Department of Cell and Molecular Biology, College of the Environment and Life Sciences, University of Rhode Island, Kingston, Rhode Island, USAGraduate School of Biological and Environmental Sciences, College of the Environment and Life Sciences, University of Rhode Island, Kingston, Rhode Island, USA
Phylogenetic reconstruction of the Shewanella genus based on the concatenated sequences of 661 conserved single-copy genes identified in the full genomes of Shewanella and five outgroup species. Support values based on 100 iterations of bootstrapping are indicated at the internal nodes. Only support values above 80 are shown. The four Shewanella species with available GEMs are marked with blue stars, and WP3 is marked with a red star.
A schematic representation of the carbon utilization pathways for various carbohydrates and their derivatives (blue), amino acids (orange), nucleic acids (red), and small carbon molecules (green) as well as their links to the central carbon metabolism (red arrows). Metabolites are represented as ovals, and metabolic and transport reactions are represented as links between metabolites. Triple arrows linking two compounds indicate that multiple reactions are involved in the conversion of one compound to the other. Genes coding for the main metabolic steps of the carbon uptake pathways are indicated as labels above the links. Abbreviations: 2A3Oxobut, l-2-amino-3-oxobutanoate; 2dr1P, 2-deoxy-d-ribose 1-phosphate; 2dr5P, 2-deoxy-d-ribose 5-phosphate; 2MaACoA, 2-methylacetoacetyl-CoA; 2PG, d-glycerate 2-phosphate; AcCoA, acetyl-CoA; Adn, adenosine; Akg, 2-oxoglutarate; Ala-L, l-alanine; Asp-l, l-aspartate; Chitob, chitobiose; Cytd, cytidine; DAd-2, deoxyadenosine; DCmp, deoxycytidine monophosphate; DCyt, deoxycytidine; DHAP, dihydroxyacetone phosphate; DUri, deoxyuridine; F6P, d-fructose 6-phosphate; G1P, d-glucose 1-phosphate; G3P, glyceraldehyde 3-phosphate; G6P, d-glucose 6-phosphate; Gal, d-galactose; Gam6P, d-glucosamine 6-phosphate; GGluABT, gammaglutamyl-gamma-aminobutyrate; Glc-D, d-glucose; Glu-L, l-glutamate; HmGCoA, hydroxymethylglutaryl-CoA; Ile-L, l-isoleucine; Lac, lactate; Leu-L, l-leucine; Malt, maltose; Maltodex, maltodextrin; MiCit, methylisocitrate; Oaa, oxaloacetate; Ptrc, putrescine; Pyr, pyruvate; R1P, alpha-d-ribose 1-phosphate; R5P, alpha-d-ribose 5-phosphate; Ser-L, l-serine; Succ, succinate; Thr-L, l-threonine; UDP-Glc, UDP-glucose; Uri, uridine.; Val-L, l-valine.
Comparison of experimentally measured and computationally simulated biomass production levels of WP3 grown with different carbon sources. Error bars represent the standard deviations of the experimentally measured biomass concentrations (gDW/liter) from three independent replicates.
(A) A schematic representation of key reactions involved in the production of ATP and PMF in WP3. (B) Comparison of biomass fluxes in the wild-type and Δatp mutant models of WP3 with GlcNac or lactate as a sole carbon source under aerobic and anaerobic conditions. (C) Biomass fluxes from anaerobic growth simulations of the WP3 wild-type model and the ΔackA, Δpta, and Δpta ΔackA mutant models using GlcNac or lactate as a sole carbon source and fumarate as a sole electron acceptor. (D) Internal reaction fluxes of the WP3 and mutant models from the simulations whose results are shown in panel C, using GlcNac as a sole carbon source. MK, menaquinone; CymA, tetraheme c-type cytochrome; ATPase, ATP synthase; Fdh, formate dehydrogenase; Ndh, NADH dehydrogenase; AckA, acetate kinase; Pra, phosphotransacetylase; Pyk, pyruvate kinase; Xpk, xylulose-5-phosphate phosphoketolase.
Comparison of NAD+/NADH homeostasis between WP3 (A) and MR-1 (B). The differences in NAD+ and NADH concentrations were calculated from simulations under anaerobic conditions with fumarate as the terminal electron acceptor for each of the carbon sources shown (x axis).
Exchange reaction constraints representing the concentrations of carbon, nitrogen, sulfur, and phosphorus sources in the minimal medium of WP3 batch culturesa
Flux bound of exchange reaction
Glucose, maltose, GlcNac, or pyruvate
↵a All other exchange reactions in the WP3 model were defined with settings in the basal constraints. The compounds pyruvate, glucose, maltose, and GlcNac were used as sole carbon sources. The lower and upper bounds of exchange reaction fluxes are shown; negative values indicate that uptake of the nutrient was permitted. The concentrations of the sole carbon sources varied from 2 mM to 40 mM; the concentrations of the sulfur, phosphorus, and nitrogen sources were set according to their concentration in the experimental medium.
Metabolic enzymes involved in ATP production and PMF generation, with their corresponding reactions, functional roles, and gene associations in the WP3 modela
swp_5155 AND swp_5156 AND swp_5157 AND swp_5158 AND swp_5159 AND swp_5160 AND swp_5161
NADH4, NADH12, NADH14
swp_1298 OR swp_2117 OR swp_4014
(swp_5024 AND swp_5025 AND swp_5023) OR (swp_5027 AND swp_5028 AND swp_5029)
↵a A schematic of key reactions and comparisons of biomass and reaction fluxes is shown in Fig. 4.
Stoichiometries of the fatty acid components in the lipid biosynthesis equation of the WP3 GEM. The stoichiometries of unsaturated, saturated, and branched-chain fatty acids were calibrated based on experimental measurements of the WP3 fatty acid composition at 20°C and 0.1 MPa (14). Download TABLE S2, PDF file, 0.01 MB.
Basal constraints for metabolic simulations performed in the WP3 and MR-1 models (see Materials and Methods). “Compound ID/Name” lists the identifiers/names of extracellular compounds with defined exchange reactions, which were used to simulate the availability of nutrients and the removal of metabolic by-products. The compound identifiers are shown for both the WP3 and MR-1 models. “Lower/Upper Bound” lists basal constraints for the lower and upper bounds of exchange reaction fluxes. Negative lower bounds indicate compounds provided as nutrient sources to the model, and a lower bound of zero indicates a compound that could only be released as a metabolic by-product but not acquired from the environment. “Type” lists the classification of the exchange compounds. “Growth supporting in WP3” lists the growth-supporting carbon sources, and terminal electron acceptors are marked as TRUE in this column. Download TABLE S3, PDF file, 0.03 MB.
Phylogenetic trees of ArgE and NagB proteins encoded in the genomes of group 1 (A and C) and group 2 (B and D) Shewanella species. Support values based on 100 iterations of bootstrapping are indicated at the internal nodes. Only support values above 80 are shown. The group 1 and group 2 copies of the corresponding proteins had no detectable homology, indicating nonhomologous replacements of the ArgE and NagB functions in the two groups of Shewanella species. Download FIG S1, PDF file, 0.4 MB.
Components of the LMO-812 minimal medium used for the experimental culture of Shewanella piezotolerans WP3. Medium components were adapted from a previously described defined marine medium (F. Widdel, p. 102–104, in HERMES Handbook of Methods for Microbial Ecology, 2005, accessible at https://epic.awi.de/29169/1/HER2005l.pdf). Download TEXT S1, PDF file, 0.1 MB.
Maximum and minimum flux values obtained from flux variability analysis (FVA), corresponding to the simulation conditions of the experiments whose results are shown in Fig. 4D. FVA was performed in the WP3 wild-type model and the Δpta, ΔackA, and Δpta ΔackA mutant models with biomass production set to its maximum (see Materials and Methods). Numbers in this table indicate raw values of the minimum and maximum fluxes before they were normalized by the biomass flux. Download TABLE S4, PDF file, 0.02 MB.
Linear models for the prediction of NAD+/NADH homeostasis in the WP3 model (see Materials and Methods). Fluxes of the ATPase reaction (black dots) were plotted based on a robustness simulation across varied fluxes of the EQ1 reaction. Linear models (red lines) were fitted to the observed correlations between EQ1 and ATPase fluxes and used to calculate the differences in NAD+ and NADH concentrations where the ATPase flux approached zero. Download FIG S3, PDF file, 0.04 MB.
Linear models for the prediction of NAD+/NADH homeostasis in the MR-1 model (see Materials and Methods). Fluxes of the ATPase reaction (black dots) were plotted based on a robustness simulation across varied fluxes of the EQ1 reaction. Linear models (red lines) were fitted to the observed correlations between EQ1 and ATPase fluxes and used to calculate the differences in NAD+ and NADH concentrations where the ATPase flux approached zero. Download FIG S4, PDF file, 0.04 MB.