Exploring the Diversity and Metabolic Potential of CO₂ fixation Mediated by RubisCO in Prokaryotes in the Japan Collection of Microorganisms

Abstract

A genome anal­ysis is essential for identifying valuable microbial resources for future applications. In the present study, we exami­ned potential CO₂-fixing microorganisms based on the presence of the Calvin–Benson–Bassham (CBB) cycle using 6,262 bacterial and 487 archaeal genomes from available cultures in the Japan Collection of Microorganisms (JCM), a well-established culture collection, in October 2023. A total of 306 strains (147 genera, eight phyla) carried CBB cycle genes, and a literature survey showed that 74 genera had experimental evidence of autotrophic growth while 73 lacked supporting information. A phylogenetic anal­ysis of the large subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase (RbcL) identified diverse forms (IA, IB, IC, IE, I+α, II, and III) with distinct metabolic associations; IA was associated with sulfur species oxidation and formed IC with hydrogen oxidation. Genome-based metabolic predictions identified the potential for CO₂ fixation in numerous species lacking experimental evidence. Our anal­yses indicate that members of Actinomycetota harboring IE RbcL were generally associated with hydrogen oxidation, possibly by using oxygen or nitrate as an electron acceptor. Additionally, 12 species in Pseudomonadota contained photosystem II reaction center proteins (PufL and PufM), suggesting phototrophic capabilities. However, the prediction of electron donors failed in some of these species. They may use the CBB cycle to regulate the intracellular redox balance under photoheterotrophic growth. The present results reveal unrecognized autotrophic potential in JCM strains and broaden our knowledge of the diversity of CO₂-fixing microorganisms. Experimental validation will clarify their roles in the global carbon cycle and their potential for biotechnological applications towards environmental sustainability.

The Calvin–Benson–Bassham (CBB) cycle is one of the most fundamental biochemical pathways for CO₂ fixation, accounting for ~95% of fixed carbon on Earth (Field et al., 1998). The CBB cycle is catalyzed by a pivotal enzyme called ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) (Banda et al., 2020; Schulz et al., 2022), which has attracted attention for its potential in industrial gas fermentation and has been genetically engineered for the production of various compounds (e.g., polyhydroxyalkanoates) in photosynthetic organisms (e.g., microalgae) and chemolithoautotrophic bacteria (Pachauri and Mayer, 2015; Nangle et al., 2020; Arora et al., 2023). In addition to the CBB cycle (Benson et al., 1948), six natural CO₂ fixation pathways that convert atmospheric CO₂ to organic carbon have been identified in autotrophic microorganisms: 1) the reverse tricarboxylic acid (rTCA) cycle including a variant (the reversed oxidative TCA cycle (roTCA)) (Evans et al., 1966; Searcy and Lee, 1998); 2) the 3-hydroxypropionate (3HP) bi-cycle (Holo and Øvreås, 1987); 3) the 4-hydroxybutyrate/3-hydroxypropionate (4HB/3HP) cycle (Berg et al., 2007); 4) the dicarboxylate/4-hydroxybutyrate (DC/4HB) cycle (Berg et al., 2007; Huber et al., 2008); 5) the reductive acetyl-CoA (Wood–Ljungdahl [WL]) pathway (Wood and Werkman, 1936); and 6) the reductive glycine pathway. Among these pathways (Sánchez-Andrea et al., 2014), the CBB cycle is distributed in photosynthetic organisms not only in bacteria, but also in Eukaryotes, such as higher plants and eukaryotic alga, in addition to various chemolithoautotrophic bacteria.

The RubisCO family of proteins has been classified into multiple structural and functional forms, including forms I, II, III, and IV, which differ from each other in their catalytic efficiency, substrate specificity, and evolutionary history (Tabita et al., 2008; Prywes et al., 2023). Besides form IV RubisCOs, which are RubisCO-like proteins that catalyze the carboxylation of ribulose-1,5-bisphosphate, all forms of RubisCO are involved in CO₂ fixation. Form II RubisCO is characterized by its poor affinity for CO₂, indicating its adapted function in low-O₂ and high-CO₂ environments (Badger and Bek, 2008). The oxygenase activity of RubisCO competes with its carboxylase activity, which is fundamental for driving CO₂ fixation. Accordingly, form I RubisCOs have recruited a small subunit, the RubisCO small subunit (RbcS), whose encoding gene (rbcS) typically colocalizes in an operon with the gene encoding the large subunit (rbcL). These evolutionary mechanisms are considered to have increased the specificity of RubisCO for CO₂ in ancient eras before oxygen was abundant on Earth (Schulz et al., 2022; Nishihara et al., 2024).

Form I RubisCOs have been further classified into the following subgroups: IA and IC (mostly in Pseudomonadota), IB (in Cyanobacteria, green algae, and plants), ID (in eukaryotic algae), IE (in Actinomycetota), and the Thermus group (Schulz et al., 2022). A recent phylogenetic anal­ysis of the rbcL gene revealed four early-diverging orthologs of form I RubisCO that lacked the rbcS gene: forms I’ (Banda et al., 2020), I-Anaero (Schulz et al., 2022), I-α, and I” (West-Roberts et al., 2021). These variants, identified from metagenome-assembled genomes, may be the key to elucidating how the small subunit of form I RubisCOs evolved to increase CO₂ affinity. Form III RubisCO has been associated with the pentose bisphosphate pathway in some methanogenic archaea lacking phosphoribulokinase (PRK) (Sato et al., 2007; Aono et al., 2012, 2015). However, it remains unclear whether these archaea perform autotrophy via this pathway because methanogens primarily rely on the Wood–Ljungdahl pathway for CO₂ fixation (Kono et al., 2017). More recently, form III RubisCO harboring PRK was reported in the chemolithoautotrophic bacterium, Thermodesulfobium acidiphilum (Thermodesulfobiota), suggesting the involvement of a modified CBB cycle in CO₂ fixation (Frolov et al., 2019). These findings indicate variations in the function of CBB-associated genes in microbial metabolism.

Among phototrophic bacteria, all purple sulfur bacteria, which belong to Gammaproteobacteria, have been reported to fix CO₂ via the CBB cycle (Shimada and Takaichi, 2024). Among purple non-sulfur bacteria, which belong to Alphaproteobacteria and Betaproteobacteria, some species fix CO₂ via the CBB cycle, but also grow well under heterotrophic conditions (Imhoff, 2009). RubisCO-mediated inorganic carbon assimilation in purple non-sulfur bacteria contributes to alleviating an over-reduced state (Shimada and Takaichi, 2024). Aerobic anoxygenic phototrophic bacteria (AAPB), which are widely distributed across Alpha-, Beta-, and Gammaproteobacteria, are obligate aerobes exhibiting heterotrophic growth; however, there is currently no evidence of autotrophic growth in these bacteria (Yurkov and Beatty, 1998). Light-stimulated CO₂ uptake has been reported in some AAPB species, such as Acidiphilium rubrum (Acetobacterales and Alphaproteobacteria) (Kishimoto et al., 1995) and Erythrobacter longus (Sphingomonadales and Alphaproteobacteria) (Shiba, 1984; Shiba and Harashima, 1986; Yurkov and Beatty, 1998). Although CO₂ fixation through the CBB cycle is insufficient to sustain autotrophic growth in AAPB, it may supply additional organic carbon intermediates for heterotrophic growth (Yurkov and Beatty, 1998). Therefore, the presence of CBB-related functional genes does not always correspond to autotrophic growth in phototrophic microorganisms, particularly in AAPB. To comprehensively investigate the activity of the CBB cycle in these bacteria, AAPB needs to be thoroughly exami­ned under culture conditions, with a focus on the uptake of CO₂ and/or CO₂-independent growth.

Despite the increasing number of bacteria that harbor the potential for autotrophic growth being identified through genomic and/or metagenomic studies, few studies have focused on the relationship between large-scale genomic information and culture-based physiological characteristics (Garritano et al., 2022). To effectively utilize microorganisms as biological resources (e.g., for bio-manufacturing), it is essential to integrate genomic information with cultivation data. The Japan Collection of Microorganisms (JCM) is one of the most well-established culture repositories in the world. Similar to other culture collections, such as the Leibniz Institute DSMZ (https://www.dsmz.de) and the Korean Collection for Type Cultures (https://kctc.kribb.re.kr/en/), JCM maintains cultivable species for reproducible experiments, including well-characterized type and non-type strains. According to data from the World Data Centre for Microorganisms (accessed in July 2025), JCM has the second-largest culture collection of archaeal and bacterial type strains (8,599 type strains) after DSMZ (12,927 type strains). The International Code of Nomenclature of Prokaryotes (ICNP) states that the culture collection plays a defining role in the description of novel species, issuing certificates of the deposition and availability of microorganisms, which are required for publishing new taxa. In addition to maintaining (type) strains, culture collections, such as JCM, are responsible for sequencing the genomes of strains, contributing to data in the Type Strain Genome Database (https://gctype.wdcm.org), an initiative under the Global Catalogue of Microorganisms project. As of July 2025, this project has sequenced 6,141 genomes out of 27,107 available type species.

In the present study, we exami­ned (1) potential CO₂-fixing microorganisms from genome-sequenced species in the JCM collection that harbor CBB cycle genes and (2) the forms of RubisCO in these organisms, integrating information from previously reported culture-based data where possible. Furthermore, based on genomic data, we predicted (3) the metabolic potential of species in which autotrophic growth has not yet been documented. The present results afford a foundation for the systematic integration of genomic and laboratory cultivation data, which will provide a robust framework for understanding the diversity and physiology of microorganisms. Moreover, this framework will enable the strategic utilization of microbial resources in basic research and biotechnological applications.

Materials and Methods

Annotation of genomes for identifying CBB cycle genes

The present study analyzed 6,749 available genomes among 13,937 strains held in the JCM collection (both type and non-type). Genomes were collected from the RefSeq database in the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov; Table S1), which represents a broad and taxonomically diverse set of microorganisms. For example, when referenced against the Genome Taxonomy Database Release 226 (April 2025), these genomes in the dataset covered 37 of the 51 valid phyla (1,616 of the 3,558 genera). Only 14 validated phyla were‍ ‍not covered, including Abditibacteriota, Atribacterota, Caldisericota, Calditrichota, Chlamydiota, Chlorobiota, Chrysiogenota, Coprothermobacterota, Cyanobacteriota, Dictyoglomota, Fibrobacterota, Thermomicrobiota, and Vulcanimicrobiota (taxonomy following NCBI).

To evaluate the genetic capacity for the CBB cycle, the collected genomes were annotated using METABOLIC (v4.0) software against the implemented hidden Markov model databases (the Kyoto Encyclopedia of Genes and Genomes [KEGG], KOfam, Pfam, TIGRfam, and custom hidden Markov models) with default options (Zhou et al., 2022). According to the annotation results, species were deemed to potentially harbor CBB cycle genes if they possessed all four selected KEGG module step hits: M00165+01 (K00855 [PRK, prkB]), M00165+02 (K01601 [rbcL, cbbL]), M00165+03 (K00927 [PGK]), and M00165+04 (K05298 [GAPA], K00150 [gap2, gapB], or K00134 [GAPDH, gapA]).

Prediction of the metabolic potential of putative autotrophs

Regarding species predicted to harbor CBB cycle genes through a genomic annotation anal­ysis, their cultivation conditions for CO₂ fixation (electron donors and electron acceptors) were exami­ned by conducting a literature survey of available publications on Google Scholar (https://scholar.google.co.jp), as of December 2024. One representative species of genera with evidence of autotrophic growth and all available strains of genera lacking evidence of autotrophic growth were selected. The metabolic potential for autotrophic growth of strains lacking evidence of autotrophic growth was predicted based on the presence of functional proteins, as annotated using METABOLIC software and cross-referenced using the MetaCyc database (Karp et al., 2019; Table S2). Predictions focused on proteins associated with electron acceptors (oxygen availability or anaerobic respiration using sulfur species and nitrate) and electron donors (hydrogen and sulfur species; Table S2). The predicted protein-coding genes of RubisCO-harboring strains were annotated using the Prokka tool (v1.14; Seemann, 2014). Additionally, photosynthesis-related genes encoding type-II reaction center proteins (pufM and pufL) were identified through a BLAST homology search against biochemically characterized proteins registered in the SwissProtKB database (Boutet et al., 2007).

Phylogenetic anal­ysis of the rbcL gene

The sequences of the rbcL gene were collected from the set of predicted protein-coding genes from species in the JCM collection. To construct a phylogenetic tree, representative sequences were selected through clustering at an 80% identity threshold using the CD-HIT tool (v.4.8.1; Fu et al., 2012). Clustering was performed among representative sequences of the same genus lacking evidence of autotrophic growth, as confirmed through a literature survey. The selected sequences were aligned using the MAFFT program (v7.427) with the options “--localpair” and “--maxiterate 1000” (Katoh et al., 2005), and the alignment was trimmed using the trimAl tool (v1.4.rev22; Capella-Gutiérrez et al., 2009) with a gap threshold of 0.8. The final alignment was used as an input for maximum-likelihood phylogenetic estimations using the IQ-TREE software package (v2.1.3; Nguyen et al., 2014). The tree was constructed using the profile mixture model, which was identified as the best-fit model by the ModelFinder tool (Kalyaanamoorthy et al., 2017). The profile mixture model was implemented in IQ-TREE (v2.1.3), and 1,000 ultrafast bootstrap replicates were performed. Transfer bootstrap support values were calculated using the BOOSTER (v0.1.2) tool (Lartillot and Philippe, 2006).

Results and Discussion

CO₂ fixation ability in species harboring CBB cycle genes

In the present study, we exami­ned the genomes of 6,262 bacterial and 487 archaeal strains, all of which are available in the JCM culture collection (Table S1). These strains were distributed across 1,731 genera within 37 phyla. A total of 306 strains (147 genera within eight phyla) were found to be potentially involved in the CBB cycle based on the presence of selected prodigal modules annotated using METABOLIC software (Zhou et al., 2022) (Table S3). Most of these strains belonged to the phylum Pseudomonadota (207 species), with the second most represented phylum being Actinomycetota (82 species). Regarding archaea, only six species belonging to Methanobacteriota were identified.

We then conducted a literature survey to establish whether strains from the identified genera have been reported to fix CO₂. The strains used for the literature survey are listed in Table S3 as representative species. Among the 147 genera, 74 contained strains with reported CO₂ fixation capabilities, while the remaining 73 lacked supporting information. The genera that lacked supporting information encompassed 173 strains lacking evidence of autotrophic growth, including 15 strains reported to lack the ability for autotrophic growth. The strains in the JCM collection harboring CBB cycle genes and demonstrating CO₂ fixation belonged to 54 genera (of 74 genera, 20 were omitted because the demonstrated species in these genera were unavailable in the JCM culture collection). In addition, possible electron donors and acceptors for CO₂ fixation in these strains were exami­ned as described in Table S4 and summarized in Table 1 (discussed later). The genera reported to exhibit autotrophic growth were distributed across five phyla: Actinomycetota, Thermodesulfobiota, Pseudomonadota, Thermodesulfobacteriota, and Methanobacteriota (Table 1 and S4).

Table 1.

Summary of autotrophs and representative growth conditions for each genus in the JCM culture collection.

Phylum	Order	Genus (representative JCM strain number)¶	Types*	Optimum oxygen concentration range**	optimum pH** (min, max)	optimum temp.** (min, max)	Electron donors in chemolithotrophs/photolithotrophs†	Electron acceptors in chemolithotrophs
Actinomycetota	Acidimicrobiales	Acidimicrobium (JCM 15462)	C	aerobic	2	45, 50	ferrous iron	oxygen
	Mycobacteriales	Mycolicibacterium (JCM 16229)#1	C	ND	ND	22, 30	hydrogen	oxygen
	Pseudonocardiales	Pseudonocardia (JCM 13855)	C	aerobic	7	30	hydrogen	oxygen
Pseudomonadota	Acidiferrobacterales	Acidiferrobacter (JCM 17358)	C	aerobic	1.2, 2	38	ferrous iron, pyrite, sulfur compounds	oxygen, iron (III)
	Acidithiobacillales	Acidithiobacillus (JCM 3867)	C	aerobic	4.5	30	sulfur, thiosulfate	oxygen
	Burkholderiales	Achromobacter (JCM 20676), Azohydromonas (JCM 20723), Cupriavidus (JCM 11282)#3, Derxia (JCM 20996)#4, Herbaspirillum (JCM 21424), Hydrogenophaga (JCM 21412), Noviherbaspirillum (JCM 17723), Roseateles (JCM 15912), Rubrivivax (JCM 21318), Thiomonas (JCM 20425), Variovorax (JCM 20526)	C, P	anaerobic, aerobic, microaerophilic (2.8–12%)	5.5, 9.0	28, 37	hydrogen, sulfur, sulfur compounds	nitrate, oxygen
	Chromatiales	Caldichromatium (JCM 39101), Halochromatium (JCM 14151), Halorhodospira (JCM 39378), Marichromatium (JCM 13911), Nitrosococcus (JCM 30415), Thioalkalivibrio (JCM 11367), Thiohalocapsa (JCM 14780), Thiolapillus (JCM 15507)	C, P	anaerobic, microaerophilic (2%), aerobic	6, 10.5	20, 50	hydrogen, ammonium, sulfur, sulfur compounds	nitrate, oxygen
	Ectothiorhodospirales	Acidihalobacter (JCM 30709), Thioalbus (JCM 15568)	C	anaerobic, aerobic	2, 7.5	28, 37	sulfur, sulfur compounds, ferrous iron, sphalerite, chalcopyrite, arsenopyrite, galena	nitrate, oxygen
	Hydrogenophilales	Hydrogenophilus (JCM 34254)	C	microaerophilic (5%)	7.0, 8.0	55	thiosulfate, sulfur	oxygen
	Hyphomicrobiales	Ancylobacter (JCM 32039), Bradyrhizobium (JCM 10834)#3, Rhodobium (JCM 9337), Rhodoplanes (JCM 14934), Rhodopseudomonas (JCM 11668), Xanthobacter (JCM 34666)	C, P	anaerobic, aerobic, microaerophilic (1%)	6.5, 8.5	22, 38	hydrogen, thiosulfate	oxygen
	Mariprofundales	Mariprofundus (JCM 30585), Ghiorsea (JCM 31637)	C	microaerophilic (0.4 or 1–3%)	6.0, 7.0	20, 25	ferrous iron, hydrogen	oxygen
	Nitrosomonadales	Ferriphaselus (JCM 18545), Parasulfuritortus (JCM 33645), Sideroxyarcus (JCM 39089), Sideroxydans (JCM 14762)#5, Thiobacillus (JCM 3870)	C	anaerobic, aerobic, microaerophilic (1–3%)	5.6, 7.5	25, 35	ferrous iron, sulfur compounds, sulfur	oxygen, nitrate, nitrite, nitrous oxide
	Pseudomonadales	Pseudomonas (JCM 12568)#6	C	microaerophilic (10%)	7	30	hydrogen	oxygen
	Rhodobacterales	Cereibacter (JCM 6121), Paracoccus (JCM 20754), Rhodobacter (JCM 14598), Rhodovulum (JCM 9220)	C, P	anaerobic, aerobic, microaerophilic (10.5%)	7-9.0	30, 37	hydrogen, sulfide, thiosulfate, sulfur	oxygen
	Rhodospirillales	Azospirillum (JCM 1247), Varunaivibrio (JCM 31027)	C	anaerobic, microaerophilic (2 or 5%)	5.5, 7.0	30, 37	hydrogen, sulfur, thiosulfate	nitrate, iron (III), oxygen
	Acetobacterales	Acidomonas (JCM 6891)#2	C	aerobic	3.0, 5.0	30, 32	methanol	oxygen
	Thiotrichales	Hydrogenovibrio (JCM 7688), Sulfurivirga (JCM 13439)	C	microaerophilic (1%,10%)	6, 6.5	37, 55	hydrogen, sulfur, thiosulfate, tetrathionate	oxygen
Thermodesulfobacteriota		Dissulfurimicrobium (JCM 19990)	C	anaerobic	6.0, 6.2	50, 52	sulfur (disproportionation)	sulfur
Thermodesulfobiota	Thermoanaerobacterales	Thermodesulfobium (JCM 11510)	C	anaerobic	4.0, 6.5	50, 55	hydrogen	sulfate
Methanobacteriota	Methanomicrobiales	Methanoculleus (JCM 15517), Methanospirillum (JCM 10133)	C	anaerobic	6.6, 7.4	30, 42	hydrogen	carbon dioxide

^¶, Representative species are indicated in Table S3. Full species names and corresponding data with references are provided in Table S4.

*, Autotrophic growth under chemolithotrophic (C) or phototrophic (P) conditions.

** Data on optimum or tested oxygen concentrations under autotrophic conditions, temperature, or pH were extracted from the reference; ND indicates “Not determined” (could not be found in the literature survey).

^†, Electron donors comprising various reduced sulfur compounds (e.g., sulfide, thiosulfate, and tetrathionate) are collectively referred to as sulfur compounds.

^#1 Autotrophic growth was not observed under conditions with CO₂ and O₂ in the gas phase.

^#2 Growth initially depended on autotrophic CO₂ fixation, and CO₂ assimilation was subsequently coupled with the ribulose monophosphate (RuMP) pathway during the early to mid-log phases of methylotrophic growth.

^#3 The strain used in the experiment demonstrating autotrophic growth is different from the JCM strain.

^#4 Source information is not available.

^#5 Unvalidated species.

^#6 Autotrophic growth via CO₂ oxidation has been reported instead of CO₂ fixation.

The results of the literature survey indicated that despite 21 genera within the phylum Actinomycetota being annotated as putative autotrophic bacteria based on genomic data, only three contained species reported to exhibit autotrophic growth: Acidimicrobium (Acidimicrobiales), Mycolicibacterium (Mycobacteriales), and Pseudonocardia (Pseudonocardiales) (Table 1, S3, and S4). All the species demonstrating autotrophic growth in these three genera were aerobic chemolithoautotrophs, utilizing hydrogen or ferrous irons as electron donors and oxygen as the electron acceptor (Table 1 and S4).

The phylum Pseudomonadota exhibited the most diverse range of metabolic types, including photosynthetic organisms (e.g., purple sulfur or non-sulfur bacteria), reflecting the wide phylogenetic diversity within this phylum. Photoautotrophs were identified in the classes Alphaproteobacteria (Rhodoplanes, Rhodopseudomonas, and Rhodobium in Hyphomicrobiales and Cereibacter, and Rhodobacter and Rhodovulum in Rhodobacterales), Betaproteobacteria (Rubrivivax in Burkholderiales), and Gammaproteobacteria (Caldichromatium, Halochromatium, Marichromatium, Thiohalocapsa, and Halorhodospira in Chromatiales). These photoautotrophs reportedly utilize hydrogen or sulfur species (e.g., thiosulfate and sulfide) as electron donors (Table 1 and S4). Among the chemolithoautotrophs identified, members of the classes “Zetaproteobacteria” (Mariprofundus and Ghiorsea in Mariprofundales) and Hydrogenophilia (Hydrogenophilus in Hydrogenophilales) required microaerophilic conditions (0.4–5% oxygen) for growth; members of the class Alphaproteobacteria (Acetobacterales, Hyphomicrobiales, Rhodobacterales, and Rhodospirillales) required microaerophilic conditions (1–10.5% oxygen) or aerobic conditions. Most bacterial species identified in the classes Betaproteobacteria (Burkholderiales and Nitrosomonadales) and Gammaproteobacteria (Acidiferrobacterales, Chromatiales, Ectothiorhodospirales, Pseudomonadales, and Thiotrichales) also required oxygen (ranging from microaerophilic to aerobic). Betaproteobacteria species specifically required microaerophilic conditions. Furthermore, some species—such as those in the genera Noviherbaspirillum, Rubrivivax, and Thiomonas (Burkholderiales)—exhibited chemolithoautotroph growth only under anaerobic conditions, utilizing nitrate as an electron acceptor (Table 1 and S4).

Species in the phyla Thermodesulfobiota (Thermodesulfobium narugense), Methanobacteriota (Methanoculleus horonobensis and Methanospirillum hungatei), and Thermodesulfobacteriota (Dissulfurimicrobium hydrothermale) have been reported to exhibit autotrophy under anaerobic conditions (Ferry et al., 1974; Mori et al., 2003; Shimizu et al., 2013; Slobodkin et al., 2016; Table‍ ‍1‍ ‍and S4). Since hydrogenotrophic methanogenic Methanobacteriota are known to fix inorganic carbon through methanogenesis via the Wood–Ljungdahl pathway (Borrel et al., 2016), the CBB cycle is not their primary autotrophic mechanism. Regarding D. hydrothermale, a complete gene set for the Wood–Ljungdahl pathway was previously reported in the genome of this bacterial species (Yvenou et al., 2022). While a previous study suggested that‍ ‍the CBB cycle gene set was incomplete in D. hydrothermale, our anal­ysis based on KEGG annotations (GhostKOALA) indicates that D. hydrothermale possesses a complete set of CBB cycle genes (Kanehisa et al., 2016; see later discussion regarding form I-α RbcL).

Forms and physiological characteristics of the RubisCO enzyme in autotrophy

To examine the forms of RubisCO found in strains in the JCM collection and their associated metabolic pathways, RbcL sequences were extracted and subjected to a phylogenetic clustering anal­ysis. The metabolic pathways related to‍ ‍autotrophic growth are shown in Fig. 1. The anal­ysis excluded form IV RubisCOs (RubisCO-like proteins) and archaeal PRK-lacking form III RubisCOs, which are associated with the pentose bisphosphate pathway (Sato et al., 2007; Tabita et al., 2008; Aono et al., 2012, 2015). Various RubisCO forms were identified across multiple species in the JCM collection, including forms IA, IB, IC, IE, I+α, II, and III. In our constructed phylogeny, the reduced phylogeny (Fig. 1 and S1) and the phylogeny including all sequences (Fig. S2) both showed the same topology: the form I-Anaero clade branched off early from the form IA/IB groups and was positioned as a sister clade to the form IC/IE and Thermus clades. A recent study reported the positional instability of the form I-Anaero clade in phylogenetic reconstructions. However, our constructed tree exhibited the same topology as that of the tree constructed based on RbcS sequences (Schulz et al., 2022; Liu et al., 2023). Therefore, instead of discussing the tree topology, we focused on the phylogenetic distribution shown in Fig. 1.

Fig. 1.

A maximum likelihood tree of amino acid sequences of RbcL. Species with confirmed CO₂ fixation were included in the phylogenetic tree as representatives of their respective genera. The constructed phylogenetic trees of RbcL are shown without grouping branches of (a) form IC, IE, I-Thermus, IB, (b) form IA, (c) form II, III, III-like, and I+α. Reference sequences were collected from published data (Badger and Bek, 2008; Schulz et al., 2022; Prywes et al., 2023). The tree was constructed using IQ-TREE with the LG+C40+G4+F substitution model and 1,000 ultrafast bootstrap replicates. Input sequences were trimmed using trimAl (-gt 0.8), resulting in 211 sequences with 438 amino acid positions. Bootstrap branch support values were calculated using BOOSTER (v0.1.2) from ultrafast bootstrap (1,000 replicates). Unsupported branches with lower bootstrap values (≤90%) were collapsed, and branches with bootstrap values <95% were colored red. The scale bar indicates 1.0 changes per amino acid site. The presence of reported metabolic activity associated with autotrophic growth is shown for each strain, based on the data provided in Table S4. See Fig. S1 for the full tree.

Form IA RbcL was identified in 42 strains, of which 23 exhibited the ability to fix CO₂ (Table S5, Fig. 1b). Form IB RbcL was previously reported in a Cyanobacteria species. In the present study, we detected form IB RbcL in Photobacterium alginatilyticum (Pseudomonadota), which suggests that this bacterium obtained rbcL via horizontal gene transfer from Cyanobacteria (Fig. 1a). Form IC was the most widely distributed, found in 97 strains, of which 22 exhibited the ability to fix CO₂ (Table S5, Fig. 1a). Form IE, associated with the Actinomycetota cluster, was identified in 61 strains, of which two exhibited the ability to fix CO₂ (Table S5, Fig. 1a). Although rbcL and other CBB cycle genes were previously shown to be absent in some species (Müller et al., 2016), all eight species in the JCM collection of the genus Thermus possessed the rbcL gene belonging to the form I Thermus clade, which is consistent with previous findings on other members of this genus (Müller et al., 2016; Schulz et al., 2022) (Fig. 1a). To the best of our knowledge, only heterotrophic growth has been observed in Thermus species lacking evidence of autotrophic growth despite the discovery of RubisCO-encoding genes in the genomes of these species. Form II was identified in 33 species, of which 19 have been reported to exhibit the ability to fix CO₂ (Table S5, Fig. 1c). Form III RbcL was only detected in T. narugense (Thermodesulfobiota). The amino acid sequence of T. narugense RbcL shared 94.87% similarity with that of T. acidiphilum RbcL, which has been shown to fix CO₂ through a RubisCO-mediated transaldolase variant of the CBB cycle (Fig. 1c). Since T. narugense possesses a complete gene set of CBB cycle genes, as described in the previous section, it may also be capable of fixing CO₂ similar to T. acidiphilum (Frolov et al., 2019).

Strains that possessed forms IA, IC, and II RbcL mainly belonged to Pseudomonadota (Fig. 1). Although most Actinomycetota RbcL belonged to form IE, the rbcL sequences from six Actinomycetota species belonged to form I (one in form IC, two in IA, and three in II). These sequences were grouped with those from Pseudomonadota in a later-diverging phylogenetic branch, which suggests that these members acquired rbcL via horizontal gene transfer from Pseudomonadota. Among the six Actinomycetota species, autotrophic growth has been tested in two species. Positive autotrophic growth was reported in Acidimicrobium ferrooxidans (form IA RbcL) (Clark and Norris, 1996); however, Acidithrix ferrooxidans (form IC RbcL) has been reported to lack the ability to fix CO₂ (Fig. 1a; Jones and Johnson, 2015). In Actinomycetota, tested strains for autotrophic growth were limited in form I as well as in form IE.

A notable result of our phylogenetic anal­ysis is that rbcL from D. hydrothermale was classified under the form I-α clade. A BLAST search of its complete genome revealed that the bacterium did not possess the rbcS gene (GCF_022026155; Fig. 1c). Three orthologs of form I RbcL that do not coexist with RbcS have been identified from metagenome-assembled genomes, namely, I’(Banda et al., 2020), I-Anaero (Schulz et al., 2022), and I-α (West-Roberts et al., 2021); however, these have not been identified in culturable species. A more detailed understanding of the mechanisms of action of these enzymes using culturable species will provide insights into the evolution of form I RbcL and the co-option of the small subunit. P. alginatilyticum (Pseudomonadota) also does not possess RbcS, and form IB rbcL was apparently acquired through horizontal gene transfer from Cyanobacteria (see above).

To investigate the relationships between metabolic diversity and RubisCO forms, we exami­ned metabolic information on autotrophic growth available in the literature, with a focus on electron donors (hydrogen or sulfur species) and acceptors as well as the capacity for photoautotrophy. Sixteen of the 22 species harboring form IC RbcL utilized hydrogen as an electron donor while four utilized sulfur species (e.g., sulfur, thiosulfate, and S°) (Fig. 1a, Table S4). In contrast, of the 24 species carrying form IA RbcL, three utilized hydrogen and 20 utilized sulfur species as electron donors (Fig. 1c, Table S4). The remaining species utilized other electron donors, such as ferrous iron (Table S4). Additionally, although many species harboring form II RbcL do not carry form I RbcL, all of the reported photoautotrophs harboring form II RbcL also harbored form IA or IC. These results suggest that different RubisCO forms are associated with metabolic variations in the utilization of electron donors, such as hydrogen versus sulfur compounds, despite the two species belonging to different phylogenetic groups (e.g., at the class and order levels).

To obtain a more detailed understanding of the ecological niches of RubisCO-harboring microbial strains, we exami­ned the relationship between RubisCO forms and the environments from which strains in the JCM collection were isolated, with a focus on strains carrying RbcL (Table S5). Although most of the strains carrying both forms IA and IC RbcL belonged the same phylum of Pseudomonadota, they differed in the distribution of the environments from which they were isolated. Of the 42 form IA–harboring strains, 18 were isolated from marine environments, 21 from terrestrial environments (including six from rhizospheres), two from engineered environments (e.g., wastewater), and one from an organism-hosted environment. In contrast, of the 97 form IC–harboring strains, 16 were isolated from marine environments, 51 from terrestrial environments (including 46 from rhizospheres), 17 from organism-hosted environments, 11 from engineered environments, and one from food. Fourteen of the 17 form IC-harboring strains isolated from organism-hosted environments came from plant-associated environments (e.g., roots). Collectively, strains isolated from plant-associated environments (roots, leaves, and the rhizosphere) accounted for 46 of the 97 form IC–harboring strains, in contrast to only six of the 42 form IA–harboring strains. These results suggest that form IC–harboring strains are more commonly associated with plant-related environments.

Metabolic predictions of potential autotrophs

As described above, 173 of the 306 strains possessing RubisCO genes lacked the capacity for CO₂ fixation or their CO₂ fixation ability had not yet been experimentally tested. Moreover, 143 of the 173 species belonged to genera in which no autotrophic strains have been reported (Table S5). Of the 143 strains, 77 encompassing 55 genera belonged to the phylum Pseudomonadota, 58 encompassing 16 genera to the phylum Actinomycetota, and seven from one genus to the phylum Deinococcota. We exami­ned the metabolic potential of these species to predict autotrophic conditions, with a focus on electron acceptors (oxygen availability or anaerobic respiration using sulfur species and nitrate) and donors (hydrogen and sulfur species; Table S5).

To identify species that utilized hydrogen as an electron donor, we exami­ned their genomes for the presence of genes encoding uptake hydrogenases. Based on the annotation results obtained using the METABOLIC software tool, hydrogenases annotated as Groups A2 and A3 ([FeFe]-hydrogenases) and Groups 1, 2a, 2b, 2c, 2d, 2e, 4h, and 4i ([NiFe]-hydrogenases) were selected as uptake hydrogenases (Søndergaard et al., 2016; Zhou et al., 2022). [NiFe]-hydrogenase genes were detected in 120 species, whereas [FeFe]-hydrogenase genes were only detected in one species. Nearly half of the identified species possessed the [NiFe]-hydrogenase gene, which indicates their potential utilization of hydrogen as an electron donor (Table 2 and S5). Among the 54 form IE-harboring strains, 33 contained the [NiFe]-hydrogenase gene, whereas only four contained genes involved in using sulfur species as an electron donor. Regarding form IA, previous studies showed a predominance of bacteria utilizing sulfur species as an electron donor (described above, also shown in Fig. 1b). However, among the strains belonging to genera in which CO₂ fixation has not yet been demonstrated, the number of strains potentially utilizing sulfur species versus hydrogen as an electron donor was similar—nine species harbored sulfur species oxidation genes and seven harbored hydrogen oxidation genes (Table S5). Most of the analyzed strains belonging to the phylum Pseudomonadota were found to possess key genes encoding both microaerophilic-type terminal oxidases, which have high affinity for oxygen (heme–copper oxidases [HCOs] family C types [ccoON] and cytochrome bd-type oxidases [cydAB]), and aerobic-type terminal oxidases, which have low affinity for oxygen (HCOs-A type terminal oxidase [coxAB or cyoABCD]) (Morris and Schmidt, 2013). In contrast, species from other phyla, particularly Bacillota and Deinococcota, generally possessed either one of these terminal oxidase types, but not both (Table 2).

Table 2.

Summary of the strain number corresponding to the predicted metabolism based on genome information. The species used in this anal­ysis were derived from genera lacking evidence of autotrophic growth. Light gray highlights indicate the total number for each phylum. Details of the genes used for metabolic predictions are given in Table S1. The full strain names and corresponding data are provided in Table S5.

Taxonomy	Total number of strains analyzed	RubisCO forms							Electron donors			Electron acceptors				Photosynthesis (pufL/pufM)
		IA	IB	IC	IE	I_thermus	II		H2 oxidation	Sulfur species oxidation		Microaerophilic (HCOs-B,C)	Aerobic (HCOs-A)	Sulfur species reduction	Nitrate reduction
total counts of strains	144	14	1	58	54	7	10		63	46		123	131	32	77	12
Actinomycetota (Phylum)	58	1		1	53		3		33	3		51	49	2	34
Acidimicrobiales	1			1									1	1
Actinomycetales	1						1					1	1	1	1
Micromonosporales	1	1										1			1
Mycobacteriales	7				7				7	1		7			4
Propionibacteriales	7				5		2					7	6		1
Pseudonocardiales	7				7				3			7	7		4
Solirubrobacterales	2				2				2			2	2		1
Streptosporangiales	32				32				21	2		26	32		22
Bacillota (Phylum)	1				1					1		1		1
Alicyclobacillales	1				1					1		1		1
Deinococcota (Phylum)	7					7			1	4			7	7	4
Thermales	7					7			1	4			7	7	4
Pseudomonadota (Phylum)	78	13	1	57			7		29	38		71	75	22	39	12
Acetobacterales	3			3						1		3	3		1	1
Burkholderiales	20			18			2		4	6		17	18	1	8	1
Hyphomicrobiales	13			13					2	3		13	13	1	3	5
Lysobacterales	2			2					1	1		2	2
Methylococcales	1	1								1		1	1
Nevskiales	1			1					1			1	1	1
Nitrosomonadales	2			2					2	1		2	2	2	2
Rhodobacterales	16	5		9			3		8	15		14	15	15	11	2
Rhodocyclales	3	1		2					2	2		3	3		3
Rhodospirillales	6	2		3			1		5	6		6	6		5	2
Sphingomonadales	1			1								1	1			1
Steroidobacterales	3			3					1			1	3		2
Thiotrichales	2	1					1		2	2		2	2	2
Vibrionales	5	3	1						1			5	5		4

In the phylum Pseudomonadota, the highest number of species was found in the genus Paraburkholderia (Burkholderiales) at nine species, followed by the genus Jiella (Hyphomicrobiales) at four species (Table S5). The remaining genera contained three or fewer species. In this phylum, hydrogen and sulfur species appeared to mainly be used as electron donors and oxygen as an electron acceptor under chemolithoautotrophic conditions. A diverse range of species have the potential to fix CO₂ through the CBB cycle, including 11 species that belong to monotypic genera under the classes Alphaproteobacteria and Gammaproteobacteria. Species from the class Alphaproteobacteria were Tistlia (form IA), Zhengella (form IC), Segnochrobactrum (form IC), Dankookia (form IC), Profundibacter (form II), Thalassobius (form II), and Insolitispirillum (form II). Species from the class Gammaproteobacteria were Methylomagnum (form IA), Cocleimonas (form IA), Metallibacterium (form IC), and Rudaea (form IC) (Table S5).

In the phylum Actinomycetota, all species possessed IE RbcL, except those in the genera Flaviflexus (Actinomycetales), Amorphoplanes (Micromonosporales), Propionicicella, and Propionicimonas (Propionibacteriales), which possessed IA or II RbcL. In this phylum, the genus containing the largest number of potential autotrophic species was Actinomadura (Streptosporangiales) with 24 species, followed by Mycobacterium (Mycobacteriales) with seven and Actinomycetospora (Pseudonocardiales) and Thermomonospora (Streptosporangiales) with five each. All Actinomadura species possessed the gene for low-affinity terminal oxidase (HCOs-A [coxA/coxB]), and 18 species additionally carried the genes encoding cytochrome bd-type oxidases (cydA/cydB), which are high-affinity terminal oxidases, suggesting the ability to grow under microaerophilic to aerobic conditions. The sulfide oxidation gene (sqr) was found in two species that possessed cytochrome bd-type oxidases (cydA/cydB). These species may grow autotrophically with sulfur species as an electron donor under microaerophilic to aerophilic conditions. Of the 15 species that carried the [NiFe]-hydrogenase gene, 12 also harbored nitrate reduction genes (narG/narH), suggesting CO₂ fixation via hydrogen oxidization using hydrogen as an electron donor and oxygen as an electron acceptor under aerobic conditions or via nitrate reduction using hydrogen as an electron donor and nitrate as an electron acceptor. The same metabolic pathways are considered to be involved in Actinomycetospora and Thermomonospora species because all the species in these two genera possessed genes for high-affinity terminal oxidases and almost half possessed the genes for [NiFe]-hydrogenase and nitrate reduction (narG/narH).

The genus Thermus consists of thermophilic species that have been isolated worldwide in geothermal environments (≥55°C) with a pH range of 5.0 to 10.5. These species have demonstrated heterotrophic growth, but some have been reported to harbor complete sets of CBB cycle genes (Da Costa et al., 2006; Müller et al., 2016). All RbcL sequences from Thermus were grouped in the same cluster, as shown in a previous study (Müller et al., 2016; Schulz et al., 2022), and the genome anal­ysis revealed that species in this genus carried the gene for low-affinity terminal oxidases (coxA/coxB). Some species of the genus had genes for complete thiosulfate oxidation (soxABXYZ), whereas others had an incomplete set of genes lacking only soxB. Chemolithoautotrophic growth was previously reported in Thermus sediminis under conditions involving the oxidation of thiosulfate to sulfate; however, this characteristic was only described in the abstract and, thus, limited information was provided (Zhou et al., 2018). Our genome anal­ysis also suggests that if CO₂ fixation occurs, it is expected to take place under thiosulfate oxidation conditions.

Regarding the autotrophic potential of phototrophic bacteria, since the presence of CBB cycle genes does not always correspond to active CO₂ fixation in phototrophs (Shimada and Takaichi, 2024), we also exami­ned photosynthesis-related genes to obtain for further insights into genera that have not been demonstrated to exhibit autotrophic growth at the genus level. Through a homologous BLAST search, we identified the pufL and pufM genes, which encode a type II reaction center involved in anoxygenic photosynthesis, in 12 species belonging to the phylum Pseudomonadota. These species belonged to eight genera, seven from the class Alphaproteobacteria (Jiella, Aliigemmobacter, Phaeovulum, Dankookia, Skermanella, Polymorphobacter, and Rhodoligotrophos) and one from the class Betaproteobacteria (Piscinibacter). All these species carried form IC RubisCO, except Phaeovulum vinaykumarii, which carried genes for both form IA and II RubisCOs. However, P. vinaykumarii has been reported to lack the ability to grow autotrophically under photosynthetic conditions (Srinivas et al., 2007). Some autotrophic growth has‍ ‍been reported in Rhodobacter and Rhodovulum (Paracoccaceae, Rhodobacterales) species.

In the genus Jiella (Alphaproteobacteria), four species were identified as potential phototrophs, representing the highest number of species found among the genera analyzed. Genes involved in hydrogen or sulfur species (such as sulfide and thiosulfate) were not detected in the genomes of these species (Table S5), requiring further anal­yses to predict the metabolism of autotrophic growth. Alternatively, it is possible that Jiella utilizes the CBB cycle under heterotrophic growth conditions and does not involve autotrophic growth, as reported in AAPB (Yurkov and Beatty, 1998). In‍ ‍Alphaproteobacteria, Hyphomicrobiales containing Jiella‍ ‍(Aurantimonadaceae) and Rhodoligotrophos (Parvibaculaceae) accommodate multiple photoautotrophic purple non-sulfur bacteria (e.g., Rhodobium and Rhodopseudomonas) (Table S4) and AAPB (e.g., Methylorubrum) (Shimada and Takaichi, 2024). In Acetobacterales, which includes Dankookia, lacking the genes involved in electron donor utilization by Jiella (Table S5), studies have reported AAPB (e.g., Acidisphaera and Craurococcus) and purple non-sulfur bacteria (e.g., Rhodopila and Rhodovastum) (Shimada and Takaichi, 2024).

Conclusion

The present study identified previously unrecognized autotrophic potential by predicting the CBB cycle in diverse accessible species in the JCM collection through an integrated anal­ysis of genome data and literature surveys. We clarified which CO₂ fixation capabilities have been experimentally demonstrated and what remains unclear regarding the CO₂ fixation potential of the isolated species. These results provide a foundation for future research focused on confirming the autotrophic capability of these species through further experimental anal­yses.

Citation

Nishihara, A., Kato, S., and Ohkuma, M. (2026) Exploring the Diversity and Metabolic Potential of CO₂ fixation Mediated by RubisCO in Prokaryotes in the Japan Collection of Microorganisms. Microbes Environ 41: ME25035.

https://doi.org/10.1264/jsme2.ME25035

Acknowledgements

This work was supported by the Japan Science and Technology Agency (JST) GteX program Grant Number JPMJGX23B0 (M.O.) and was partially supported by the Japan Society for the Promotion of Science Grant-in-Aid for Early-Career Scientists 24K17163 (A.N.).

Conflicts of interest

The authors declare that there are no conflicts of interest.

References

•  Aono,  R.,  Sato,  T.,  Yano,  A.,  Yoshida,  S.,  Nishitani,  Y.,  Miki,  K., et al. (2012) Enzymatic characterization of AMP phosphorylase and ribose-1,5-bisphosphate isomerase functioning in an archaeal AMP metabolic pathway. J Bacteriol 194: 6847–6855.
•  Aono,  R.,  Sato,  T.,  Imanaka,  T., and  Atomi,  H. (2015) A pentose bisphosphate pathway for nucleoside degradation in Archaea. Nat Chem Biol 11: 355–360.
•  Arora,  Y.,  Sharma,  S., and  Sharma,  V. (2023) Microalgae in bioplastic production: A comprehensive review. Arabian J Sci Eng 48: 7225–7241.
•  Badger,  M.R., and  Bek,  E.J. (2008) Multiple Rubisco forms in proteobacteria: Their functional significance in relation to CO₂ acquisition by the CBB cycle. J Exp Bot 59: 1525–1541.
•  Banda,  D.M.,  Pereira,  J.H.,  Liu,  A.K.,  Orr,  D.J.,  Hammel,  M.,  He,  C., et al. (2020) Novel bacterial clade reveals origin of form I Rubisco. Nat Plants 6: 1158–1166.
•  Benson,  A.A.,  Bassham,  J.A., and  Calvin,  M. (1948) The path of carbon in photosynthesis. Science 107: 476–480.
•  Berg,  I.A.,  Kockelkorn,  D.,  Ramos-Vera,  W.H.,  Say,  R.F.,  Zarzycki,  J.,  Hügler,  M., et al. (2007) A 3-hydroxypropionate/4-hydroxybutyrate autotrophic carbon fixation cycle in Archaea. Proc Natl Acad Sci U S A 104: 15370–15375.
•  Borrel,  G.,  Adam,  P.S., and  Gribaldo,  S. (2016) Methanogenesis and the Wood–Ljungdahl pathway: An ancient, versatile, and fragile association. Genome Biol Evol 8: 1706–1711.
• Boutet, E., Lieberherr, D., Tognolli, M., Schneider, M., and Bairoch, A. (2007) Plant bioinformatics: Methods and protocols. In Plant Bioinformatics: Methods and Protocols. Edwards, D. (ed). Totowa, NJ: Humana Press, pp. 89–112.
•  Capella-Gutiérrez,  S.,  Silla-Martínez,  J.M., and  Gabaldón,  T. (2009) trimAl: A tool for automated alignment trimming in large-scale phylogenetic anal­yses. Bioinformatics 25: 1972–1973.
•  Clark,  D.A., and  Norris,  P.R. (1996) Acidimicrobium ferrooxidans gen. nov., sp. nov.: Mixed-culture ferrous iron oxidation with Sulfobacillus species. Microbiology (London, U K) 142: 785–790.
• Da Costa, M.S., Rainey, F.A., and Nobre, M.F. (2006) The Prokaryotes: Volume 7: Proteobacteria: Delta, Epsilon Subclass. In The Prokaryotes. Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K-H., Stackebrandt, E. (eds). New York, NY: Springer, pp. 797–812.
•  Evans,  M.C.W.,  Buchanan,  B.B., and  Arnon,  D.I. (1966) A new ferredoxin-dependent carbon reduction cycle in a photosynthetic bacterium. Proc Natl Acad Sci U S A 55: 928–934.
•  Ferry,  J.G.,  Smith,  P.H., and  Wolfe,  R.S. (1974) Methanospirillum, a new genus of methanogenic bacteria, and characterization of Methanospirillum hungatii sp. nov. Int J Syst Evol Microbiol 24: 465–469.
•  Field,  C.B.,  Behrenfeld,  M.J.,  Randerson,  J.T., and  Falkowski,  P. (1998) Primary production of the biosphere: Integrating terrestrial and oceanic components. Science 281: 237–240.
•  Frolov,  E.N.,  Kublanov,  I.V.,  Toshchakov,  S.V.,  Lunev,  E.A.,  Pimenov,  N.V.,  Bonch-Osmolovskaya,  E.A., et al. (2019) Form III RubisCO-mediated transaldolase variant of the Calvin cycle in a chemolithoautotrophic bacterium. Proc Natl Acad Sci U S A 116: 18638–18646.
•  Fu,  L.,  Niu,  B.,  Zhu,  Z.,  Wu,  S., and  Li,  W. (2012) CD-HIT: Accelerated for clustering the next-generation sequencing data. Bioinformatics 28: 3150–3152.
•  Garritano,  A.N.,  Song,  W., and  Thomas,  T. (2022) Carbon fixation pathways across the bacterial and archaeal tree of life. PNAS Nexus 1: pgac226.
•  Holo,  H., and  Øvreås,  O.H.K.H. (1987) 3-Hydroxypropionate, a possible intermediate in autotrophic CO₂ fixation in Chloroflexus aurantiacus. Arch Microbiol 147: 291–295.
•  Huber,  H.,  Gallenberger,  M.,  Jahn,  U.,  Eylert,  E.,  Berg,  I.A.,  Kockelkorn,  D., et al. (2008) A dicarboxylate/4-hydroxybutyrate cycle in Crenarchaeota. Proc Natl Acad Sci U S A 105: 7851–7856.
• Imhoff, J.F. (2009) Phototrophic purple bacteria. In Encyclopedia of Life Sciences (ELS). Chichester: John Wiley & Sons.
•  Jones,  R.M., and  Johnson,  D.B. (2015) Acidithrix ferrooxidans gen. nov., sp. nov.: A filamentous and obligately heterotrophic, acidophilic Actinobacterium that catalyzes dissimilatory iron redox cycling. Res Microbiol 166: 111–120.
•  Kalyaanamoorthy,  S.,  Minh,  B.Q.,  Wong,  T.K.F.,  von Haeseler,  A., and  Jermiin,  L.S. (2017) ModelFinder: Fast model selection for accurate phylogenetic estimates. Nat Methods 14: 587–589.
•  Kanehisa,  M.,  Sato,  Y., and  Morishima,  K. (2016). BlastKOALA and GhostKOALA: KEGG tools for functional characterization of genome and metagenome sequences. J Mol Biol 428: 726–731.
•  Karp,  P.D.,  Billington,  R.,  Caspi,  R.,  Fulcher,  C.A., and  Latendresse,  M., et al. (2019). The BioCyc collection of microbial genomes and metabolic pathways. Briefing Bioinf 20: 1085–1093.
•  Katoh,  K.,  Kuma,  K.-I.,  Toh,  H., and  Miyata,  T. (2005) MAFFT version 5: Improvement in accuracy of multiple sequence alignment. Nucleic Acids Res 33: 511–518.
•  Kishimoto,  N.,  Fukaya,  F.,  Inagaki,  K.,  Sugio,  T.,  Tanaka,  H., and  Tano,  T. (1995) Distribution of bacteriochlo­rophyll a among aerobic and acidophilic bacteria and light-enhanced CO₂-incorporation in Acidiphilium rubrum. FEMS Microbiol Ecol 16: 291–296.
•  Kono,  T.,  Mehrotra,  S.,  Endo,  C.,  Kizu,  N.,  Matsuda,  M.,  Kimura,  H., et al. (2017) A RuBisCO-mediated carbon metabolic pathway in methanogenic archaea. Nat Commun 8: 14007.
•  Lartillot,  N., and  Philippe,  H. (2006) Computing Bayes factors using thermodynamic integration. Syst Biol 55: 195–207.
•  Liu,  A.K.,  Kaeser,  B.,  Chen,  L.,  West-Roberts,  J.,  Taylor-Kearney,  L.J.,  Lavy,  A., et al. (2023) Deep-branching evolutionary intermediates reveal structural origins of form I Rubisco. Curr Biol 33: 5316–5325.
•  Mori,  K.,  Kim,  H.,  Kakegawa,  T., and  Hanada,  S. (2003) A novel lineage of sulfate-reducing microorganisms: Thermodesulfobiaceae fam. nov., Thermodesulfobium narugense gen. nov., sp. nov. Extremophiles 7: 283–290.
•  Morris,  R.L., and  Schmidt,  T.M. (2013) Shallow breathing: Bacterial life at low O₂. Nat Rev Microbiol 11: 205–212.
•  Müller,  W.J.,  Tlalajoe,  N.,  Cason,  E.D.,  Litthauer,  D.,  Reva,  O.,  Brzuszkiewicz,  E., and  van Heerden,  E. (2016) Whole-genome comparison of Thermus sp. NMX2.A1 reveals principal carbon metabolism differences with closest relation Thermus scotoductus SA-01. G3:Genes, Genomes, Genet 6: 2791–2797.
•  Nangle,  S.N.,  Ziesack,  M.,  Buckley,  S.,  Trivedi,  D.,  Loh,  D.M.,  Nocera,  D.G., and  Silver,  P.A. (2020) Valorization of CO₂ through lithoautotrophic production of sustainable chemicals in Cupriavidus necator. Metab Eng 62: 207–220.
•  Nguyen,  L.-T.,  Schmidt,  H.A.,  von Haeseler,  A., and  Minh,  B.Q. (2014) IQ-TREE: A fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol 32: 268–274.
•  Nishihara,  A.,  Tsukatani,  Y.,  Azai,  C., and  Nobu,  M.K. (2024) Illuminating the coevolution of photosynthesis and Bacteria. Proc Natl Acad Sci U S A 121: e2322120121.
• Pachauri, R.K., and Mayer, L. (2015) Climate Change 2014: Synthesis Report. Geneva, Switzerland: Intergovernmental Panel on Climate Change.
•  Prywes,  N.,  Phillips,  N.R.,  Tuck,  O.T.,  Valentin-Alvarado,  L.E., and  Savage,  D.F. (2023) Rubisco function, evolution, and engineering. Annu Rev Biochem 92: 385–410.
•  Sánchez-Andrea,  I.,  Guedes,  I.A.,  Horn,  M.A., and  Plugge,  C.M. (2014) The reductive glycine pathway allows autotrophic growth in Desulfovibrio desulfuricans. mBio 5: e00944-14.
•  Sato,  T.,  Atomi,  H., and  Imanaka,  T. (2007) Archaeal type III RuBisCOs function in a pathway for AMP metabolism. Science 315: 1003–1006.
•  Schulz,  L.,  Guo,  Z.,  Zarzycki,  J.,  Steinchen,  W.,  Schuller,  J.M.,  Heimerl,  T., et al. (2022) Evolution of increased complexity and specificity at the dawn of form I Rubiscos. Science 378: 155–160.
•  Searcy,  D.G., and  Lee,  S.H. (1998) A model for the origin of biological carbon fixation. BioSystems 45: 51–60.
•  Seemann,  T. (2014) Prokka: Rapid prokaryotic genome annotation. Bioinformatics 30: 2068–2069.
•  Shiba,  T. (1984) Utilization of light energy by the strictly aerobic bacterium Erythrobacter sp. OCH 114. J Gen Appl Microbiol 30: 239–244.
•  Shiba,  T., and  Harashima,  K. (1986) Aerobic photosynthetic bacteria. Microbiol Sci 3: 376–378.
• Shimada, K., and Takaichi, S. (2024) Anoxygenic Phototrophic Bacteria. Amsterdam: Elsevier.
•  Shimizu,  S.,  Ueno,  A.,  Tamamura,  S.,  Naganuma,  T., and  Kaneko,  K. (2013) Methanoculleus horonobensis sp. nov., a methanogenic archaeon isolated from a deep diatomaceous shale formation. Int J Syst Evol Microbiol 63: 4320–4323.
•  Slobodkin,  A.I.,  Slobodkina,  G.B.,  Panteleeva,  A.N.,  Chernyh,  N.A.,  Novikov,  A.A., and  Bonch-Osmolovskaya,  E.A. (2016) Dissulfurimicrobium hydrothermale gen. nov., sp. nov., a thermophilic, autotrophic, sulfur-disproportionating deltaproteobacterium isolated from a hydrothermal pond. Int J Syst Evol Microbiol 66: 1022–1026.
•  Søndergaard,  D.,  Pedersen,  C.N.S., and  Greening,  C. (2016) HydDB: A web tool for hydrogenase classification and analysis. Sci Rep 6: 34212.
•  Srinivas,  T.N.,  Kumar,  P.,  Sasikala,  C.,  Ramana,  C.V., and  Imhoff,  J.F. (2007) Rhodobacter vinaykumarii sp. nov., a marine phototrophic alphaproteobacterium from tidal waters, and emended description of the genus Rhodobacter. Int J Syst Evol Microbiol 57: 1984–1987.
•  Tabita,  F.R.,  Satagopan,  S.,  Hanson,  T.E.,  Kreel,  N.E., and  Scott,  S.S. (2008) Distinct form I, II, III, and IV Rubisco proteins from the three kingdoms of life provide clues about Rubisco evolution and structure/function relationships. J Exp Bot 59: 1515–1524.
• West-Roberts, J.A., Matheus-Carnevali, P.B., Schoelmerich, M.C., Al-Shayeb, B., Thomas, A.D., Sharrar, A., et al. (2021) The Chloroflexi supergroup is metabolically diverse and representatives have novel genes for non-photosynthesis-based CO₂ fixation. bioRxiv https://doi.org/10.1101/2021.08.23.457424.
•  Wood,  H.G., and  Werkman,  C.H. (1936) The utilization of CO₂ by Clostridium acidi-urici and C. acidi-propionici. Biochem J 30: 48–53.
•  Yurkov,  V.V., and  Beatty,  J.T. (1998) Aerobic anoxygenic phototrophic bacteria. Microbiol Mol Biol Rev 62: 695–724.
•  Yvenou,  S.,  Allioux,  M.,  Slobodkin,  A.,  Slobodkina,  G.,  Jebbar,  M., and  Alain,  K. (2022) Genetic potential of Dissulfurimicrobium hydrothermale, an obligate sulfur-disproportionating thermophilic microorganism. Microorganisms 10: 60.
•  Zhou,  E.M.,  Xian,  W.D.,  Mefferd,  C.C.,  Thomas,  S.C.,  Adegboruwa,  A.L.,  Williams,  N., et al. (2018) Thermus sediminis sp. nov., a thiosulfate-oxidizing and arsenate-reducing organism isolated from Little Hot Creek in the Long Valley Caldera, California. Extremophiles 22: 983–991.
•  Zhou,  Z.,  Tran,  P.Q.,  Breister,  A.M.,  Liu,  Y.,  Kieft,  K.,  Cowley,  E.S., et al. (2022) METABOLIC: High-throughput profiling of microbial genomes for functional traits, metabolism, biogeochemistry, and community-scale functional networks. Microbiome 10: 33.