* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download From CO2 to cell: energetic expense of creating biomass using the
Basal metabolic rate wikipedia , lookup
Artificial gene synthesis wikipedia , lookup
Photosynthetic reaction centre wikipedia , lookup
Biochemical cascade wikipedia , lookup
Fatty acid synthesis wikipedia , lookup
Fatty acid metabolism wikipedia , lookup
Biosynthesis wikipedia , lookup
Amino acid synthesis wikipedia , lookup
Carbon sink wikipedia , lookup
Metabolic network modelling wikipedia , lookup
Isotopic labeling wikipedia , lookup
Nicotinamide adenine dinucleotide wikipedia , lookup
Pharmacometabolomics wikipedia , lookup
Electron transport chain wikipedia , lookup
Light-dependent reactions wikipedia , lookup
NADH:ubiquinone oxidoreductase (H+-translocating) wikipedia , lookup
Adenosine triphosphate wikipedia , lookup
Metabolomics wikipedia , lookup
Biosequestration wikipedia , lookup
Evolution of metal ions in biological systems wikipedia , lookup
Photosynthesis wikipedia , lookup
Oxidative phosphorylation wikipedia , lookup
Biochemistry wikipedia , lookup
FEMS Microbiology Letters, 363, 2016, fnw054 doi: 10.1093/femsle/fnw054 Advance Access Publication Date: 2 March 2016 Research Letter R E S E A R C H L E T T E R – Physiology & Biochemistry From CO2 to cell: energetic expense of creating biomass using the Calvin–Benson–Bassham and reductive citric acid cycles based on genome data Department of Integrative Biology, University of South Florida, 4202 East Fowler Avenue, Tampa, FL 33620, USA ∗ Corresponding author: Department of Integrative Biology, University of South Florida, 4202 East Fowler Avenue, Tampa, FL 33620, USA. Tel: 813-974-5173; Fax: 813-974-3263; E-mail: [email protected] One sentence summary: The first genome sequence-based estimate of the cost to build cells from CO2 using two common autotrophic pathways predicts that the Calvin–Benson–Bassham cycle is 122% as expensive as the reductive citric acid cycle. Editor: Rich Boden ABSTRACT The factors driving the dominance of the Calvin–Benson–Bassham cycle (CBB) or reductive citric acid cycle (rCAC) in autotrophic microorganisms in different habitats are debated. Based on costs for synthesizing a few metabolic intermediates, it has been suggested that the CBB poses a disadvantage due to higher metabolic cost. The purpose of this study was to extend this estimate of cost from metabolite synthesis to biomass synthesis. For 12 gammaproteobacteria (CBB) and five epsilonproteobacteria (rCAC), the amount of ATP to synthesize a gram of biomass from CO2 was calculated from genome sequences via metabolic maps. The eleven central carbon metabolites needed to synthesize biomass were all less expensive to synthesize via the rCAC (66%–89% of the ATP needed to synthesize them via CBB). Differences in cell compositions did result in differing demands for metabolites among the organisms, but the differences in cost to synthesize biomass were small among organisms that used a particular pathway (e.g. rCAC), compared to the difference between pathways (rCAC versus CBB). The rCAC autotrophs averaged 0.195 moles ATP per g biomass, while their CBB counterparts averaged 0.238. This is the first in silico estimate of the relative expense of both pathways to generate biomass. Keywords: autotroph; deep-sea; hydrothermal; Calvin cycle; reductive citric acid cycle; carbon fixation INTRODUCTION The vast majority of biomass carbon begins its transit through the biosphere via autotrophic carbon fixation pathways. Given the habitat, phylogenetic and physiological diversity of autotrophic organisms, the diversity of biochemistries that can catalyze autotrophic carbon fixation is perhaps unsurprising; six pathways have been described (reviewed in Berg 2011). The Calvin–Benson–Bassham cycle (CBB) and reductive citric acid cycle (rCAC) are present in several phyla of Bacteria; the hydroxypropionate bicycle is present in Chloroflexi; the 3-hydroxypropionate/butyrate and the dicarboxylate/4- hydroxybutarate cycles operate in Crenarchaea; and the WoodLjungdahl pathway is present in both Archaea and Bacteria (Berg 2011). Each autotrophic pathway dovetails with central carbon metabolism to synthesize all of the metabolic intermediates necessary for cellular biosynthesis (e.g. pyruvate, oxaloacetate, et cetera; Fig. 1). In autotrophs, all of the carbon atoms in these metabolic intermediates originated from CO2 and/or HCO3 − , with electrons provided from cellular reductant pools (e.g. NADH, NADPH, reduced ferredoxin). Regardless of pathway, each metabolic intermediate carries the same number of carbon atoms and acquires the same number of electrons from cellular reductant Received: 22 January 2016; Accepted: 29 February 2016 C FEMS 2016. All rights reserved. For permissions, please e-mail: [email protected] 1 Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016 Mary Mangiapia and Kathleen Scott∗ 2 FEMS Microbiology Letters, 2016, Vol. 363, No. 7 pools (e.g. three carbons and 10 electrons for pyruvate). However, the enzymes responsible for the synthesis of each compound from CO2 /HCO3 − vary among the different autotrophic pathways, and introduce differences in the energetic requirements for the synthesis of these molecules. For example, phosphoenolpyruvate synthesis from CO2 via the CBB consumes ATP due to the activities of two kinases unique to the CBB (phosphofructokinase and sedoheptulose-7-phosphate kinase), while its synthesis via the rCAC requires ATP consumption via pyruvate:water dikinase or phosphoenolpyruvate carboxykinase (Fig. 1; Tables S1 and 2, Supporting Information). Additionally, the energetic requirements for the synthesis of each compound vary based on the identity of the cellular electron donor used by the enzymes catalyzing its synthesis; electron donors with more negative redox potentials (e.g. ferredoxin) are more energetically expensive to reduce (Fuchs 2011). Given these differences in energetic expense, it has been suggested that some pathways might be more favorable in energy-limited or energy-replete habitats (Berg 2011; Boyle 2011; Fuchs 2011; Bar-Even, Noor and Milo 2012; Konneke et al. 2014). Deep-sea hydrothermal vents are inhabited by autotrophic microorganisms that use a diversity of biochemistries for carbon fixation, and therefore provide a context for comparing the potential advantages and disadvantages of these biochemistries for their host organisms. The microbes responsible for primary productivity at these sites rely on geothermally-generated electron donors (e.g. H2 S, H2 ) supplied by hydrothermal fluid emitted from the crust (Sievert and Vetriani 2012). Sulfur-oxidizing autotrophic gamma- and epsilonproteobacteria are some of the dominant microorganisms at these sites. The gammaproteobacterial sulfur oxidizers include Thiomicrospiras, which are common isolates from many sulfide-rich habitats, including deep-sea hydrothermal vents in the Atlantic and Pacific Oceans (Ruby, Wirsen and Jannasch 1981; Muyzer et al. 1995; Takai et al. 2004), shallower marine sediments from temperate (Kuenen and Veldkamp 1972; Brinkhoff et al. 1998; Brinkhoff et al. 1999c), Arctic (Knittel et al. 2005) and Antarctic sites (Mikucki and Priscu 2007) and hypersaline lakes and springs (Brinkhoff and Muyzer 1997; Sorokin et al. 2006). The epsilonproteobacterial sulfur oxidizers inhabit many of these same habitats (Campbell et al. 2006). At hydrothermal vents, the epsilonproteobacterial sulfur-oxidizers typically are numerically dominant (LopezGarcia et al. 2003; Campbell et al. 2013); however, sometimes the gammaproteobacterial sulfur-oxidizers prevail (Brazelton and Baross 2010). The factors proposed to drive the relative dominance of these two groups of bacteria include epsilonproteobacterial sensitivity to oxygen and versatility with respect to electron donors and acceptors (Campbell et al. 2006). Another trait that distinguishes these two groups of bacteria are the pathways they use for autotrophic carbon fixation. Most of the sulfur-oxidizing vent gammaproteobacteria use the CBB, Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016 Figure 1. Central carbon metabolism for autotrophic microorganisms using the rCAC (left) or CBB cycle (right) for carbon fixation. Key central carbon metabolites are in red, and for CBB autotrophs, the citric acid cycle is designated as a ‘cycle’ since many autotrophs do not operate a complete oxidative citric acid cycle. Mangiapia and Scott MATERIALS AND METHODS Sulfur-oxidizing autotrophic gamma- and epsilonproteobacterial organisms were selected based on the quality of their sequence data (complete genome or high-quality draft). Repre- senting gammaproteobacterial CBB autotrophs, Thiomicrospira crunogena, KP2, JR2, Milos T1 and T2 and MA2-6 were isolated from hydrothermal vents (Jannasch et al. 1985; Muyzer et al. 1995; Brinkhoff et al. 1999c), while T. pelophila, arctica, chilensis and kuenenii originated from coastal sediments (Kuenen and Veldkamp 1972; Brinkhoff et al. 1999a,b; Knittel et al. 2005), T. halophila was isolated from a hypersaline lake (Sorokin et al. 2006), and Hydrogenovibrio marinus was isolated from surface ocean water (Nishihara, Igarashi and Kodama 1991). The epsilonproteobacterial rCAC autotrophs included in this study are hydrothermal vent isolates Sulfurimonas autotrophica, Nitratiruptor Sb155-2 and Sulfurovum NBC37-1 (Inagaki et al. 2003; Nakagawa et al. 2005), pelagic isolate S. gotlandica GD1 (Labrenz et al. 2013) and sediment isolate S. denitrificans (Timmer-Ten Hoor 1975). Overview of calculations Estimating the cost of synthesizing a cell from CO2, using the CBB or rCAC, was informed by genome data via KEGG maps (http://www.genome.jp/kegg/ and http://img.jgi.doe.gov/; Kanehisa and Goto 2000; Markowitz et al. 2010; Table S3, Supporting Information), and supplemented with relevant literature. This estimate began by calculating the cost (ATP, reducing equivalents, CO2 ) for synthesizing a mole of each of the key central carbon metabolites from CO2 (Fig. 1; Tables S1–3, Supporting Information). This was followed by calculating the cost of synthesizing building blocks (e.g. amino acids, nucleotides, fatty acids) from these metabolites, and then the cost of polymerizing these building blocks into macromolecules (Fig. 2). To calculate the final cost of synthesizing the cells, the costs of the macromolecules were normalized to their relative amounts in a cell. The macromolecular composition of the cells was based on data from E. coli (Neidhardt, Ingraham and Schaechter 1990), since these analyses have not been conducted on the taxa included in this study. Further, details of the calculations are provided in Supplementary Methods. RESULTS AND DISCUSSION Central carbon metabolites from CO2 The amount of ATP and reducing equivalents needed to synthesize central carbon metabolites from CO2 is consistent with those calculated previously (Tables S1 and 2, Supporting Information; (Berg et al. 2010; Boyle 2011; Bar-Even, Noor and Milo 2012; Konneke et al. 2014). For organisms using either carbon fixation pathway, the number of electrons per mol of metabolite produced is the same, but the cellular electron donors differ (Table 1). For example, all the electrons for pyruvate synthesis are provided by NADH when the CBB cycle is used, while use of the rCAC requires other electron donors (e.g. PFOR utilizes reduced ferredoxin). Once electron donors are converted to ‘ATP equivalents’ to account for the energetic expense of reducing them (see Supplementary Methods), it is apparent that the central carbon metabolites are more expensive to synthesize for the gammaproteobacteria (Thiomicrospiras and H. marinus), which use the CBB, compared to the epsilonproteobacteria, which use the rCAC (Fig. 3). Energetic expense for central carbon metabolites averaged ∼80% (rCAC/CBB), with largest differences (rCAC/CBB) for 2-oxoglutarate (66%) and acetyl-CoA (68%), and smallest for fumarate (88%; Fig. 4). Driving this increased expense are the dephosphorylations of bisphosphorylated Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016 while the vent epsilonproteobacteria use the rCAC (Fig. 1; Felbeck 1981; Hügler et al. 2005; Takai et al. 2005; Scott et al. 2006). The enzymes of the CBB cycle are oxygen-tolerant, but oxygen is a competitive substrate with CO2 for RubisCO and results in a wasteful oxygenase reaction. Aerobic cells must minimize the chances of the oxygenase reaction, which results in the loss of a carbon dioxide molecule as well as the consumption of ATP to regenerate RuBP (Tabita et al. 2008). The rCAC is even more sensitive to oxygen than the CBB. The two oxidoreductases necessary for this pathway (pyruvate: ferredoxin oxidoreductase, PFOR and 2-oxoglutarate: ferredoxin oxidoreductase) are typically inactivated by oxygen, which generally restricts this cycle to anaerobes and microaerobes (Berg 2011). However, some rCAC autotrophs (Hydrogenbacter thermophilus and Aquifex pyrophilus) live in microaerobic environments and may have oxygen tolerant versions of these enzymes (Shiba et al. 1985; Beh et al. 1993). Accordingly, sensitivity to oxygen has also been suggested to be a major driver in the distribution of these two autotrophic pathways (Berg 2011). At hydrothermal vents, multiple factors are likely to dictate the distribution of the autotrophs and their biochemistries. Given that dissolved oxygen concentrations can vary from 0 to ∼100 μM over spatial scales of less than 1 cm and temporal scales ranging from seconds to days, oxygen sensitivity is likely to play a role. Additionally, electron donor availabilities are also quite variable; for example, sulfide concentrations also range from 0 to ∼100 μM and vary over the same distances and timescales as oxygen (Fisher et al. 1988; Johnson, Childress and Beehler 1988; Goffredi et al. 1997). Given the variability in redox substrate availability, energetic expense of autotrophic pathways may also factor into the success of different organisms at the vents. The degree to which carbon fixation pathways are more or less expensive (in terms of ATP and cellular reductant use) has been addressed previously, to a limited extent. Previous efforts limited their calculations to the biosynthesis of pyruvate (Berg et al. 2010; Fuchs 2011; Bar-Even, Noor and Milo 2012), calculated the costs of other metabolites in terms of ATP consumption without adding in the cost of the cellular electron donors (Konneke et al. 2014), or kept the tallies of ATP and cellular electron donors separate (Boyle 2011), which complicates comparisons between different autotrophic pathways, since the different pathways utilize different cellular reductants in differing amounts, and these cellular reductants differ in the energy needed to regenerate them (Fuchs 2011). These approaches have been illuminating; however, the objective of autotrophy is the synthesis of cells, not metabolic intermediates, so the application of these prior estimates to understanding physiology or ecology is very problematic. Happily, the recent sequencing of many autotrophic bacterial genomes makes it possible to expand this initial calculation to include all of the metabolic intermediates, building blocks (e.g. amino acids) and macromolecules, while accounting for variations in pathways among organisms. The objective of this work was to use genome data from 17 species of sulfur-oxidizing gamma- and epsilonproteobacteria to compare the expense of synthesizing autotrophic cells from CO2 using either the CBB or rCAC. 3 4 FEMS Microbiology Letters, 2016, Vol. 363, No. 7 Table 1. ATP and cellular electron donors needed to synthesize central carbon metabolites from CO2 using the CBB and rCAC cyclesa . Metabolite Acetyl-CoA Pyruvate Oxaloacetate Fumarate 2-oxoglutarate Phosphoenolpyruvated 3-phosphoglycerate Erythrose-4-P Ribose-5-P Fructose 6-P Sedoheptulose 7-P ATP, CBB ATP, rCAC Electrons, CBB Electrons, rCAC 7 7 8 8 15 8 8 12.33 15 18 21 2 2 3.33 3.33 4.33 4 4 6.67 8.33 10 11.67 4 NADH 5 NADH 5 NADH 5 NADH, 1 Qred b –1 NADPH, 9 NADH 5 NADH 5 NADH 8 NADH 10 NADH 12 NADH 14 NADH 2 Fdred , 1 NADPH, 2 NADH 4 Fdred , 1 NADPH, 2 NADH 4 Fdred , 1 NADPH, 2 NADH 4 Fdred , 1 NADPH, 3 NADH 6 Fdred , 1 NADPH, 4 NADHc 4 Fdred , 1 NADPH, 2 NADH 4 Fdred , 1 NADPH, 2 NADH 5.33 Fdred , 1.33 NADPH, 4 NADH 6.67 Fdred , 1.67 NADPH, 5 NADH 8 Fdred , 2 NADPH, 6 NADH 9.33 Fdred , 2.33 NADPH, 7 NADH a Steps to synthesize these compounds are detailed in Tables S1 and 2 (Supporting Information). Amounts in table are the moles necessary to synthesize 1 mol of each central carbon metabolite. b Qred = quinol (ubiquinol for Thiomicrospiras and H. marinus). c For Sulfurovum sp. NBC37-1 and S. gotlandica the electrons needed for 2-oxoglutarate synthesis are 6 Fdred , 1 NADPH, 3 NADH and 1 Qred (Table S2, Supporting Information). d For S. denitrificans, phosphoenolpyruvate synthesis requires 4.33 ATP, 4 Fdred , 1 NADPH and 2 NADH (Table S2, Supporting Information). compounds necessary to drive the CBB forward (fructose 1,6-bisphosphate, 1,3-bisphosphoglycerate, sedoheptulose 1,7phosphate; File 1, Supporting Information). While dephosphorylation of two of these compounds (fructose 1,6-bisphosphate, 1,3-bisphosphoglycerate; Table S1, Supporting Information) is also necessary for rCAC autotrophs to synthesize the six metabolites produced via gluconeogenesis and the pentose phosphate pathway (Fig. 1), CBB autotrophs have the added bur- den of having to send all of the carbon they fix through multiple dephosphorylation steps in the CBB cycle before feeding them into central carbon metabolism to synthesize all of the 11 central carbon metabolites needed for biosynthesis. Another added expense for CBB autotrophs arises from the syntheses of acetyl-CoA and 2-oxoglutarate, which require oxidative decarboxylations to synthesize them from pyruvate and isocitrate, respectively (Fig. 1). Some of the losses are recouped Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016 Figure 2. Methods overview. Steps used to determine the amount of ATP required to build a cell from CO2 via the rCAC or CBB from CO2 . Mangiapia and Scott 5 as NADH and NADPH via pyruvate dehydrogenase and isocitrate dehydrogenase, but these central carbon metabolites remain more expensive for CBB autotrophs. Interestingly, some of the Thiomicrospiras have genes encoding an oxoacid oxidoreductase (T. crunogena, halophile, JR2 and MA2-6; Tcr 1709 and 1710 in T. crunogena). Cell extracts of T. crunogena reduce benzyl viologen when pyruvate is added, consistent with the presence of PFOR (Thomas Hanson, unpublished data). Use of PFOR could make acetyl-coA and 2-oxoglutarate less expensive, as the reduced ferredoxin is worth more ATP-equivalents (see above). Bucking this trend, a few steps for rCAC-mediated central carbon metabolite synthesis are a bit profligate relative to the CBB. Use of ferredoxin electrons to synthesize pyruvate and 2-oxoglutarate is expensive as described above. The synthesis of phosphoenolpyruvate has the added expense of using phosphoenolpyruvate carboxykinase (S. denitrificans) or pyruvate, water dikinase (Table S2, Supporting Information), which hydrolyzes ATP to AMP and results in a two ATP difference in expense between phosphoenolpyruvate and pyru- vate in rCAC autotrophs. For the CBB autotrophs, the reverse reaction is catalyzed by pyruvate kinase, with a 1 ATP difference in expense between these two compounds (Table S1, Supporting Information). Lastly, some of the rCAC autotrophs (S. autotrophica, S. denitrificans, Nitratiruptor sp. SB155-2) have a fumarate reductase whose predicted subunit composition is consistent with using NADH instead of quinol as the reductant, adding to the expense for synthesis of 2-oxoglutarate and other compounds downstream of this reaction (Table S2, Supporting Information). Macromolecules from central carbon metabolites As expected, protein synthesis has a high demand for pyruvate, oxaloacetate, 2-oxoglutarate, phosphoenolpyruvate and 3-phosphoglycerate for carbon skeletons for amino acids (Table 2; Fig. S1, Supporting Information). Metabolite requirements are high relative to the other macromolecules due in part to the relative abundance of proteins compared to other macro- Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016 Figure 3. Moles of ATP to synthesize each central carbon metabolite. The ATP and cellular reductant necessary for metabolite synthesis from CO2 were tallied (Table 1). As described in Supplementary Methods, to facilitate comparison, cellular reductant (Fdred , NADPH, NADH, Qred ) was converted to equivalents of ATP by estimating their cost to proton potential (3H+ = 1 ATP) if reduced via reverse electron transport (NADH), membrane-bound transhydrogenase (NADPH) and Rnf or NfnAB complex (Fdred ). Values for gammaproteobacteria are depicted as blue-hued bars (Thiomicrospira spp., H. marinus), while those for epsilonproteobacteria are indicated with red-hued bars (S. autotrophica, gotlandica and denitrificans; Nitratiruptor Sb155-2; Sulfurovum NBC37-1). 6 FEMS Microbiology Letters, 2016, Vol. 363, No. 7 molecules in cells. NADPH is required for reductive amination of alphaketoacids, while NADH is produced as a byproduct of some amino acid synthesis pathways (e.g. those synthesized from 3phosphoglycerate: glycine, serine, cysteine). Nucleic acid synthesis draws heavily on oxaloacetate, 3phosphogycerate and ribose 5-phosphate pools, with NADH produced from GMP and pyrimidine synthesis (Table 2; Fig. S1, Supporting Information). With respect to the thioredoxin required for deoxynucleotide synthesis, since the electrons carried by thioredoxin for deoxyribonucleotide biosynthesis originate from NADPH via thioredoxin reductase (Arner and Holmgren 2000), their cost in ATP equivalents is equivalent to that for NADPH electrons. The other macromolecules require heavy input of metabolites needed for their synthesis [polar lipids: acetyl-CoA, 3phosphoglycerate and cellular reductant (NADH and NADPH); peptidoglycan: acetyl-coA, phosphoenolpyruvate and fructose 6-phosphate for the glycans, with pyruvate, oxaloacetate, 2ketoglytarate for the tetrapeptide linkers; lipooligosaccharide: acetyl-CoA and reductant for fatty acids, phosphoenolpyruvate, sugar phosphates for KDO2 and core oligosaccharide; Table 2; Fig. S1, Supporting Information]. Differences among species in the moles of metabolites needed to synthesize proteins, nucleic acids and polar lipids sufficient for 1 g of cells are most apparent for polar lipids (Table 2; Fig. S1, Supporting Information). These differences between taxa are, however, quite small compared to the differences apparent for central carbon metabolite synthesis (Table 1). No difference between species is shown for peptidoglycan or lipooligosaccharide synthesis (Table 2; Fig. S1, Supporting Information), as it is not known how these macromolecules vary among the taxa studied here. Biomass from macromolecules When the metabolites necessary for synthesizing macromolecules were tallied to estimate the amounts necessary for synthesizing a gram of biomass, differences between taxa were small, despite accommodating species-specific differences in macromolecule composition whenever these data were available (Table 2; Fig. S1, Supporting Information). However, when the moles of metabolite needed to synthesize 1 g biomass are multiplied by the moles of ATP needed to synthesize each (Table 1; Fig. 3), differences become apparent (Fig. 4). The added expense of central carbon metabolite synthesis from the CBB (Table 1) is substantial enough to propagate through the differences in cell composition, resulting in calculated expense for 1 g biomass production by rCAC autotrophs that is 82% of the average calculated for CBB autotrophs (Table S4, Supporting Information). Differences among rCAC and CBB autotrophs are quite small (0.194–0.196 moles ATP per g biomass for rCAC; 0.237–0.239 for CBB; Table S4, Supporting Information), as within these two subclusters the differences in macromolecule composition did not have a substantial impact on the relative amounts of central carbon metabolites necessary to synthesize a cell (Table 2). Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016 Figure 4. Moles of ATP needed to synthesize sufficient metabolites for 1 g biomass (dry weight). The moles of metabolites needed to synthesize 1 g biomass (Fig. S1, Supporting Information) were multiplied by the moles of ATP necessary to synthesize them in each organism based on carbon fixation pathway and central carbon metabolism (Table 1). Values for gammaproteobacteria are depicted as blue-hued bars (Thiomicrospira spp., H. marinus), while those for epsilonproteobacteria are indicated with red-hued bars (S. autotrophica, gotlandica and denitrificans; Nitratiruptor Sb155-2; Sulfurovum NBC37-1). Mangiapia and Scott 7 Table 2. Millimoles of metabolites needed to synthesize 1g biomass from CO2 using the CBB and rCAC cyclesa . Proteins Metabolite Acetyl-CoA Pyruvate Oxaloacetate Fumarate 2-oxoglutarate Phosphoenolpyruvate 3-phosphoglycerate Erythrose-4-P Ribose-5-P Fructose 6-P Sedoheptulose 7-P ATP NADPH NADH Nucleic acids Polar lipids Peptidoglycan LOSb Biomass CBB rCAC CBB rCAC CBB rCAC CBB and rCAC CBB and rCAC CBB rCAC 0.44 2.34 1.55 − 0.29 0.85 0.71 0.58 0.36 0.10 0 0 23.40 8.60 − 1.46 0.41 2.34 1.71 − 0.24 0.71 0.81 0.55 0.41 0.09 0 0 23.25 9.01 − 1.36 0 0 0.87 − 0.56 0 0 0.38 0 0.69 0 0 8.75 1.42 7.36 0 0 0.88 − 0.57 0 0 0.38 0 0.69 0 0 8.74 1.44 7.36 1.97 0.01 0 0 0 0 0.23 0 0.03 0 0 2.34 2.03 1.63 1.91 0 0 0 0 0 0.23 0 0 0 0 2.16 1.91 1.58 0.05 0.07 0.02 0 0.02 0.02 0 0 0 0.05 0 0.44 0.19 0 0.54 0 0 0 0 0.03 0 0 0.03 0.03 0.05 0.80 0.47 0.47 3.00 2.42 2.45 −0.85 0.87 0.76 1.20 0.36 0.85 0.08 0.05 35.72 12.71 −0.25 2.91 2.42 2.62 −0.80 0.74 0.86 1.16 0.41 0.80 0.08 0.05 35.38 13.03 −0.20 a CONCLUSIONS This work substantially strengthens the argument for a key advantage of the rCAC, given that it is based not on a few intermediates, but the whole process of cellular biosynthesis, tailored to each organism using genome sequence-informed predictions of all pathways from CO2 to biomass, supplemented with all available information from the literature. Based on this in silico study, biomass production from CO2 is less energetically expensive using the rCAC than the CBB. This difference in expense reflects biosynthetic expense of the central carbon metabolites; surprisingly, differences in biomass composition exerted a relatively weak effect on the final tally of ATP equivalents per g biomass. Given a low-oxygen habitat where it can operate unhindered, the rCAC is predicted to provide a substantial advantage over any CBB competitors, which begs empirical verification: Given identical growth conditions, do rCAC autotrophs have greater growth yields than CBB autotrophs? In order to make such a comparison, organisms would need to be selected based on similarities in their electron transport chains to keep the potential yield per reductant constant. Further, growth conditions would need to be tailored to avoid activating energetically expensive auxiliary capabilities that cells would need to induce to maintain either the rCAC or CBB (e.g. oxidative damage response for rCAC autotrophs; carbon concentrating mechanisms for CBB autotrophs). One possible comparison would be T. crunogena (CBB) and S. autotrophica (rCAC), given the similarities in their electron transport chains: both genomes encode the Sox complex for thiosulfate oxidation (soxABCXYZ genes: Tcr 0601, 1549, 0156, 0604, 0603, 0602; Saut 0994, 0995, 2096, 0991–3; Scott et al. 2006; Sikorski et al. 2010), and both are capable of robust growth using oxygen as a sole terminal electron acceptor (Jannasch et al. 1985; Inagaki et al. 2003), presumably by using cbb3 -type cytochrome c oxidase complexes (Pitcher and Watmough 2004, cbb I – III:Tcr 1965–1963; Saut 1965, 1964, 1962). If both species were cultivated in the presence of elevated concentrations of CO2 and low-oxygen tensions, their growth yields could be compared to the results from this calculation. If indeed the measured growth yields match those predicted from these calculations, they could explain why epsilonproteobacterial autotrophs dominate under anoxic coditions (Campbell et al. 2006; Han and Perner 2015), while Thiomicrospira can dominate in habitats perfused with oxygenated water, e.g. the Lost City hydrothermal field (Brazelton and Baross 2010), at which any advantage in rCAC-mediated growth yield is presumably negated by inactivation by oxygen. An explanation for the dominance of CBB or rCAC in different habitats is key to understanding primary productivity in these habitats, and could be helpful in synthetic biology-based efforts to engineer organisms to generate compounds of industrial interest from CO2 . SUPPLEMENTARY DATA Supplementary data are available at FEMSLE online. ACKNOWLEDGEMENTS The authors are grateful to Sydney Russell for her assistance in assessing metabolic maps in IMG, to Thomas Hanson for use of unpublished data and to anonymous reviewers for their helpful suggestions. FUNDING Sequencing and annotating the Thiomicrospiras genomes was performed by the United States Department of Energy Joint Genome Institute, under contract no. DE-AC02-05CH11231. This work was supported by the National Science Foundation [NSFMCB-0643713 and NSF-IOS-1257532 to KMS]. Conflict of interest. None declared. REFERENCES Arner ESJ, Holmgren A. Physiological functions of thioredoxin and thioredoxin reductase. Eur J Biochem 2000;267:6102–9. Bar-Even A, Noor E, Milo R. A survey of carbon fixation pathways through a quantitative lens. J Exp Bot 2012;63:2325–42. Beh M, Strauss G, Huber R et al. Enzymes of the reductive citricacid cycle in the autotrophic eubacterium Aquifex pyrophilus and in the Archaebacterium Thermoproteus neutrophilus. Arch Microbiol 1993;160:306–11. Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016 Averages from 11 species of gammaproteobacterial CBB autotrophs and five epsilonproteobacterial rCAC autotrophs are presented here. Data from individual species are depicted in (Figs 1 and 2, Supporting Information). b LOS = lipooligosaccharides. 8 FEMS Microbiology Letters, 2016, Vol. 363, No. 7 Jannasch H, Wirsen C, Nelson D et al. Thiomicrospira crunogena sp. nov., a colorless, sulfur-oxidizing bacterium from a deep-sea hydrothermal vent. Int J Syst Bacteriol 1985;35:422–4. Johnson KS, Childress JJ, Beehler CL. Short term temperature variability in the Rose Garden hydrothermal vent field. DeepSea Res 1988;35:1711–22. Kanehisa M, Goto S. KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Res 2000;28:27–30. Knittel K, Kuever J, Meyerdierks A et al. Thiomicrospira arctica sp. nov. and Thiomicrospira psychrophila sp. nov., psychrophilic, obligately chemolithoautotrophic, sulfur-oxidizing bacteria isolated from marine Arctic sediments. Int J Syst Evol Micr 2005;55:781–6. Konneke M, Schubert DM, Brown PC et al. Ammonia-oxidizing archaea use the most energy-efficient aerobic pathway for CO2 fixation. P Natl Acad Sci USA 2014;111:8239–44. Kuenen JG, Veldkamp H. Thiomicrospira pelophila, gen. n., sp. n., a new obligately chemolithoautotrohic colorless sulfur bacterium. Anton Van Lee 1972;38:241–56. Labrenz M, Grote J, Mammitzsch K et al. Sulfurimonas gotlandica sp. nov., a chemoautotrophic and psychrotolerant epsilonproteobacterium isolated from a pelagic redoxcline, and an emended description of the genus Sulfurimonas. Int J Syst Evol Micr 2013;63:4141–8. Lopez-Garcia P, Duperron S, Philippot P et al. Bacterial diversity in hydrothermal sediment and epsilonproteobacterial dominance in experimental microcolonizers at the Mid-Atlantic Ridge. Environ Microbiol 2003;5:961–76. Markowitz VM, Chen I-MA, Palaniappan K et al. The integrated microbial genomes system: an expanding comparative analysis resource. Nucleic Acids Res 2010;38(Database issue):D382– 90. Mikucki JA, Priscu JC. Bacterial diversity associated with Blood Falls, a subglacial outflow from the Taylor Glacier, Antarctica. Appl Environ Microb 2007;73:4029–39. Muyzer G, Teske A, Wirsen CO et al. Phylogenetic relationships of Thiomicrospira species and their identification in deeop-sea hydrothermal vent samples by denaturing gradient gel electrophoresis of 16S rDNA fragments. Arch Microbiol 1995;164:165–72. Nakagawa S, Takai K, Inagaki F et al. Distribution, phylogenetic diversity and physiological characteristics of epsilonProteobacteria in a deep-sea hydrothermal field. Environ Microbiol 2005;7:1619–32. Neidhardt FC, Ingraham JL, Schaechter M. Composition and organization of the bacterial cell. In: Physiology of the Bacterial Cell: A Molecular Approach. Neidhardt FC, Ingraham JL, Schaechter M (eds). Sunderland: Sinauer Associates, Inc., 1990, 1–29. Nishihara H, Igarashi Y, Kodama T. Hydrogenovibrio marinus gen. nov., sp. nov., a marine obligately chemolithoautotrophic hydrogen-oxidizing bacterium. Int J Syst Bacteriol 1991;41:130–3. Pitcher RS, Watmough NJ. The bacterial cytochrome cbb3 oxidases. Biochim Biophys Acta 2004;1655:388–99. Ruby EG, Wirsen CO, Jannasch HW. Chemolithotrophic SulfurOxidizing Bacteria from the Galapagos Rift Hydrothermal Vents. Appl Environ Microb 1981;42:317–24. Scott KM, Sievert SM, Abril FN et al. The genome of deep-sea vent chemolithoautotroph Thiomicrospira crunogena. PLoS Biol 2006;4:1–17. Shiba H, Kawasumi T, Igarashi Y et al. CO2 assimilation via the reductive tricarboxylic-acid cycle in an obligatley autotrophic, aerobic hydrogen-oxidizing bacterium, Hydrogenobacter thermophilus. Arch Microbiol 1985;141:198–203. Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016 Berg I. Ecological aspects of the distribution of different autotrophic CO2 fixation pathways. Appl Environ Microb 2011;77:1925–36. Berg I, Kockelkorn D, Ramos-Vera W et al. Autotrophic carbon fixation in archaea. Nat Rev Microbiol 2010;8:447–59. Boyle NR. Computation of metabolic fluxes and efficiencies for biological carbon dioxide fixation. Metab Eng 2011;11:150–8. Brazelton WJ, Baross JA. Metagenomic comparison of two Thiomicrospira lineages inhabiting contrasting deep-sea hydrothermal environments. PLoS One 2010;5:e13530. Brinkhoff T, Muyzer G. Increased species diversity and extended habitat range of sulfur-oxidizing Thiomicrospira spp. Appl Environ Microb 1997;63:3789–96. Brinkhoff T, Muyzer G, Wirsen CO et al. Thiomicrospira chilensis sp. nov., a mesophilic obligately chemolithoautotrophic sulfuroxidizing bacterium isolated from a Thioploca mat. Int J Syst Evol Micr 1999a;49:875–9. Brinkhoff T, Muyzer G, Wirsesn CO et al. Thiomicrospira kuenenii sp. nov. and Thiomicrospira frisia sp. nov., two mesophilic obligately chemolithoautotrophic sulfur-oxidizing bacteria isolated from an intertidal mud flat. Int J Syst Evol Micr 1999b;49:385–92. Brinkhoff T, Santegoeds CM, Sahm K et al. A ployphasic approach to study the diversity and vertical distribution of sulfur-oxidizing Thiomicrospira species in coastal sediments of the german wadden sea. Appl Environ Microb 1998;64: 4650–7. Brinkhoff T, Sievert SM, Kuever J et al. Distribution and diversity of sulfur-oxidizing Thiomicrospira spp. at a shallow-water hydrothermal vent in the Aegean Sea (Milos, Greece). Appl Environ Microb 1999c;65:3843–9. Campbell BJ, Engel AS, Porter ML et al. The versatile epsilonproteobacteria: key players in sulphidic habitats. Nat Rev Microbiol 2006;4:458–68. Campbell BJ, Polson SW, Allen LZ et al. Diffuse flow environments within basalt- and sediment-based hydrothermal vent ecosystems harbor specialized microbial communities. Front Microbiol 2013;4:15. Felbeck H. Chemoautotrophic potential of the hydrothermal vent tube worm, Riftia pachyptila Jones (Vestimentifera). Science 1981;213:336–8. Fisher C, Childress J, Arp A et al. Microhabitat variation in the hydrothermal vent mussel, Bathymodiolus thermophilus, at the Rose Garden vent on the Galapagos Rift. Deep-Sea Res 1988;35:1769–91. Fuchs G. Alternative pathways of carbon dioxide fixation: insights into the early evolution of life? Annu Rev Microbiol 2011;65:631–58. Goffredi SK, Childress JJ, Desaulniers NT et al. Inorganic carbon acquisition by the hydrothermal vent tubeworm Riftia pachyptila depends upon high external P-CO2 and upon proton-equivalent ion transport by the worm. J Exp Biol 1997;200:883–96. Han Y, Perner M. The globally widespread genus Sulfurimonas: versatile energy metabolisms and adaptations to redox clines. Front Microbiol 2015;6:989. Hügler M, Wirsen CO, Fuchs G et al. Evidence for autotrophic CO2 fixation via the reductive tricarboxylic acid cycle by members of the epsilon subdivision of proteobacteria. J Bacteriol 2005;187:3020–7. Inagaki F, Takai K, Kobayashi H et al. Sulfurimonas autotrophica gen. nov., sp nov., a novel sulfur-oxidizing epsilonproteobacterium isolated from hydrothermal sediments in the Mid-Okinawa Trough. Int J Syst Evol Micr 2003;53: 1801–5. Mangiapia and Scott Sievert SM, Vetriani C. Chemoautotrophy at deep-sea vents. Oceanography 2012;25:219–33. Sikorski J, Munk C, Lapidus A et al. Complete genome sequence of Sulfurimonas autotrophica type strain (OK10). Stand Genomic Sci 2010;3:194–202. Sorokin DY, Tourova TP, Kolganova TV et al. Thiomicrospira halophila sp. nov., a moderately halophilic, obligately chemolithoautotrophic, sulfur-oxidizing bacterium from hypersaline lakes. Int J Syst Evol Micr 2006;56:2375–80. Tabita FR, Hanson TE, Satagopan S et al. Phylogenetic and evolutionary relationships of RubisCO and the RubisCO-like proteins and the functional lessons provided by diverse molecular forms. Philos T Roy Soc B 2008;363:2629–40. 9 Takai K, Campbell BJ, Cary SC et al. Enzymatic and genetic characterization of carbon and energy metabolisms by deep-sea hydrothermal chemolithoautotrophic isolated of Epsilonproteobacteria. Appl Environ Microb 2005;71:7310–20. Takai K, Hirayama H, Nakagawa T et al. Thiomicrospira thermophila sp nov., a novel microaerobic, thermotolerant, sulfur-oxidizing chemolithomixotroph isolated from a deepsea hydrothermal fumarole in the TOTO caldera, Mariana Arc, Western Pacific. Int J Syst Evol Micr 2004;54: 2325–33. Timmer-Ten Hoor A. A new type of thiosulphate oxidizing, nitrate reducing microorganism: Thiomicrospira denitrificans sp. nov. Neth J Sea Res 1975;9:344–50. Downloaded from http://femsle.oxfordjournals.org/ by guest on April 5, 2016