 3-(N-morpholino)propanesulfonic acid (MOPS) buffer was replaced with 2-[bis(2-hydroxyethyl)amino]acetic acid (Bicine) with the pH adjusted to 8.0.

General growth conditions.

Cells were grown overnight in LB. For aerobic LB cultures, overnight cultures were used to inoculate fresh LB at 1%, and these new cultures were grown for 5 h. The cells were then pelleted, rinsed, and resuspended in appropriate minimal medium. For anaerobic cultures, the washed minimal medium-grown cells were used to inoculate minimal medium with glycerol, ethanolamine, and CN-B₁₂. These cultures were then grown for 5 h. The cells were then pelleted and repeatedly resuspended in minimal medium appropriate for the specific assay. The doubling times presented in Tables 2 and 4 were calculated from growth in 96-well plates in a Bio-Tek Synergy HT plate reader. Cultures were handled exactly as described above but were scaled down to 200 μl. Anaerobic cultures were grown in capped tubes. Tubes were filled almost completely with medium, sealed with a rubber stopper lubricated with vacuum grease, and then crimp capped. Care was taken to minimize the aerobic gas headspace. For anaerobic incubations without tetrathionate, the inoculum size was 40%. This large inoculum size was used to highlight ethanolamine consumption without corresponding growth of the culture. All growth measurements were made at 37°C with vigorous agitation of the tubes.

Table 2.

Suppressors of eutD

Genotype	DNA sequence change	Effect on protein	Doubling time on ethanolamine (h/generation)	Relative fitness^a
Wild type (TR10000)	No change (eut⁺)		3.9	1.00
eutD372	In-frame deletion of eut-coding sequence	No EutD protein	>16	<0.24
Extended deletions that restore growth on ethanolamine to eutD deletion mutant
eutQTDMN387	Deletes eutQ bp 471 to 8 bp before eutE; removes eutTDMN	Truncated EutQ, no EutTDMN; wild-type EutE	3.4	1.14
eutQTDMNE388	Deletes eutQ bp 343 to eutE bp 51; removes eutTDMN	Fuses EutQ to EutE (−17AA); EutE⁺	3.9	0.99
Single gene mutations that restore growth on ethanolamine to a eutD deletion mutant
eutQ386	C94A	P32T substitution	4.6	0.84
eutM389	G13A	G5R substitution	3.7	1.05
eutM390	A25C	T9P substitution	4.0	0.97
eutM391	C26A	T9N substitution	4.0	0.97
eutM392	Deletion of bp 80–94 (15 bp), 9-bp join point	Internal deletion of 5 amino acids	3.7	1.05
eutM393	Deletion of bp 81–92 (12 bp), 2-bp join point	Internal deletion of 4 amino acids	3.5	1.09
eutM394	Duplication of bp 117–121, 5-bp insertion	Frameshift, premature termination	3.8	1.00
eutM395	Deletion of bp 137–144 (8 bp), 0-bp join point	Deletion, frameshift, premature termination	3.6	1.06
eutM396	G142T	D48Y substitution	3.6	1.06
eutM397	C152T	A51V substitution	4.3	0.90
eutM398	Deletion of bp 176–183 (8 bp), 3-bp join point	Deletion, frameshift, premature termination	3.6	1.06
eutM399	Deletion of bp 191; G192C, C193A, T198C	Frameshift, premature termination	3.6	1.07
eutM400	G250T	Nonsense mutation. premature termination	3.5	1.02
eutE401	G19A	E7K substitution in signal sequence	3.6	1.07
eutE402	T38A	L13Q substitution in signal sequence	3.3	1.17

Open in a new tab

Fitness expressed as the ratio of the wild-type doubling time to that of the mutants.

Table 4.

Suppression of the EutD phenotype by constructed deletion mutations

Genotype^a	Strain	Doubling time on ethanolamine (h/generation)	Relative fitness^b
Wild type	TR10000	3.1	1.00
eutD	TT24804	>16	<0.19
eutDM	TT19189	2.9	1.06
eutD eutL	TT26707	4.1	0.76
eutD eutK	TT26708	4.7	0.67
eutD eutQ	TT26704	4.0	0.77
eutD eutS	TT26702	5.4	0.58
eutD eutP	TT26703	6.5	0.48
eutD eutG	TT26705	8.5	0.37
eutD eutJ	TT26706	9.5	0.33

Open in a new tab

Each mutation listed is an in-frame deletion of the entire coding sequence.

Fitness expressed as the ratio of the wild-type doubling time to that of the mutants; values above 0.5 are considered a significant improvement.

Growth of a eutD mutant.

Three independent cultures were grown overnight in LB. Overnight cultures were used to inoculate fresh LB at a 1:100 dilution. The new cultures were grown for 5 h. These cultures were either used directly to inoculate ethanolamine minimal medium cultures in the manner of Starai et al. (12) or pelleted and resuspended in minimal medium with no added carbon source prior to use as the inoculum. Multiple rinse cycles were found to make no difference. Unrinsed or rinsed cells were used to inoculate at 1% 10-ml minimal medium cultures supplemented with ethanolamine and CN-B₁₂.

Selecting Eut⁺ suppressors of a eutD mutant.

Revertants (Eut⁺) of an in-frame eutD deletion mutant (Eut⁻) were selected by plating on minimal medium with ethanolamine as the sole carbon source. Prior to plating, cells of the eutD strain TT24804 were grown for ∼16 h in LB medium. Cells were pelleted and resuspended in the medium of Neidhardt et al. (24) with no carbon source. Washed cells (10⁶, 10⁷, or 10⁸) were plated on the same medium (at either pH 7.0 or pH 8.0) with added 0.5% ethanolamine and 150 nM B₁₂ solidified with 3% Noble agar. Eut⁺ colonies began appearing 5 days after plating, and isolated colonies were picked on days 5 and 6.

The isolated revertants were confirmed to be phenotypically Eut⁺ by patching and printing to minimal ethanolamine medium. The mutations conferring the Eut⁺ phenotype were found to be linked to the eut operon region by P22 transductional crosses (linkage to cysA). Sequencing of selected eut operon regions identified the mutations responsible for the suppression phenotype. No systematic variation in the nature of the suppressor mutations isolated was observed under any of the various selection conditions tested (pH, number of cells plated).

β-Galactosidase assays.

Assays for expression of the eut operon used the eut-38::Mud(Lac) reporter system, and β-galactosidase was assayed as previously described (25–27). Four replicate cultures were grown overnight in LB medium and then used to inoculate 1% NCE with glycerol and CN-B₁₂ and with or without ethanolamine. These cultures were grown for 6 h and then assayed as described above. Miller units were calculated by continuous monitoring of product formation using a Bio-Tek Synergy HT plate reader.

Use of NMR to assay metabolite concentrations in growth media.

Samples (0.5 ml) taken from cultures at various times were assayed for cell density (optical density at 600 nm [OD₆₀₀]) and were analyzed by ¹H nuclear magnetic resonance (NMR) for the presence of metabolites in the medium. Samples of anaerobic cultures were drawn through the rubber stopper using a 1-ml syringe fitted with a needle. Small amounts of atmospheric gas were likely drawn into the sealed tubes during sampling, but the introduced gas was confirmed to not support significant growth by drawing samples from tubes to which the external electron acceptor tetrathionate was not added and monitoring the subsequent lack of growth. Samples were stored frozen at −20°C until analysis.

Thin-wall 300-MHz Wilmad NMR tubes (5-mm diameter, 7 in. long) were loaded with 450 μl of sample, 4.5 μl of 1 M sodium benzoate, and 50 μl of 99.96% deuterium oxide. Measurements were taken on a Bruker Avance-500 spectrometer at the UC-Davis NMR core facility. Residual H₂O was suppressed with continuous wave presaturation during the delay using a 50-Hz-effective-field-strength presaturating field. Relevant experimental parameters were as follows: temperature, 298 K; number of dummy scans, 4; number of experimental scans, 16; relaxation delay, 5 s; acquisition time, 2.73 s; and number of data points collected for each assay, 32,768. Collected spectra were Fourier transformed and baseline straightened using the Bruker fifth-order linear correction program (absf).

Peaks were picked by hand, as were regions for integration. Chemical shifts are reported relative to 3-(trimethylsilyl)-1-propanesulfonic acid (residual HOD = 4.8 ppm). Measurements were collected for solutions of known single concentrations of individual analytes in growth medium to confirm the specific peaks that corresponded to each analyte. Standard curves to ensure a linear relationship between integrated peaks and concentration were not constructed. Consistent sampling conditions (pulse width, delays, temperature) were employed throughout the described experiments, and the probe was tuned to the proper resonant frequency and impedance for every sample. These experimental conditions should ensure that the derived concentration data closely describe the actual concentration of the analytes in the growth media.

The absolute concentration of Tricine was quantified using the peaks from benzoate: a doublet at ∼7.75 δ (2 ¹H) and triplets at ∼7.45 δ (1 ¹H) and ∼7.35 δ (2 ¹H). The concentration of Tricine, as determined by singlet peaks at ∼3.55 δ (2 ¹H) and ∼3.65 δ (6 ¹H), was then used to quantify the relative concentrations of the other analytes.

The ethanolamine concentration was determined by averaging the readings of triplets at ∼3.7 δ and 3.05 δ. Ethanol was determined using a triplet peak at ∼1.05 δ, and acetate was determined from a singlet peak at ∼1.8 δ. Total acetaldehyde species were determined by adding together doublet peaks corresponding to acetaldehyde (∼2.15 δ), 1,1-ethanediol (∼1.4 δ), and other acetaldehyde species (possibly paraldehyde; ∼1.2 to 1.4 δ).

Growth on acetate and ethanol.

Three independent cultures were grown overnight in LB and used to inoculate minimal medium with glycerol and ethanolamine. The inoculum size was 1%, and the cultures were allowed to grow for 5 h at 37°C. After 5 h, 1-ml samples of the cultures were pelleted in a tabletop centrifuge, rinsed, and resuspended in minimal medium with no carbon source.

For acetate growth assays, the resuspended cells were used to inoculate a 10-ml culture of minimal medium supplemented with ethylenediamine and/or sodium acetate. The inoculum size was 1%. For ethanol growth assays, the resuspended cells were used to inoculate a 10-ml culture of minimal medium supplemented with ethylenediamine and/or ethanol. The inoculum size was 10%.

RESULTS

An unexpected phenotype of eutD mutants.

The EutD enzyme (phosphotransacetylase) catalyzes one of the side reactions described in Fig. 1. Because this reaction is not central to the conventional view of the pathway, it was surprising that EutD was required for growth on ethanolamine under all conditions tested. Two conflicting phenotypes have previously been reported for eutD mutants. Initially, it was observed that eutD mutants fail to use ethanolamine as an aerobic source of carbon and energy (6, 7). In subsequent tests, constructed eutD mutations caused only a slight defect in ethanolamine use under the conditions used (12). Below, we confirm the latter observations but provide evidence that EutD is essential for long-term growth on ethanolamine in an otherwise wild-type cell.

Escalante-Semerena and coworkers observed a minor growth defect of eutD mutants when they diluted cells grown in rich medium (LB) 100-fold into minimal medium with ethanolamine as a sole carbon and energy source (12). Tested in this way, wild-type Salmonella grows to full density with a doubling time of 3.1 h and a eutD mutant grows to a similar density with a doubling time of 6.3 h. We confirmed this observation (Fig. 2A).

A different result was obtained when LB-grown cells were washed prior to inoculation of minimal ethanolamine medium (Fig. 2B). Wild-type Salmonella grew as previously observed, albeit with a longer lag time and a slightly slower doubling time (4.5 h). The eutD mutant, however, showed very little growth over the course of a 48-hour experiment.

The importance of EutD for growth on ethanolamine was also seen when single cells of a eutD mutant were plated on minimal ethanolamine plates. Wild-type cells formed healthy colonies within 2 days, while eutD mutant cells formed extremely tiny colonies that first appeared after 5 days of incubation. The trace of growth seen for eutD mutants was due to nutrients contaminating the agar rather than slow growth on ethanolamine. Colonies like those formed by eutD mutants were also observed for a deletion mutant lacking the entire eut operon and for wild-type cells plated on minimal medium lacking any added carbon source (data not shown).

Apparently, traces of LB medium transferred into minimal ethanolamine medium with the inoculum allowed eutD mutants to divide in a manner roughly similar to wild-type growth. This growth was not seen when eutD cells were pregrown on minimal medium with either glucose or glycerol as the carbon source. Neither the dilute LB transferred with the inoculum nor ethanolamine alone in minimal medium was sufficient to allow eutD mutants to grow, but together they allowed wild-type-like growth for at least 5 generations. This Eut⁻ phenotype was surprising since it was expected that the acetyl-CoA produced by the EutBC and EutE activities could be used directly in the TCA cycle, especially under aerobic conditions. The conclusion that EutD activity is required for growth on ethanolamine was confirmed by the reversion experiments described below.

Selecting suppressors of eutD mutations.

To elucidate the role of the EutD enzyme, Eut⁺ revertants of a eutD mutant were selected and analyzed. A strain with a constructed eutD deletion mutation (nonpolar) was grown on LB, and washed cells were plated on minimal ethanolamine medium containing B₁₂. Lawn growth was minimal, and revertant (Eut⁺) colonies accumulated over the course of 6 days. Nineteen unique revertants were analyzed, and all proved to carry a suppressor mutation genetically linked to the eut operon in P22-mediated transduction crosses. Sequencing the eut operon of these mutant strains identified the suppressor mutations responsible for allowing growth.

As seen in Table 2, most of the spontaneous eutD suppressor mutations affected either eutQ or eutM, two genes near eutD (Fig. 1 and Table 2). Two suppressor mutations had removed these genes as part of a large deletion that removed the entire eutQ, eutT, eutD, eutM, and eutN region. One of these deletions fused the N-terminal 114 amino acids of eutQ to the bulk of the eutE reading frame (without its first 17 codons). One eutM mutant had multiple base substitutions that together converted an imperfectly palindromic run in the wild-type sequence to a perfect palindrome, probably by mismatch correction, as previously described (28). Growth was restored to the eutD mutant by two eutE point mutations (E7K and L13Q) that cause amino acid substitutions in the N-terminal end of EutE. These changes seem likely to affect packaging of EutE into the compartment, since analogous positions E7, I10, and L14 contribute to packaging of PduP into microcompartments in the 1,2-propanediol pathway of Salmonella (15). These mutations do not destroy the catalytic function of EutE, since the mutants were able to grow on ethanolamine.

The finding of many suppressors in the eutM gene suggested the involvement of a microcompartment since EutM is one of the five shell proteins. It was curious that suppressor mutations were not found in the other four shell protein genes. If suppressors act by destroying the compartment, one might have expected that mutations in any of the shell protein genes could serve to suppress. No function is known for EutQ. The proximity of the eutQ and eutM genes to eutD suggested the possibility that the Eut⁻ phenotype of the strain with the eutD mutation might be due to transcriptional polar effects on expression of distal genes (i.e., eutE [acetaldehyde dehydrogenase] and eutBC [ethanolamine-ammonia lyase]) that were relieved by the suppressor mutations. This possibility seemed unlikely, in that the suppressed eutD mutation was a constructed in-frame deletion and many of the suppressors were missense mutations, which were not expected to be polar. However, this possibility was tested and eliminated by measuring the effects of the eutD deletion and its suppressors on expression of downstream genes.

Evidence that the eutD mutation and its suppressors do not affect expression of distal genes in the operon.

The possibility that the strain with the eutD deletion originally constructed owed its growth phenotype to a polar effect on expression of eutE and eutBC was tested using a lac reporter gene (eut-38::MudA) inserted at the promoter-distal end of the operon (25, 29). This insertion was outside all coding sequence and immediately adjacent to the transcription terminator and has been used extensively in characterizing eut operon transcription (25). This reporter fusion was added to the parent eutD deletion mutant and a series of its derived suppressor strains (Table 3).

Table 3.

Effect of eut mutations on expression of distal genes in operon

Genotype^a	Strain	Mutation type	β-Galactosidase expression from eut-38::Mud(Lac)^b	Ratio of mutant to wild-type activity
Control	TT26869	No mutation added	277 ± 10	1.00
eutD	TT26870	In-frame deletion of eutD-coding sequence	291 ± 17	1.05
eutQTDMN	TT26877	Truncated EutQ, no EutTDMN; EutE⁺	316 ± 15	1.14
eutD eutQ	TT26871	eutQ P32T missense	271 ± 16	0.98
eutD eutM	TT26873	eutM G5R missense	259 ± 14	0.94
eutD eutM	TT26874	eutM deletion, frameshift, termination	263 ± 25	0.95
eutD eutM	TT26875	eutM frameshift, termination	261 ± 12	0.94
eutD eutM	TT26876	eutM E84STOP nonsense, termination	263 ± 11	0.95
eutD eutE	TT26872	eutE E7K missense	281 ± 23	1.01

Open in a new tab

All strains had eut-38::Mud(Lac). The eut-38::Mud insertion fuses the lacZ gene to the eut operon at a point distal to all eut-coding sequences.

Cells were grown on minimal glycerol ethanolamine medium. β-Galactosidase was assayed as described in Materials and Methods, and β-galactosidase expression is expressed in units described by Miller (26).

Neither the initial eutD deletion nor any of the added suppressor mutations showed any polar effect on expression of downstream genes in the operon. This was true, even though some of the suppressors created translation stops that might have had a polar effect. Apparently, the shortness of the affected genes and the proximity of the suppressor mutation to the initiation signals of the next gene prevent polar effects on transcription.

Effects of constructed deletion mutations on eutD phenotype.

The spontaneously selected suppressor mutations were difficult to interpret in terms of function. While eutM mutations suggested that disruption of the microcompartment might suppress, this did not explain why suppressors were not found in the other shell protein genes. To pursue this, a series of eut deletion mutations was constructed and added individually to the parental eutD deletion mutant. The growth of these strains on ethanolamine is presented in Table 4, where strains are ordered by the growth improvement provided by the added mutation.

As seen in Table 4, the lack of any of 3 compartment shell proteins (eutM, eutL, or eutK) corrected the growth defect of a eutD mutant on ethanolamine, with the eutL and eutK mutations providing weaker suppression. (The contribution of EutN was not tested due to difficulties in making a eutN eutD double-deletion mutant.) We arbitrarily defined any relative fitness above 0.5 to be significant suppression; this was done because of the variability of growth rates and because the parent eutD deletion mutant showed detectable growth. Removal of the eutS shell protein had very little effect on fitness. The weaker effects of shell protein mutations other than eutM mutations may explain why such mutants were not recovered among spontaneous suppressors. These weaker suppressive effects of some mutations may suggest that the remaining shell components can produce compartment subassemblies that have residual function in binding enzymes. Alternatively, abnormal subassemblies may impair growth on ethanolamine for reasons apart from compartment integrity. (All of the strains in Table 4 had equivalent fitness when tested for growth on glycerol.) Essentially no suppression was caused by removal of the unknown proteins EutP and EutJ or the alcohol dehydrogenase EutG. Other proteins could not be tested for suppression because they are essential for growth on ethanolamine: EutBC (ethanolamine-ammonia lyase) and EutE (aldehyde dehydrogenase).

Compounds excreted by cells during growth on ethanolamine.

Wild-type cells growing on ethanolamine excrete some acetaldehyde, and more is released by strains defective for microcompartment shell proteins (4). Acetate excretion and subsequent reassimilation were previously reported for cells growing aerobically on ethanolamine (12). The experiments described below demonstrate that aerobic cultures excreted and reassimilated acetate, while anaerobic cultures showed only excretion, even when tested with the alternative electron acceptor tetrathionate. Compounds appearing in the medium during growth on ethanolamine were detected by ¹H NMR. Figure 3A describes a wild-type culture growing aerobically on ethanolamine (∼30 mM). The rapid disappearance of ethanolamine was accompanied by accumulation of acetate, acetaldehyde, and ethanol. Acetate accumulates to a maximum, the acetate switch point (30), at which cells shift from net excretion to reassimilation of acetate. As seen in Fig. 3A, prior to the switch, aerobic cultures accumulated nearly equimolar amounts of ethanol and acetate, as expected, as reactions within the compartment are coupled. The slight excess of excreted ethanol over acetate may reflect consumption of some acetyl-PO₄ for use as cell carbon. The disappearance of acetaldehyde and ethanol from the medium suggests reassimilation but could have been due in part to loss from the culture as gases.

Cultures grown under anaerobic conditions with tetrathionate as an electron acceptor did not show the acetate switch, probably because of their inability to reassimilate the excreted acetate or the other excreted compounds (Fig. 3B). This failure to use acetate was also seen in direct tests and probably reflects the high cost (2 high-energy PO₄ molecules) of converting assimilated acetate to acetyl-PO₄. An earlier report of anaerobic acetate used a higher concentration of tetrathionate and may have reflected in part an OD accumulation due to precipitated sulfur during reduction of tetrathionate (11).

Under anaerobic conditions with no electron acceptor (O₂, tetrathionate, or otherwise) (Fig. 3C), cells could not grow and seemed to be unable to ferment ethanolamine. However, despite their lack of growth, cells degraded ethanolamine gratuitously. These nongrowing cells showed parallel excretion of ethanol and acetate but no excess of ethanol, supporting the idea that the excess seen aerobically reflects use of some acetyl-PO₄ or acetate as a carbon source.

The production and possible consumption of ethanol during aerobic growth (Fig. 3A) were unexpected, since the enzymes that produce ethanol (Adh) in Salmonella are generally expressed only anaerobically (31, 32). During anaerobic growth, ethanol is excreted to dispose of excess electrons accumulated in the form of NADH from oxidative degradation processes. Under aerobic conditions, these electrons are consumed by respiration, generally obviating the need to excrete metabolites in order to achieve redox balance. The production of ethanol during both aerobic and anaerobic growth on ethanolamine suggested a nonstandard form of metabolism. To explain these results and the phenotypes of eutD mutants and their suppressors described above, a model for the ethanolamine catabolic pathway is proposed below.

Model for ethanolamine catabolism.

The model diagrammed in Fig. 4 avoids the previous paradox of a compartment that allows diffusion of large cofactors and yet restricts movement of volatile aldehyde. In this model, the compartment restricts the movement of large cofactors and allows diffusion of at least some small metabolites. The microcompartment is proposed to drive the catabolic pathway in an unconventional way. That is, it may not increase the internal concentration of acetaldehyde, which is difficult to confine, but achieves the same end by injecting the problematic metabolite into a very reactive environment with local high concentrations of enzymes and cofactors. In this model, acetyl-CoA exists in two pools, with one within the compartment serving only as a substrate for EutD. The other pool is outside the compartment and can serve as the substrate for the housekeeping enzyme Pta, leading to acetyl-PO₄ and (with Ack) to acetate and ATP, or alternatively, the external pool can be used by citrate synthase (GltA; leading to the TCA cycle). The recycling of cofactors enforces a redox balance within the compartment that is reflected in the excreted products (ethanol and acetate) measured by NMR. The balance is still apparent under aerobic conditions because so little TCA function is required to satisfy the requirements of biosynthesis and generation of a proton motive force.

The model explains the Eut⁻ phenotype of a eutD mutant and its suppression by microcompartment defects. It was previously expected that a eutD mutant would convert ethanolamine to acetyl-CoA (using EutBC and EutE), which could enter the TCA cycle and allow growth. The observed inability of eutD mutants to use ethanolamine is now explained because acetyl-CoA cannot exit the microcompartment but accumulates and traps all the microcompartment internal CoA. In the absence of EutD, the EutE reaction is secondarily blocked by a lack of free CoA. The model proposes that disruption of the compartment suppresses eutD mutants by allowing acetyl-CoA to reach the external Pta and TCA cycle enzymes and provide cells with energy and carbon.

The model explains the Eut⁻ phenotype of eutG mutants. This phenotype (like that of the eutD mutants described above) is subtle and was missed in initial tests. However, single cells of a eutG mutant are unable to form colonies on ethanolamine, and this defect is suppressed by mutations in genes for microcompartment shell proteins (D. L. Huseby and J. R. Roth, unpublished data). These results will be described in more detail elsewhere but are explained by the model presented here. Alcohol dehydrogenase (EutG) is peripheral to the pathway, as initially imagined, but in the light of the model becomes central to recycling of NAD⁺ within the microcompartment. The failure of a EutG mutant to use ethanolamine reflects the fact that all three reactions (EutG, EutE, and EutD) stop when NADH cannot be recycled to NAD⁺ within the microcompartment. Without EutG, reduced NADH accumulates, blocking further EutE activity, and without EutE, CoA accumulates in its free form (Fig. 4).

The model explains the release of ethanol observed during growth on ethanolamine under both aerobic and anaerobic conditions. The coupling of EutG and EutE reactions (by NAD recycling) enforces a strict 1:1 relationship between ethanol (produced by EutG) and products derived from the acetyl-CoA (acetate and cell carbon). This relationship is enforced by the strict coupling of the EutE and EutG reactions to one another via the NAD⁺/NADH pools inside the microcompartment. That is, for every 2 molecules of acetaldehyde generated from ethanolamine by EutBC, one molecule must be converted to ethanol and the other must be converted to acetyl-CoA and thence to acetyl-PO₄, which is released for use by housekeeping enzymes.

One speculative aspect of the model described in Fig. 4 is the location of the EutBC enzyme outside or possibly on the surface of the compartment. This positioning is not tested here but seems reasonable because the EutBC cofactor adenosyl-cobalamin (Ado-B₁₂) is subject to radical damage and must be replaced frequently. The needed Ado-B₁₂ can be made by housekeeping adenosyltransferase (CobA) as well as the operon-specific EutT enzyme (33). If EutT were inside the compartment, repair would require transport of Ado-B₁₂ (molecular weight, 1,500) across a compartment wall that is proposed to block diffusion of NAD, CoA, and acetaldehyde.

The model proposes that the ethanolamine pathway is poised to support fermentation. The acetyl-PO₄ produced by the coupled compartment reactions diffuses freely out of the compartment. Most is then converted to acetate by the AckA enzyme with coproduction of ATP. The bias toward acetate explains the near equivalence of excreted ethanol and acetate. The slight excess of ethanol over acetate reflects the small fraction of acetyl-PO₄ converted back to acetyl-CoA by Pta outside the compartment and directed into the TCA cycle. The small fraction of acetyl-PO₄ that enters the TCA cycle can produce carbon precursors and generate some reducing power for biosynthesis.

Testing predictions of the model.

In the model described above, the acetyl-CoA produced within the compartment is not available for housekeeping enzymes outside the compartment (e.g., GltA, Pta) but becomes available when the compartment is disrupted by a mutation in one of the shell proteins. Thus, the microcompartment serves as a barrier than can be genetically removed to test aspects of the model. The tests described below examine the ability of several Eut enzymes to contribute to growth on acetate and ethanol with and without the barrier in place.

Growth phenotypes on ethanolamine.

In principle, the effect of shell protein mutants on ethanolamine catabolism should test aspects of the model. In practice, this direct test has been difficult. The only ethanolamine growth phenotype for shell protein mutants is a failure to grow on ethanolamine plates at high pH (8.0) (4). These laboratory growth tests require concentrations of ethanolamine (10 to 40 mM) much higher than those likely to be encountered naturally, because the growth yield on ethanolamine is so low. It is difficult to measure growth on low ethanolamine concentrations, because the substrate is consumed before measureable growth can occur.

Despite these difficulties, the ethanolamine growth phenotypes for two types of eut mutants support the model described here. As described above, both the eutD and eutG mutants failed to grow on ethanolamine (even at a high concentration) when tested under sufficiently stringent conditions. The model predicts that the EutD and EutG reactions are coupled to the central EutE activity because of obligatory recycling of their cofactors, NAD or CoA. Thus, elimination of any single activity should block all of them, as long as the compartment is intact. Consistent with the model, mutations that disrupt the compartment (and thus destroy the coupling) suppress the Eut⁻ phenotypes of mutants lacking either of the peripheral enzymes, EutD or EutG.

Normally, the ethanolamine pathway of Salmonella probably operates at low levels of ethanolamine. In a host gut, cells are exposed to a low but continuous supply of host-derived ethanolamine (34). The Salmonella ethanolamine diffusion facilitator (EutH) contributes to growth only at ethanolamine concentrations below 25 μM (19). Because it is difficult to assess growth phenotypes at such low natural concentrations, we have resorted to devious ways to test the function of the compartment proposed by the model.

Growth phenotypes on acetate.

In the test for growth phenotypes on acetate, the EutD enzyme substitutes for the housekeeping enzyme Pta in degrading exogenous acetate. The standard acetate pathways are diagrammed in Fig. 5A. During anaerobic fermentation of sugars, cells produce acetate (and ATP) from acetyl-CoA by the sequential activity of Pta (phosphotransacylase) and AckA (acetate kinase) (35–37). On high concentrations of external acetate, cells can reverse this pathway (Fig. 5A, left). On low concentrations of acetate, cells use the Acs enzyme (acetyl-CoA synthetase), which converts ATP to AMP and thus consumes two high-energy phosphates to produce acetyl-CoA (37, 38).

Fig 5 — Pathways for assimilation of exogenous acetate. (A) Standard pathways for assimilating exogenous acetate; (B) pathway for the test strain used to examine microcompartment function. The housekeeping enzymes Pta and Acs were genetically removed, leaving a strain that possesses AckA but lacks a complete route from acetate to acetyl-CoA. Induction of the ethanolamine operon leads to production of the EutD protein but does not allow growth on acetate (left line from acetyl-PO₄), unless the compartment is disrupted by mutations that remove shell proteins (right arrow from acetyl-PO₄).

The ethanolamine model predicts that during growth on acetate, the EutD (phosphotransacetylase) should substitute for the housekeeping enzyme Pta and convert the acetyl-PO₄ produced by AckA to acetyl-CoA (diagrammed in Fig. 5B). This should be possible only if the eut operon is induced and the microcompartment is disrupted so that the acetyl-PO₄ produced by AckA has access to EutD. Without this disruption, acetyl-PO₄ (produced by AckA) might diffuse into the compartment and be acted on by EutD, but the produced acetyl-CoA could not escape because of its large size. According to the previous view of the ethanolamine pathway, EutD is needed only anaerobically or to supplement the insufficient activity of Pta during growth on ethanolamine (10, 12). This view predicts that induced EutD should replace Pta with no disruption of the compartment. These predictions were tested using a pta acs double mutant, which cannot use acetate as a carbon source (Fig. 5B).

These tests require induction of expression of the eut operon, which is normally induced by the simultaneous presence of ethanolamine and B₁₂ (25). However, under these conditions, Salmonella would simply utilize ethanolamine for growth. Removal of EutBC (ethanolamine-ammonia lyase) would prevent this growth but might disrupt the microcompartment. Therefore, the operon was induced by the gratuitous inducer ethylenediamine, which induces the eut operon ∼1.6-fold better than ethanolamine at the 10 mM concentration used. Ethylenediamine neither serves as a carbon source nor interferes with use of ethanolamine in wild-type cells (A. Britton-Soulette, P. Anderson, and D. L. Huseby, unpublished results).

In Fig. 6, it can been seen that neither the ethylenediamine nor acetate alone supported significant growth of a pta acs double mutant. Addition of ethylenediamine plus acetate did not allow the pta acs double mutant to grow on acetate when cells had a wild-type eut operon (and presumably intact microcompartments) (Fig. 6A). The failure of induced EutD to substitute for the missing Pta seems to be caused by the microcompartment, since strong growth was observed on ethylenediamine plus acetate by strains carrying the eutS, eutMN, and eutLK deletion mutations, in addition to the pta and acs mutations (Fig. 6B). This growth depends on the induced EutD enzyme since growth on acetate was blocked in strains with a eutD mutation, in addition to the shell protein defects.

In this test, growth on acetate required EutD but no other eut enzyme. This allowed tests of the effect of various individual eut proteins on microcompartment integrity. Strains that all carried the acs and pta mutations plus one or more mutations in the eut operon were constructed. Mutations that affect microcompartment integrity were arbitrarily defined as mutations that provided growth on acetate at a rate greater than 70% of that seen for strain TT26710 with deletions of all shell protein genes, eutS, eutMN, eutLK, pta, and acs (defined as 100%). By this definition, disruption was caused by single deletion mutations in the shell protein genes eutM, eutN, and eutK, the eutMN and eutLK multiple mutations, and the eutS, eutMN, and eutLK multiple mutations. Partial disruption was defined as restoration of less than 70% but greater than 60% of the standard. This was seen for the deletion mutations removing eutE (aldehyde dehydrogenase), eutJ (unknown function), and eutB and eutC (ethanolamine-ammonia lyase) or the shell protein gene eutL.

Mutations judged to have no effect on microcompartment formation (growth rate on acetate, less than 50% of the standard) were eutS (microcompartment shell), eutP (unknown function), eutG (alcohol dehydrogenase), eutT (B₁₂ adenosyltransferase), and eutA (reactivation of lyase following B₁₂ damage). As expected from the nature of the assay, mutations that removed the EutD protein eliminated all growth on acetate with or without a microcompartment.

Strains carrying a eutD mutation grew significantly less well on acetate than did strains with any of the other mutations that failed to stimulate growth on acetate. Tester strains (carrying only the pta and acs mutations) showed extremely limited growth on acetate in the presence of ethylenediamine, but even this background growth was eliminated by a eutD deletion.

These results support the idea that disruption of the microcompartment allows growth of a pta acs mutant on acetate. The failure of the eutS deletion to improve growth was not unexpected, since this deletion also failed to improve growth in the eutD suppression assay. Surprisingly, eutQ mutations had no effect in the acetate assay (Table 5) but did serve to suppress eutD mutations on ethanolamine (Table 4). Also surprising was the finding that strains with mutations in the eutB, eutC, or eutE genes, which encode central enzymes, showed partially improved growth on acetate, suggesting that these proteins are needed for assembly or maintenance of the microcompartment. Other genes that showed no effect on improvement of growth on acetate were eutP (no known function) and eutG (alcohol dehydrogenase). The EutA and EutT proteins seemed to play no role in compartment formation, and both are required for reactivation of EutBC (ethanolamine-ammonia lyase), required following frequent radical damage to the B₁₂ cofactor (39). The EutA protein extracts the damaged cofactor (40), and EutT adenosylates B₁₂ to produce the replacement cofactor (33). The failure of these proteins to contribute to microcompartment structure is consistent with the idea that they act on the EutBC enzyme located outside the compartment. The ethanolamine-ammonia lyase itself may be more intimately associated with the microcompartment, since its absence compromised the compartment slightly and allowed some growth on acetate.

Table 5.

Effect of eut deletion mutations on microcompartment integrity as judged by acetate growth

Genotype^a	Strain	Doubling time (h/generation)	Relative fitness^b	Microcompartment disruption
Wild type	TT26709	>16	<0.23	−
eutSΔ eutMNΔ eutLKΔ	TT26710	3.69	1.00	+
eutMNΔ eutLKΔ	TT26711	3.97	0.93	+
eutMΔ	TT26713	3.63	1.02	+
eutNΔ	TT26714	4.60	0.80	+
eutKΔ	TT26716	4.72	0.78	+
eutLΔ	TT26715	5.47	0.67	±
eutBΔ	TT26718	6.07	0.61	±
eutCΔ	TT26719	5.70	0.65	±
eutEΔ	TT26721	5.51	0.67	±
eutJΔ	TT26726	5.78	0.64	±
eutSΔ	TT26712	9.00	0.41	−
eutQΔ	TT26724	7.95	0.46	−
eutPΔ	TT26723	11.52	0.32	−
eutTΔ	TT26725	11.50	0.32	−
eutAΔ	TT26717	>16	<0.23	−
eutGΔ	TT26722	>16	<0.23	−
eutDΔ	TT26720	>16	<0.23	−
eutDMΔ	TT26727	>16	<0.23	+

Open in a new tab

All strains were pta acs mutants.

Relative fitness was scaled to the doubling time of the eutSΔ eutMNΔ eutLKΔ mutant grown with acetate and ethylenediamine, whose fitness was defined to be equal to 1.

Growth phenotypes on ethanol.

Normally, Salmonella cannot use ethanol as a carbon and energy source under aerobic conditions, which prevent expression of alcohol dehydrogenase (31). However, mutants of both Salmonella enterica and Escherichia coli that express alcohol dehydrogenases aerobically can grow on ethanol (31, 41). To learn about the nature of the microcompartment, we used ethanol as a source of carbon and energy. The initial expectation was that EutG would convert ethanol to acetaldehyde, which EutE might convert to acetyl-CoA (Fig. 7). The microcompartment was expected to interfere because both Eut reactions would produce NADH, making it impossible to recycle NAD within the compartment. As seen below, the experiment worked as expected, in that growth on ethanol did depend on disruption of the compartment and induction of the eut operon. Unexpectedly, this growth depended on the housekeeping alcohol dehydrogenase AdhP and not on EutG. These points are developed below.

Fig 7 — Use of the EutE enzyme to support growth on ethanol. Gray arrows, the initial expectation that the EutG and EutE enzymes would degrade external ethanol to acetaldehyde and disruption would be needed to recycle NADH; black arrows, actual results suggesting that growth on ethanol depends on the outside enzyme AdhP, which produces acetaldehyde that is then oxidized by EutE. Compartment disruption seems to be required to allow acetaldehyde to access EutE.

Growth on ethanol is described in Fig. 8. As predicted by the model, growth required both induction of the eut operon and disruption of the compartment. As before, this assay was then used to test the effect of all the eut operon mutants on compartment integrity. These experiments (Table 6) supported the general idea of the model but revealed an unexpected result: the EutG alcohol dehydrogenase was not required for growth.

Table 6.

Effect of eut mutations on ability to grow on ethanol

Line no.	Genotype	Strain	Doubling time (h/generation)	Relative fitness^a	Microcompartment disruption
1	Wild type	TR10000	>16	<0.23	−
2	eutSΔ eutMNΔ eutLKΔ	TT26680	3.6	1.00	+
3	eutMNΔ eutLKΔ	TT26679	3.9	0.94	+
4	eutMNEΔ	TT26730	>16	<0.23	No EutE activity
5	eutEΔ	TT22815	>16	<0.23	No EutE activity
6	eutGΔ	TT23036	>16	<0.23	−
7	eutMΔ eutGΔ	TT26729	4.6	0.80	+
8	eutMΔ eutGΔ adhP	TT26737	>16	<0.23	No EutG or AdhP dehydrogenase
9	eutSΔ eutMNΔ eutLKΔ adhP	TT26738	>16	<0.23	No AdhP dehydrogenase
10	eutDΔ	TT24804	>16	<0.23	−
11	eutDΔ eutMΔ	TT26694	3.5	1.03	+
12	eutDΔ eutE (signal sequence)	TT26700	3.6	1.00	+
13	eutQΔ	TT24802	5.1	0.72	+
14	eutMΔ	TT22523	4.3	0.85	+
15	eutLΔ	TT22570	4.1	0.88	+
16	eutSΔ	TT22522	9.6	0.38	−
17	eutKΔ	TT22571	>16	<0.23	−
18	eutNΔ	TT22569	>16	<0.23	−
19	eutAΔ	TT24805	14.7	0.25	−
20	eutBΔ	TT24806	>16	<0.23	−
21	eutCΔ	TT24807	>16	<0.23	−
22	eutTΔ	TT24803	>16	<0.23	−
23	eutPΔ	TT24801	>6	<0.23	−
24	eutJΔ	TT26728	>16	<0.23	−

Open in a new tab

Fitness is the ratio of the doubling time for the eutSΔ eutMNΔ eutLKΔ mutant to that of the other strains.

Using the same strategy used for growth on acetate, the eut operon was induced by ethylenediamine and various constructed mutants were tested for their ability to grow on ethanol. The initial expectation was that ethanol would diffuse into the compartment and have free access to the EutG alcohol dehydrogenase, but ethanol oxidation would be limited because the microcompartment would prevent recycling of the produced NADH (Fig. 7, gray arrows). This problem was expected to be severe because during growth on ethanol, both EutG (alcohol dehydrogenase) and EutE (acetaldehyde oxidoreductase) would generate NADH, which could not be recycled to NAD (diagrammed in Fig. 7). Disruption of the compartment was expected to allow NAD recycling so that EutG and EutE could work together in support of growth on ethanol.

Results of these tests are in Table 6. In all cases, cells were grown with ethylenediamine to induce the eut operon. The results in the top three lines of this table are as expected. Wild-type cells failed to grow on ethanol, but growth was restored by mutations that disrupted the compartment (as seen in Fig. 8). Also as expected, growth depended on the EutE enzyme, acetaldehyde dehydrogenase (lines 4 and 5). The surprise (lines 6 and 7) was that growth did not depend on EutG (ethanol dehydrogenase). Even a eutG mutant (line 6) was allowed to grow by a shell protein mutation (line 7). Thus, cells having EutE could grow on ethanol, even without the EutG ethanol dehydrogenase. How was ethanol converted to acetaldehyde?

The responsible enzyme turned out to be AdhP, a little-known alcohol dehydrogenase, present in both S. enterica and E. coli, that is known to be induced by ethanol (42). Mutations in this gene were tested because they had previously been seen to contribute to ethanolamine metabolism in certain mutant strains (T. Fazzio, unpublished results). An adhP mutation prevented EutE-dependent growth on ethanol, whether or not the strain had EutG activity (lines 8 and 9), but the growth of an adhP⁺ strain occurred only if the eut operon was induced and the compartment was disrupted. This presumed role of AdhP is diagrammed in Fig. 7 (dark arrows). We confirmed the role of AdhP by observing (by NMR) the accumulation of small amounts of acetaldehyde in cultures exposed to ethanol, but only if the tested strain contained a wild-type adhP gene; strains in which the adhP gene was disrupted failed to show this accumulation (data not shown). This acetaldehyde production did not require induction of the eut operon or disruption of the compartment.

The involvement of AdhP is surprising because it suggests that acetaldehyde produced outside the compartment (by the housekeeping enzyme AdhP) is being used for growth. If aldehyde can diffuse into the compartment, the situation would resemble normal growth on ethanolamine and catabolism could proceed with an intact compartment and normal cofactor recycling. However, growth on ethanol required disruption of the compartment. If catabolism of externally generated acetaldehyde requires compartment disruption, this suggests strongly that the compartment restricts diffusion of acetaldehyde.

The ability of various mutations to disrupt the compartment and restore growth is included in Table 6. Growth rates were measured and fitness was calculated relative to the growth rate and fitness of a strain lacking all five shell proteins, whose fitness was defined to be 1.00. Fitness values greater than 50% were defined as positive suppression, and those below 50% were defined as a lack of suppression. The distinction between these was quite clear for strains with all mutations with the exception of the strain with the eutS mutation, which showed a small but significant improvement in fitness.

As predicted by the model, EutD (phosphotransacetylase) is not required for growth on ethanol in this assay since once the compartment is disrupted to allow entry of acetaldehyde, recycling of NAD is made possible even without EutG (Table 6, lines 10 and 11). Apparently, disruption of the compartment by a shell protein defect gives acetyl-CoA access to the TCA cycle and allows recycling of CoA and NADH. Alteration of the EutE signal sequence allowed growth on ethanol with no shell protein defect (compare lines 10 and 12 in Table 6), just as it restored the growth of a eutD mutant on ethanolamine (see above). This suggests that for growth on ethanol, it is sufficient for EutE to be outside the compartment, where it has access to acetaldehyde produced by AdhP and can produce acetyl-CoA with access to the TCA cycle. Similarly, a eutQ mutation allowed growth on ethanol, just as it allowed a eutD mutant to grow on ethanolamine. This suggests that EutQ may be needed to enclose EutE in a compartment.

DISCUSSION

A genetic approach has shed light on the physiological role and functional mechanism of microcompartments in the ethanolamine (eut) pathway of Salmonella. Early views of this pathway assumed that compartments could retain the small and uncharged acetaldehyde molecule while allowing the passage of substrates and products, including ethanolamine and the huge cofactors CoA/acetyl-CoA and NAD/NADH. These models did not suggest a central role for the accessory enzymes EutG and EutD but proposed that they might contribute during anaerobic growth. The need for a new view of the pathway was suggested by the failure of mutants lacking the peripheral EutG or EutD enzymes to grow on ethanolamine under either aerobic or anaerobic conditions. In addition, mutants lacking these enzymes showed an increased loss of acetaldehyde to the gas phase, suggesting a central role in the pathway (4). These early results were consistent with an ability of microcompartments to retain acetaldehyde but were hard to pursue genetically because mutants lacking shell proteins showed only minor defects in growth on ethanolamine. The lack of mutant phenotypes could mean that the compartments provide a benefit too minor to note in the lab but sufficient to ensure selective maintenance of these genes in natural populations. Alternatively, the compartments may be essential for use of ethanolamine, but only under some unidentified natural condition, such as very low ethanolamine concentrations. The robust phenotypes for shell proteins reported here and those seen on ethanolamine at pH 8.0 (4) suggest that the compartments are essential under some conditions.

Natural conditions under which ethanolamine use is important have recently been identified (34, 43). The ethanolamine operon is part of a constellation of genes that is central to defining Salmonella as a species. Together, these genes constitute nearly 2% of the Salmonella genome and encode proteins for catabolism of ethanolamine and 1,2-propanediol, synthesis of B₁₂ (cob and cbi), and use of tetrathionate as an electron acceptor (ttr, phs, asr) (11, 44, 45). These genes are found in all Salmonella isolates and were used individually as taxonomic criteria for identifying Salmonella even before it was clear that they acted together (46, 47). Recent results show that this constellation of genes allows Salmonella to grow in an inflamed gut (34, 43) using two poor carbon sources (ethanolamine and 1,2-propanediol). These compounds can be used anaerobically because of the alternative electron acceptor (tetrathionate), which is continuously reoxidized by oxidative bursts from the host tissue (43). This suggests that Salmonella may grow on very low concentrations of carbon sources that are fed very slowly. We suggest and are testing the possibility that shell protein mutants may have more striking phenotypes during slow growth on limiting levels of ethanolamine.

Since simple shell protein mutants showed no ethanolamine growth defect under standard laboratory conditions, we used alternative conditions under which these mutants had a clear phenotype. The unexpected strong Eut⁻ phenotype of a eutD mutant allowed selection of suppressor mutants, whose microcompartment defects allowed growth. This suggested that the compartment is central to the ethanolamine pathway and led to the model in which the three enzymes involved in acetaldehyde catabolism function within the microcompartment and use private cofactor pools that must be recycled. This view was supported by the compartment preventing the EutD phosphotransacetylase from contributing to growth on acetate (in place of the Pta enzyme). Another support was the compartment preventing EutE from supporting growth on ethanol that had been converted outside the compartment to acetaldehyde by AdhP. The disruption required for growth on ethanol suggested that the compartment blocked the influx and presumably also the efflux of acetaldehyde (see below). Had acetaldehyde been able to diffuse in, one would have expected the intact compartment to facilitate growth on ethanol by the coupled activity of EutE, EutD, and EutG, as occurs during normal ethanolamine degradation.

The model proposes that microcompartments do more than contain aldehyde. They may present aldehyde to a local high concentration of enzymes and cofactors. The internal recycling of cofactors (CoA and NADH) couples reactions and lets them act together to accelerate consumption of acetaldehyde, with no required accumulation. It has previously been proposed that the aldehyde microcompartments (of the eut and pdu pathways) might cycle NAD/NADH internally (4, 15). Recently published work of Cheng et al. has provided evidence for NADH recycling in the analogous pdu microcompartments of Salmonella (15). The model presented here includes a role for contained phosphotransacetylase (EutD) function and recycling of CoA. By retaining injected aldehyde and coupling the reactions that degrade it, the compartment accelerates use of a volatile intermediate and stimulates growth on ethanolamine.

The model makes a prediction (Fig. 4) that the first enzyme, ethanolamine-ammonia lyase (EutBC), is associated with the compartment but lies outside its lumen. While this possibility is not addressed directly by the data presented, it seems attractive in the light of the other results. The radical-based B₁₂ reaction mechanism leads to frequent destruction of the cofactor. Thus, EutBC activity requires frequent cofactor removal and replacement by intact Ado-B₁₂. Removal is accomplished by the EutA enzyme (33, 40), and replacement requires Ado-B₁₂, which can be provided either by the EutT adenosyltransferase encoded in the eut operon or by the housekeeping enzyme CobA (33, 48). If EutBC (and, by extension, B₁₂) were within the compartment with the rest of the enzymes, then removal and replacement of the damaged cofactor by fresh Ado-B₁₂ would require a continual influx of B₁₂ and ATP and possibly an efflux of damaged Ado-B₁₂. These fluxes seem unlikely if the compartment blocks passage of large molecules.

We suggest instead that EutBC is associated with the compartment and functions in a vectoral manner. That is, it acquires its cofactor and substrate ethanolamine outside the compartment and releases its product, acetaldehyde, inside the compartment. The enzymes involved in EutBC reactivation (EutA, EutT) could be associated with the cytoplasmic side of the lyase. This is similar to a model that has been proposed for carbonic anhydrase in some types of carboxysomes (49). The association of B₁₂ with the compartment is supported by the demonstration that the B₁₂-dependent enzyme of the 1,2-propanediol pathway is recovered among the proteins of purified microcompartments, suggesting at least a tight association (50). The accessibility of EutBC to external B₁₂ is also supported by the observations that growth of a EutT mutant (lacking B₁₂ adenosyltransferase) is allowed by the equivalent housekeeping transferase, CobA, suggesting that the Ado-B₁₂ can be provided outside the compartment (33).

For the analogous pdu operon, an alternate method of maintaining the Ado-B₁₂ cofactor within the microcompartment has been proposed (51). The gated pores seen in purified shell protein multimers have been proposed to allow passage of large molecules (cofactors) across the shell of the microcompartment, while still maintaining the impermeability of the compartment to small molecules (aldehydes) (52). Further study is needed to distinguish between these models.

The cofactor recycling proposed for the ethanolamine microcompartment begs the question of how this model might apply to the homologous carboxysomes, which contain RuBisCo in CO₂-fixing cyanobacteria. Carboxysomes are assumed to retain CO₂, produced within the compartment by carbonic anhydrase. Thus, CO₂ is immediately exposed to a high local concentration of RuBisCo. This alone could contribute to rapid fixation, whether or not diffusion of CO₂ is restricted. Restricted CO₂ diffusion would require the compartment to allow entry of ribulose bisphosphate and bicarbonate (HCO₃⁻) and exit of glyceraldehyde 3-phosphate while blocking exit of CO₂. This would be only slightly more difficult than the discrimination proposed here for ethanolamine microcompartments, which appears to allow passage of ethanol and acetyl-PO₄ while restricting exit of acetaldehyde. RuBisCo does not require cofactors that must be recycled and thus does not require the recycling proposed here for aldehyde compartments. The underlying principle of all compartments may be presentation of a difficult-to-contain substrate to a high local concentration of catalyst. This benefit is enhanced if efflux is restricted. In the case of aldehyde metabolism, this basic mechanism can be realized only if the necessary cofactors can be recycled internally.

Microcompartments are shells composed of 5 or more protein types that encapsulate one or more enzymes. Recent work has shed light on how shell proteins oligomerize and how they might restrict the flow of metabolites. The oligomeric structure of individual shell proteins is a thin hexagonal prism (53). A pore formed at the intersection of the 6 subunits of the hexamer is presumed to gate the access of substrates and products to and from the interior of the microcompartment (53, 54). The nature of these pores could dictate which molecules are permitted to enter and leave the compartment. In previous pathway models, these pores would have to allow very large molecules—NAD (663 Da), CoA (767 Da), and, possibly, Ado-B₁₂ (1,570 Da)—to pass but restrict the diffusion of acetaldehyde (44 Da). In the new model, pores either have no role in metabolite transport or may discriminate between modest-sized charged metabolites and small nonpolar compounds (aldehydes, O₂, or CO₂) while blocking movement of large cofactors.

A major problem is how to assemble the compartment around the requisite enzymes and cofactors. No scaffold seems to be required to assemble the carboxysome (55). In contrast, deletions that eliminated EutB, EutC, or EutE appeared to disrupt the compartment so as to allow growth on acetate. None of these enzyme activities is required for acetate growth, so their loss seems to disturb the compartment sufficiently that acetyl-CoA (produced from acetyl-PO₄ by EutD) can escape the compartment and gain access to enzymes of the TCA cycle. This suggests that these enzymes play a role in assembly, perhaps as scaffold proteins.

Interpretation of these results is difficult, however, because defects in different shell proteins appeared to have different effects on microcompartment integrity, depending on the assay used (discussed below). Table 7 summarizes the effects of various shell protein deletion mutations on the assays described by the data in Tables 4 to 6.

Table 7.

Role of various proteins on compartment integrity as judged by three assays

Mutation added	Growth of^a:
Mutation added	eutD mutant on ethanolamine with release of acetyl-CoA or EutE	pta acs mutant on acetate with release of acetyl-CoA or EutD	Wild type (adhP⁺) on ethanol with entry of acetaldehyde or release of EutE
None (wild type)	−	−	−
eutM	+	+	+
eutL	+	±	+
eutK	+	+	−
eutN	ND	+	−
eutQ	+	−	+
eutS	±	−	−
eutE	^*	±	^*
eutE (signal)^b	+	ND	+
eutB	^*	±	−
eutC	^*	±	−
eutD	−	^*	−
eutG	−	−	−
eutJ	−	±	−
eutP	−	−	−
eutT	ND	−	−
eutA	^*	−	−

Open in a new tab

+, the added mutation disrupts the compartment as judged by this assay; ±, the added mutation leaves compartment intact as judged by this assay; −, the added mutation partially disrupts the compartment; ND, the added mutation was not tested in this assay; *, this protein is needed for growth in this assay and could not be tested for its contribution to the compartment.

Signal sequence mutation.

A simple form of the model (Fig. 4) might predict that the compartment would be completely destroyed by removal of any one of the five proteins thought to be components of the shell (EutSMNKL). This clearly did not occur. Removal of the EutM and EutL proteins disrupted the compartment, as judged by all of the assays. However, removal of EutS had little or no effect in any of the assays. Removal of EutN and EutK appeared to disrupt the compartment, as judged by the acetate assay but not as judged by the ethanol growth assay. These inconsistencies may suggest that some residual functions may be retained by compartments that assemble without all of their shell proteins. This possibility is consistent with the observation of abnormal compartment-like structures in cells lacking particular proteins, which has been extensively documented for the pdu compartments (16) and to a more limited extent for those of eut (13). Another possibility is that disruption of the compartment may leave assemblages that inhibit the activity of certain enzymes.

Another curious aspect of these results is the failure of the EutG enzyme to convert ethanol to acetaldehyde, even after the compartment is disrupted. That is, all growth on ethanol depended on the AdhP enzyme, which is apparently not a normal part of the ethanolamine pathway. We suggest that EutG may operate within the intact compartment during growth on ethanolamine but become inactive or oxygen sensitive once it is released from a disrupted compartment. We suspect that EutG is functional within eut microcompartments, even under aerobic conditions, based upon the Eut⁻ phenotype of eutG mutants and the suppression of this phenotype, like that for eutD, by mutations that disrupt the microcompartment (Huseby and Roth, unpublished). The putative oxygen sensitivity of EutG is supported by the observed oxygen sensitivity of the analogous pdu enzyme, PduQ (56), but the apparent continued function of EutG in the metabolism of ethanolamine under aerobic conditions suggests that the microcompartment may confer some protection from molecular oxygen.

We conclude that the microcompartment restricts diffusion of acetaldehyde. This is primarily based on the observation that acetaldehyde produced by the housekeeping AdhP enzyme did not support growth unless the compartment was disrupted. Had acetaldehyde diffused into the intact compartment, we would have expected it to be rapidly consumed by the coupled EutE, EutD, and EutG reactions. Similarly, attempts to show cell growth on exogenous acetaldehyde have not been successful. However, considerable acetaldehyde is released from wild-type cultures growing on ethanolamine (4), suggesting that the diffusion barrier is not complete. These release assays involved continuously removing and exchanging air above a growing culture, a process that could drive any minimal acetaldehyde escape. Similarly, we have observed the accumulation and subsequent disappearance of acetaldehyde in the medium of cells growing on ethanolamine (Fig. 3). It is not clear whether this disappearance of acetaldehyde reflected consumption within compartments or cytosolic conversion to ethanol or acetate by housekeeping enzymes. Our best estimate is that the compartment restricts but may not completely block diffusion of acetaldehyde.

ACKNOWLEDGMENTS

This work was supported in part by NIH grant AI088122 to Andreas Bäumler and J.R.R. and grant GM27068 to J.R.R. NMR analysis was done at the UC-Davis NMR facility, which is supported by NSF DBIO 722538 and NIH RR11973.

We thank Andreas Bäumler, Sebastian Winter, Alfred Spormann, Jerry Dallas, Joseph Penrod, and Valley Stewart for valuable suggestions in the course of this work.

Footnotes

Published ahead of print 12 April 2013

REFERENCES

1. Bobik TA, Havemann GD, Busch RJ, Williams DS, Aldrich HC. 1999. The propanediol utilization (pdu) operon of Salmonella enterica serovar Typhimurium LT2 includes genes necessary for formation of polyhedral organelles involved in coenzyme B12-dependent 1,2-propanediol degradation. J. Bacteriol. 181:5967–5975 [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Kofoid E, Rappleye C, Stojiljkovic I, Roth J. 1999. The 17-gene ethanolamine (eut) operon of Salmonella typhimurium encodes five homologues of carboxysome shell proteins. J. Bacteriol. 181:5317–5329 [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Shively JM, Ball F, Brown DH, Saunders RE. 1973. Functional organelles in prokaryotes: polyhedral inclusions (carboxysomes) of Thiobacillus neapolitanus. Science 182:584–586 [DOI] [PubMed] [Google Scholar]
4. Penrod JT, Roth JR. 2006. Conserving a volatile metabolite: a role for carboxysome-like organelles in Salmonella enterica. J. Bacteriol. 188:2865–2874 [DOI] [PMC free article] [PubMed] [Google Scholar]
5. Sampson EM, Bobik TA. 2008. Microcompartments for B₁₂-dependent 1,2-propanediol degradation provide protection from DNA and cellular damage by a reactive metabolic intermediate. J. Bacteriol. 190:2966–2971 [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Roof DM, Roth JR. 1988. Ethanolamine utilization in Salmonella typhimurium. J. Bacteriol. 170:3855–3863 [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Roof DM, Roth JR. 1989. Functions required for vitamin B₁₂-dependent ethanolamine utilization in Salmonella typhimurium. J. Bacteriol. 171:3316–3323 [DOI] [PMC free article] [PubMed] [Google Scholar]
8. Zhu H, Gonzalez R, Bobik TA. 2011. Coproduction of acetaldehyde and hydrogen during glucose fermentation by Escherichia coli. Appl. Environ. Microbiol. 77:6441–6450 [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Brinsmade SR, Escalante-Semerena JC. 2004. The eutD gene of Salmonella enterica encodes a protein with phosphotransacetylase enzyme activity. J. Bacteriol. 186:1890–1892 [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Bologna F, Campos-Bermudez V, Saavedra D, Andreo C, Drincovich M. 2010. Characterization of Escherichia coli EutD: a phosphotransacetylase of the ethanolamine operon. J. Microbiol. 48:629–636 [DOI] [PubMed] [Google Scholar]
11. Price-Carter M, Tingey J, Bobik TA, Roth JR. 2001. The alternative electron acceptor tetrathionate supports B₁₂-dependent anaerobic growth of Salmonella enterica serovar Typhimurium on ethanolamine or 1,2-propanediol. J. Bacteriol. 183:2463–2475 [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Starai VJ, Garrity J, Escalante-Semerena JC. 2005. Acetate excretion during growth of Salmonella enterica on ethanolamine requires phosphotransacetylase (EutD) activity, and acetate recapture requires acetyl-CoA synthetase (Acs) and phosphotransacetylase (Pta) activities. Microbiology 151:3793–3801 [DOI] [PubMed] [Google Scholar]
13. Brinsmade SR, Paldon T, Escalante-Semerena JC. 2005. Minimal functions and physiological conditions required for growth of Salmonella enterica on ethanolamine in the absence of the metabolosome. J. Bacteriol. 187:8039–8046 [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Fan C, Bobik TA. 2011. The N-terminal region of the medium subunit (PduD) packages adenosylcobalamin-dependent diol dehydratase (PduCDE) into the Pdu microcompartment. J. Bacteriol. 193:5623–5628 [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Fan C, Cheng S, Sinha S, Bobik TA. 2012. Interactions between the termini of lumen enzymes and shell proteins mediate enzyme encapsulation into bacterial microcompartments. Proc. Natl. Acad. Sci. U. S. A. 109:14995–15000 [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Cheng S, Sinha S, Fan C, Liu Y, Bobik TA. 2011. Genetic analysis of the protein shell of the microcompartments involved in coenzyme B₁₂-dependent 1,2-propanediol degradation by Salmonella. J. Bacteriol. 193:1385–1392 [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Davis RW, Botstein D, Roth JR. 1980. Advanced bacterial genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY [Google Scholar]
18. Schmieger H. 1971. A method for detection of phage mutants with altered transducing ability. Mol. Gen. Genet. 110:378–381 [DOI] [PubMed] [Google Scholar]
19. Penrod JT, Mace CC, Roth JR. 2004. A pH-sensitive function and phenotype: evidence that EutH facilitates diffusion of uncharged ethanolamine in Salmonella enterica. J. Bacteriol. 186:6885–6890 [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Datsenko KA, Wanner BL. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. U. S. A. 97:6640–6645 [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Yu D, Sawitzke JA, Ellis H, Court DL. 2003. Recombineering with overlapping single-stranded DNA oligonucleotides: testing a recombination intermediate. Proc. Natl. Acad. Sci. U. S. A. 100:7207–7212 [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Lawes M, Maloy S. 1995. MudSacI, a transposon with strong selectable and counterselectable markers: use for rapid mapping of chromosomal mutations in Salmonella typhimurium. J. Bacteriol. 177:1383–1387 [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Berkowitz D, Hushon JM, Whitfield HJ, Roth J, Ames BN. 1968. Procedure for identifying nonsense mutations. J. Bacteriol. 96:215–220 [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Neidhardt FC, Bloch PL, Smith DF. 1974. Culture medium for enterobacteria. J. Bacteriol. 119:736–747 [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Roof DM, Roth JR. 1992. Autogenous regulation of ethanolamine utilization by a transcriptional activator of the eut operon in Salmonella typhimurium. J. Bacteriol. 174:6634–6643 [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Miller JH. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY [Google Scholar]
27. Sheppard DE, Roth JR. 1994. A rationale for autoinduction of a transcriptional activator: ethanolamine ammonia-lyase (EutBC) and the operon activator (EutR) compete for adenosyl-cobalamin in Salmonella typhimurium. J. Bacteriol. 176:1287–1296 [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Ripley LS. 1982. Model for the participation of quasi-palindromic DNA sequences in frameshift mutation. Proc. Natl. Acad. Sci. U. S. A. 79:4128–4132 [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Hughes KT, Roth JR. 1984. Conditionally transposition-defective derivative of Mu d1(Amp Lac). J. Bacteriol. 159:130–137 [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Wolfe AJ. 2005. The acetate switch. Microbiol. Mol. Biol. Rev. 69:12–50 [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Dailly Y, Mat-Jan F, Clark DP. 2001. Novel alcohol dehydrogenase activity in a mutant of Salmonella able to use ethanol as sole carbon source. FEMS Microbiol. Lett. 201:41–45 [DOI] [PubMed] [Google Scholar]
32. Dailly YP, Bunch P, Clark DP. 2000. Comparison of the fermentative alcohol dehydrogenases of Salmonella typhimurium and Escherichia coli. Microbios 103:179–196 [PubMed] [Google Scholar]
33. Sheppard DE, Penrod JT, Bobik T, Kofoid E, Roth JR. 2004. Evidence that a B₁₂-adenosyl transferase is encoded within the ethanolamine operon of Salmonella enterica. J. Bacteriol. 186:7635–7644 [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Thiennimitr P, Winter SE, Winter MG, Xavier MN, Tolstikov V, Huseby DL, Sterzenbach T, Tsolis RM, Roth JR, Baumler AJ. 2011. Intestinal inflammation allows Salmonella to use ethanolamine to compete with the microbiota. Proc. Natl. Acad. Sci. U. S. A. 108:17480–17485 [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Brown TDK, Jones-Mortimer MC, Kornberg HL. 1977. The enzymic interconversion of acetate and acetyl-coenzyme A in Escherichia coli. J. Gen. Microbiol. 102:327–336 [DOI] [PubMed] [Google Scholar]
36. LeVine SM, Ardeshir F, Ames GF. 1980. Isolation and characterization of acetate kinase and phosphotransacetylase mutants of Escherichia coli and Salmonella typhimurium. J. Bacteriol. 143:1081–1085 [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Clark DP, Cronan JP., Jr 1996. Two-carbon compounds and fatty acids as carbon sources, p 343–357 In Neidhardt FC, Curtiss R, III, Ingraham JL, Lin ECC, Low KB, Magasanik B, Reznikoff WS, Riley M, Schaechter M, Umbarger HE. (ed), Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed, vol 1 American Society for Microbiology, Washington, DC [Google Scholar]
38. Kumari S, Tishel R, Eisenbach M, Wolfe AJ. 1995. Cloning, characterization, and functional expression of acs, the gene which encodes acetyl coenzyme A synthetase in Escherichia coli. J. Bacteriol. 177:2878–2886 [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Toraya T. 2003. Radical catalysis in coenzyme B₁₂-dependent isomerization (eliminating) reactions. Chem. Rev. 103:2095–2128 [DOI] [PubMed] [Google Scholar]
40. Mori K, Bando R, Hieda N, Toraya T. 2004. Identification of a reactivating factor for adenosylcobalamin-dependent ethanolamine ammonia lyase. J. Bacteriol. 186:6845–6854 [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Clark D, Cronan JE. 1980. Escherichia coli mutants with altered control of alcohol dehydrogenase and nitrate reductase. J. Bacteriol. 141:177–183 [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Shafqat J, Höög J-O, Hjelmqvist L, Oppermann UCT, Ibáñez C, Jörnvall H. 1999. An ethanol-inducible MDR ethanol dehydrogenase/acetaldehyde reductase in Escherichia coli. Eur. J. Biochem. 263:305–311 [DOI] [PubMed] [Google Scholar]
43. Winter SE, Thiennimitr P, Winter MG, Butler BP, Huseby DL, Crawford RW, Russell JM, Bevins CL, Adams LG, Tsolis RM, Roth JR, Baumler AJ. 2010. Gut inflammation provides a respiratory electron acceptor for Salmonella. Nature 467:426–429 [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Lawrence JG, Roth JR. 1996. Evolution of coenzyme B₁₂ synthesis among enteric bacteria: evidence for loss and reacquisition of a multigene complex. Genetics 142:11–24 [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Anjum MF, Marooney C, Fookes M, Baker S, Dougan G, Ivens A, Woodward MJ. 2005. Identification of core and variable components of the Salmonella enterica subspecies I genome by microarray. Infect. Immun. 73:7894–7905 [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Patil MD, Parhad NM. 1986. Growth of salmonellas in different enrichment media. J. Appl. Microbiol. 61:19–24 [DOI] [PubMed] [Google Scholar]
47. Rambach A. 1990. New plate medium for facilitated differentiation of Salmonella spp. from Proteus spp. and other enteric bacteria. Appl. Environ. Microbiol. 56:301–303 [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Buan NR, Suh S-J, Escalante-Semerena JC. 2004. The eutT gene of Salmonella enterica encodes an oxygen-labile, metal-containing ATP:corrinoid adenosyltransferase enzyme. J. Bacteriol. 186:5708–5714 [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Dou Z, Heinhorst S, Williams EB, Murin CD, Shively JM, Cannon GC. 2008. CO₂ fixation kinetics of Halothiobacillus neapolitanus mutant carboxysomes lacking carbonic anhydrase suggest the shell acts as a diffusional barrier for CO₂. J. Biol. Chem. 283:10377–10384 [DOI] [PubMed] [Google Scholar]
50. Havemann GD, Bobik TA. 2003. Protein content of polyhedral organelles involved in coenzyme B₁₂-dependent degradation of 1,2-propanediol in Salmonella enterica serovar Typhimurium LT2. J. Bacteriol. 185:5086–5095 [DOI] [PMC free article] [PubMed] [Google Scholar]
51. Cheng S, Bobik TA. 2010. Characterization of the PduS cobalamin reductase of Salmonella enterica and its role in the Pdu microcompartment. J. Bacteriol. 192:5071–5080 [DOI] [PMC free article] [PubMed] [Google Scholar]
52. Tanaka S, Sawaya MR, Yeates TO. 2010. Structure and mechanisms of a protein-based organelle in Escherichia coli. Science 327:81–84 [DOI] [PubMed] [Google Scholar]
53. Kerfeld CA, Sawaya MR, Tanaka S, Nguyen CV, Phillips M, Beeby M, Yeates TO. 2005. Protein structures forming the shell of primitive bacterial organelles. Science 309:936–938 [DOI] [PubMed] [Google Scholar]
54. Crowley CS, Cascio D, Sawaya MR, Kopstein JS, Bobik TA, Yeates TO. 2010. Structural insight into the mechanisms of transport across the Salmonella enterica Pdu microcompartment shell. J. Biol. Chem. 285:37838–37846 [DOI] [PMC free article] [PubMed] [Google Scholar]
55. Menon BB, Dou Z, Heinhorst S, Shively JM, Cannon GC. 2008. Halothiobacillus neapolitanus carboxysomes sequester heterologous and chimeric RuBisCo species. PLoS One 3:e3570. 10.1371/journal.pone.0003570 [DOI] [PMC free article] [PubMed] [Google Scholar]
56. Cheng S, Fand C, Sinha S, Bobik TA. 2012. The PduQ enzyme is an alcohol dehydrogenase used to recycle NAD⁺ internally within the Pdu microcompartment of Salmonella enterica. PLoS One 7:e47144. 10.1371/journal.pone.0047144 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B1] 1. Bobik TA, Havemann GD, Busch RJ, Williams DS, Aldrich HC. 1999. The propanediol utilization (pdu) operon of Salmonella enterica serovar Typhimurium LT2 includes genes necessary for formation of polyhedral organelles involved in coenzyme B12-dependent 1,2-propanediol degradation. J. Bacteriol. 181:5967–5975 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B2] 2. Kofoid E, Rappleye C, Stojiljkovic I, Roth J. 1999. The 17-gene ethanolamine (eut) operon of Salmonella typhimurium encodes five homologues of carboxysome shell proteins. J. Bacteriol. 181:5317–5329 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B3] 3. Shively JM, Ball F, Brown DH, Saunders RE. 1973. Functional organelles in prokaryotes: polyhedral inclusions (carboxysomes) of Thiobacillus neapolitanus. Science 182:584–586 [DOI] [PubMed] [Google Scholar]

[B4] 4. Penrod JT, Roth JR. 2006. Conserving a volatile metabolite: a role for carboxysome-like organelles in Salmonella enterica. J. Bacteriol. 188:2865–2874 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B5] 5. Sampson EM, Bobik TA. 2008. Microcompartments for B₁₂-dependent 1,2-propanediol degradation provide protection from DNA and cellular damage by a reactive metabolic intermediate. J. Bacteriol. 190:2966–2971 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B6] 6. Roof DM, Roth JR. 1988. Ethanolamine utilization in Salmonella typhimurium. J. Bacteriol. 170:3855–3863 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B7] 7. Roof DM, Roth JR. 1989. Functions required for vitamin B₁₂-dependent ethanolamine utilization in Salmonella typhimurium. J. Bacteriol. 171:3316–3323 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B8] 8. Zhu H, Gonzalez R, Bobik TA. 2011. Coproduction of acetaldehyde and hydrogen during glucose fermentation by Escherichia coli. Appl. Environ. Microbiol. 77:6441–6450 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B9] 9. Brinsmade SR, Escalante-Semerena JC. 2004. The eutD gene of Salmonella enterica encodes a protein with phosphotransacetylase enzyme activity. J. Bacteriol. 186:1890–1892 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B10] 10. Bologna F, Campos-Bermudez V, Saavedra D, Andreo C, Drincovich M. 2010. Characterization of Escherichia coli EutD: a phosphotransacetylase of the ethanolamine operon. J. Microbiol. 48:629–636 [DOI] [PubMed] [Google Scholar]

[B11] 11. Price-Carter M, Tingey J, Bobik TA, Roth JR. 2001. The alternative electron acceptor tetrathionate supports B₁₂-dependent anaerobic growth of Salmonella enterica serovar Typhimurium on ethanolamine or 1,2-propanediol. J. Bacteriol. 183:2463–2475 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B12] 12. Starai VJ, Garrity J, Escalante-Semerena JC. 2005. Acetate excretion during growth of Salmonella enterica on ethanolamine requires phosphotransacetylase (EutD) activity, and acetate recapture requires acetyl-CoA synthetase (Acs) and phosphotransacetylase (Pta) activities. Microbiology 151:3793–3801 [DOI] [PubMed] [Google Scholar]

[B13] 13. Brinsmade SR, Paldon T, Escalante-Semerena JC. 2005. Minimal functions and physiological conditions required for growth of Salmonella enterica on ethanolamine in the absence of the metabolosome. J. Bacteriol. 187:8039–8046 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B14] 14. Fan C, Bobik TA. 2011. The N-terminal region of the medium subunit (PduD) packages adenosylcobalamin-dependent diol dehydratase (PduCDE) into the Pdu microcompartment. J. Bacteriol. 193:5623–5628 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B15] 15. Fan C, Cheng S, Sinha S, Bobik TA. 2012. Interactions between the termini of lumen enzymes and shell proteins mediate enzyme encapsulation into bacterial microcompartments. Proc. Natl. Acad. Sci. U. S. A. 109:14995–15000 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B16] 16. Cheng S, Sinha S, Fan C, Liu Y, Bobik TA. 2011. Genetic analysis of the protein shell of the microcompartments involved in coenzyme B₁₂-dependent 1,2-propanediol degradation by Salmonella. J. Bacteriol. 193:1385–1392 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B17] 17. Davis RW, Botstein D, Roth JR. 1980. Advanced bacterial genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY [Google Scholar]

[B18] 18. Schmieger H. 1971. A method for detection of phage mutants with altered transducing ability. Mol. Gen. Genet. 110:378–381 [DOI] [PubMed] [Google Scholar]

[B19] 19. Penrod JT, Mace CC, Roth JR. 2004. A pH-sensitive function and phenotype: evidence that EutH facilitates diffusion of uncharged ethanolamine in Salmonella enterica. J. Bacteriol. 186:6885–6890 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B20] 20. Datsenko KA, Wanner BL. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. U. S. A. 97:6640–6645 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B21] 21. Yu D, Sawitzke JA, Ellis H, Court DL. 2003. Recombineering with overlapping single-stranded DNA oligonucleotides: testing a recombination intermediate. Proc. Natl. Acad. Sci. U. S. A. 100:7207–7212 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B22] 22. Lawes M, Maloy S. 1995. MudSacI, a transposon with strong selectable and counterselectable markers: use for rapid mapping of chromosomal mutations in Salmonella typhimurium. J. Bacteriol. 177:1383–1387 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B23] 23. Berkowitz D, Hushon JM, Whitfield HJ, Roth J, Ames BN. 1968. Procedure for identifying nonsense mutations. J. Bacteriol. 96:215–220 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B24] 24. Neidhardt FC, Bloch PL, Smith DF. 1974. Culture medium for enterobacteria. J. Bacteriol. 119:736–747 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B25] 25. Roof DM, Roth JR. 1992. Autogenous regulation of ethanolamine utilization by a transcriptional activator of the eut operon in Salmonella typhimurium. J. Bacteriol. 174:6634–6643 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B26] 26. Miller JH. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY [Google Scholar]

[B27] 27. Sheppard DE, Roth JR. 1994. A rationale for autoinduction of a transcriptional activator: ethanolamine ammonia-lyase (EutBC) and the operon activator (EutR) compete for adenosyl-cobalamin in Salmonella typhimurium. J. Bacteriol. 176:1287–1296 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B28] 28. Ripley LS. 1982. Model for the participation of quasi-palindromic DNA sequences in frameshift mutation. Proc. Natl. Acad. Sci. U. S. A. 79:4128–4132 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B29] 29. Hughes KT, Roth JR. 1984. Conditionally transposition-defective derivative of Mu d1(Amp Lac). J. Bacteriol. 159:130–137 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B30] 30. Wolfe AJ. 2005. The acetate switch. Microbiol. Mol. Biol. Rev. 69:12–50 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B31] 31. Dailly Y, Mat-Jan F, Clark DP. 2001. Novel alcohol dehydrogenase activity in a mutant of Salmonella able to use ethanol as sole carbon source. FEMS Microbiol. Lett. 201:41–45 [DOI] [PubMed] [Google Scholar]

[B32] 32. Dailly YP, Bunch P, Clark DP. 2000. Comparison of the fermentative alcohol dehydrogenases of Salmonella typhimurium and Escherichia coli. Microbios 103:179–196 [PubMed] [Google Scholar]

[B33] 33. Sheppard DE, Penrod JT, Bobik T, Kofoid E, Roth JR. 2004. Evidence that a B₁₂-adenosyl transferase is encoded within the ethanolamine operon of Salmonella enterica. J. Bacteriol. 186:7635–7644 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B34] 34. Thiennimitr P, Winter SE, Winter MG, Xavier MN, Tolstikov V, Huseby DL, Sterzenbach T, Tsolis RM, Roth JR, Baumler AJ. 2011. Intestinal inflammation allows Salmonella to use ethanolamine to compete with the microbiota. Proc. Natl. Acad. Sci. U. S. A. 108:17480–17485 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B35] 35. Brown TDK, Jones-Mortimer MC, Kornberg HL. 1977. The enzymic interconversion of acetate and acetyl-coenzyme A in Escherichia coli. J. Gen. Microbiol. 102:327–336 [DOI] [PubMed] [Google Scholar]

[B36] 36. LeVine SM, Ardeshir F, Ames GF. 1980. Isolation and characterization of acetate kinase and phosphotransacetylase mutants of Escherichia coli and Salmonella typhimurium. J. Bacteriol. 143:1081–1085 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B37] 37. Clark DP, Cronan JP., Jr 1996. Two-carbon compounds and fatty acids as carbon sources, p 343–357 In Neidhardt FC, Curtiss R, III, Ingraham JL, Lin ECC, Low KB, Magasanik B, Reznikoff WS, Riley M, Schaechter M, Umbarger HE. (ed), Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed, vol 1 American Society for Microbiology, Washington, DC [Google Scholar]

[B38] 38. Kumari S, Tishel R, Eisenbach M, Wolfe AJ. 1995. Cloning, characterization, and functional expression of acs, the gene which encodes acetyl coenzyme A synthetase in Escherichia coli. J. Bacteriol. 177:2878–2886 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B39] 39. Toraya T. 2003. Radical catalysis in coenzyme B₁₂-dependent isomerization (eliminating) reactions. Chem. Rev. 103:2095–2128 [DOI] [PubMed] [Google Scholar]

[B40] 40. Mori K, Bando R, Hieda N, Toraya T. 2004. Identification of a reactivating factor for adenosylcobalamin-dependent ethanolamine ammonia lyase. J. Bacteriol. 186:6845–6854 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B41] 41. Clark D, Cronan JE. 1980. Escherichia coli mutants with altered control of alcohol dehydrogenase and nitrate reductase. J. Bacteriol. 141:177–183 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B42] 42. Shafqat J, Höög J-O, Hjelmqvist L, Oppermann UCT, Ibáñez C, Jörnvall H. 1999. An ethanol-inducible MDR ethanol dehydrogenase/acetaldehyde reductase in Escherichia coli. Eur. J. Biochem. 263:305–311 [DOI] [PubMed] [Google Scholar]

[B43] 43. Winter SE, Thiennimitr P, Winter MG, Butler BP, Huseby DL, Crawford RW, Russell JM, Bevins CL, Adams LG, Tsolis RM, Roth JR, Baumler AJ. 2010. Gut inflammation provides a respiratory electron acceptor for Salmonella. Nature 467:426–429 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B44] 44. Lawrence JG, Roth JR. 1996. Evolution of coenzyme B₁₂ synthesis among enteric bacteria: evidence for loss and reacquisition of a multigene complex. Genetics 142:11–24 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B45] 45. Anjum MF, Marooney C, Fookes M, Baker S, Dougan G, Ivens A, Woodward MJ. 2005. Identification of core and variable components of the Salmonella enterica subspecies I genome by microarray. Infect. Immun. 73:7894–7905 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B46] 46. Patil MD, Parhad NM. 1986. Growth of salmonellas in different enrichment media. J. Appl. Microbiol. 61:19–24 [DOI] [PubMed] [Google Scholar]

[B47] 47. Rambach A. 1990. New plate medium for facilitated differentiation of Salmonella spp. from Proteus spp. and other enteric bacteria. Appl. Environ. Microbiol. 56:301–303 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B48] 48. Buan NR, Suh S-J, Escalante-Semerena JC. 2004. The eutT gene of Salmonella enterica encodes an oxygen-labile, metal-containing ATP:corrinoid adenosyltransferase enzyme. J. Bacteriol. 186:5708–5714 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B49] 49. Dou Z, Heinhorst S, Williams EB, Murin CD, Shively JM, Cannon GC. 2008. CO₂ fixation kinetics of Halothiobacillus neapolitanus mutant carboxysomes lacking carbonic anhydrase suggest the shell acts as a diffusional barrier for CO₂. J. Biol. Chem. 283:10377–10384 [DOI] [PubMed] [Google Scholar]

[B50] 50. Havemann GD, Bobik TA. 2003. Protein content of polyhedral organelles involved in coenzyme B₁₂-dependent degradation of 1,2-propanediol in Salmonella enterica serovar Typhimurium LT2. J. Bacteriol. 185:5086–5095 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B51] 51. Cheng S, Bobik TA. 2010. Characterization of the PduS cobalamin reductase of Salmonella enterica and its role in the Pdu microcompartment. J. Bacteriol. 192:5071–5080 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B52] 52. Tanaka S, Sawaya MR, Yeates TO. 2010. Structure and mechanisms of a protein-based organelle in Escherichia coli. Science 327:81–84 [DOI] [PubMed] [Google Scholar]

[B53] 53. Kerfeld CA, Sawaya MR, Tanaka S, Nguyen CV, Phillips M, Beeby M, Yeates TO. 2005. Protein structures forming the shell of primitive bacterial organelles. Science 309:936–938 [DOI] [PubMed] [Google Scholar]

[B54] 54. Crowley CS, Cascio D, Sawaya MR, Kopstein JS, Bobik TA, Yeates TO. 2010. Structural insight into the mechanisms of transport across the Salmonella enterica Pdu microcompartment shell. J. Biol. Chem. 285:37838–37846 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B55] 55. Menon BB, Dou Z, Heinhorst S, Shively JM, Cannon GC. 2008. Halothiobacillus neapolitanus carboxysomes sequester heterologous and chimeric RuBisCo species. PLoS One 3:e3570. 10.1371/journal.pone.0003570 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B56] 56. Cheng S, Fand C, Sinha S, Bobik TA. 2012. The PduQ enzyme is an alcohol dehydrogenase used to recycle NAD⁺ internally within the Pdu microcompartment of Salmonella enterica. PLoS One 7:e47144. 10.1371/journal.pone.0047144 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Evidence that a Metabolic Microcompartment Contains and Recycles Private Cofactor Pools

Douglas L Huseby

John R Roth

Abstract

INTRODUCTION

Fig 1.

MATERIALS AND METHODS

Strains and crosses.

Table 1.

Chemicals and growth media.

General growth conditions.

Table 2.

Table 4.

Growth of a eutD mutant.

Selecting Eut+ suppressors of a eutD mutant.

β-Galactosidase assays.

Use of NMR to assay metabolite concentrations in growth media.

Growth on acetate and ethanol.

RESULTS

An unexpected phenotype of eutD mutants.

Fig 2.

Selecting suppressors of eutD mutations.

Evidence that the eutD mutation and its suppressors do not affect expression of distal genes in the operon.

Table 3.

Effects of constructed deletion mutations on eutD phenotype.

Compounds excreted by cells during growth on ethanolamine.

Fig 3.

Model for ethanolamine catabolism.

Fig 4.

Testing predictions of the model.

Growth phenotypes on ethanolamine.

Growth phenotypes on acetate.

Fig 5.

Fig 6.

Table 5.

Growth phenotypes on ethanol.

Fig 7.

Fig 8.

Table 6.

DISCUSSION

Table 7.

ACKNOWLEDGMENTS

Footnotes

REFERENCES

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Selecting Eut⁺ suppressors of a eutD mutant.