Metabolism and Occurrence of Methanogenic and Sulfate-Reducing Syntrophic Acetate Oxidizing Communities in Haloalkaline Environments

Anaerobic syntrophic acetate oxidation (SAO) is a thermodynamically unfavorable process involving a syntrophic acetate oxidizing bacterium (SAOB) that forms interspecies electron carriers (IECs). These IECs are consumed by syntrophic partners, typically hydrogenotrophic methanogenic archaea or sulfate reducing bacteria. In this work, the metabolism and occurrence of SAOB at extremely haloalkaline conditions were investigated, using highly enriched methanogenic (M-SAO) and sulfate-reducing (S-SAO) cultures from south-western Siberian hypersaline soda lakes. Activity tests with the M-SAO and S-SAO cultures and thermodynamic calculations indicated that H2 and formate are important IECs in both SAO cultures. Metagenomic analysis of the M-SAO cultures showed that the dominant SAOB was ‘Candidatus Syntrophonatronum acetioxidans,’ and a near-complete draft genome of this SAOB was reconstructed. ‘Ca. S. acetioxidans’ has all genes necessary for operating the Wood–Ljungdahl pathway, which is likely employed for acetate oxidation. It also encodes several genes essential to thrive at haloalkaline conditions; including a Na+-dependent ATP synthase and marker genes for ‘salt-out‘ strategies for osmotic homeostasis at high soda conditions. Membrane lipid analysis of the M-SAO culture showed the presence of unusual bacterial diether membrane lipids which are presumably beneficial at extreme haloalkaline conditions. To determine the importance of SAO in haloalkaline environments, previously obtained 16S rRNA gene sequencing data and metagenomic data of five different hypersaline soda lake sediment samples were investigated, including the soda lakes where the enrichment cultures originated from. The draft genome of ‘Ca. S. acetioxidans’ showed highest identity with two metagenome-assembled genomes (MAGs) of putative SAOBs that belonged to the highly abundant and diverse Syntrophomonadaceae family present in the soda lake sediments. The 16S rRNA gene amplicon datasets of the soda lake sediments showed a high similarity of reads to ‘Ca. S. acetioxidans’ with abundance as high as 1.3% of all reads, whereas aceticlastic methanogens and acetate oxidizing sulfate-reducers were not abundant (≤0.1%) or could not be detected. These combined results indicate that SAO is the primary anaerobic acetate oxidizing pathway at extreme haloalkaline conditions performed by haloalkaliphilic syntrophic consortia.


Stoichiometry of acetate oxidation
In the M-SAO and S-SAO cultures, around 4-5 mM and 1.6 mM acetate was consumed that did not result in stoichiometric methane and sulfide formation, respectively. M. natronophilus needs acetate for growth (Zhilina et al., 2013), whereas Desulfonatronovibrio magnus does not need it but grows better with it (Sorokin et al., 2011). Pure growing cultures of M. natronophilus consumed on average 1.1 mM (±0.5) acetate and pure cultures of Desulfonatronovibrio magnus consumed on average 0.6 mM (±0.4) acetate when growing with H2 as electron donor ( Supplementary Fig 11). Anabolic acetate consumption can therefore explain the gap in acetate stoichiometry, since a total of 8 MAGs were recovered from the enrichment culture that represent 8 organisms that could have consumed part of the acetate for growth (Supplementary (Kevbrin et al., 1998;Sorokin et al., 2008;Zhilina et al., 2013;Sorokin et al., 2015). All bacterial and archaeal MAGs contained genes for acetate activation either via ACK/PTA (MSAO_Bac2, MSAO_Bac3) or via AMP-forming acetyl-CoA synthethase (MSAO_Bac1, MSAO_Bac4 and all archaeal MAGs). Besides 'Ca. S. acetioxidans', only MSAO_Bac3 (related to Desulfonatronospira sp.) contained all genes for operating the Wood-Ljungdahl (WL) pathway, but pure culture representatives of this genus did not use acetate for catabolic purposes (Zhilina et al., 1997;Sorokin et al., 2008;Sorokin et al., 2010) and it therefore probably uses the WL pathway for CO2 fixation. The other archaeal MAGs were a lithotrophic methanogen (MSAO_Arc2) and a methylotrophic methanogen (MSAO_Arc3), both incapable to use acetate for methanogenesis.

Other metabolic properties
The genome of 'Ca. S. acetioxidans' encodes all enzymes of the glycolysis (except for the pyruvate kinase isozymes) which shows that it probably has the ability to degrade sugars or to perform gluconeogenesis The genome also encodes for the non-oxidative branch of the pentose phosphate pathway and can therefore produce glyceraldehyde-3-phosphate from ribulose-5-phosphate and vice versa ( Supplementary Fig 7). 'Ca. S. acetioxidans' does not encode for a complete TCA cycle. Conversion of malate to oxaloacetate proceeds via activity of a malate dehydrogenase, which is not present in the genome. However, the genome does encode for an enzyme that could bypass this conversion by producing pyruvate from malate using a NAD +dependent oxaloacetate-decarboxylating malate dehydrogenase (ME2; k121-4746-cds5). The pyruvate could also come from acetate via activity of pyruvate synthase (k121-561). Pyruvate could also be produced from oxaloacetate via oxaloacetate decarboxylase (k121-5682) ( Supplementary Fig 7). 'Ca. S. acetioxidans' does not encode for the full enzyme of pyruvate carboxylase since it only encodes for subunit B and could therefore not produce oxaloacetate from pyruvate. It therefore probably has to proceed via PEP. The enzyme that normally goes in the direction from PEP to pyruvate, pyruvate kinase is encoded in the genome (k121-951). The other missing part of the TCA cycle is a gene that encodes for the enzyme for conversion of citrate to isocitrate via cis-aonitate; aconitate hydratase (acnA/acnB/ACO). The genome does not encode for an aconitase. The genome does encode for a possible homologue of aconitase, the 3isopropylmalate dehydratase (k121-5682). 3-isopropylmalate dehydratase (alphaisopropylmalate isomerase) was described to be a possible aconitase homologue that also belongs to the aconitase superfamily where all members show a similar overall structure and domain organization. 3-isopropylmalate dehydratase is normally involved in leucine biosynthesis. The gene for 3-isopropylmalate dehydratase was indeed found next to oxaloacetate decarboxylase and leucine biosynthesis genes (k121-5682) and is therefore probably involved in leucine biosynthesis and not in aconitase activity. This however needs to be proven. 'Ca. S. acetioxidans' therefore also probably does not have the potential to fix carbon using the reverse TCA cycle, even with the recently discovered reversibility of citrate synthase (Mall et al., 2018;Nunoura et al., 2018), since the genome also does not encode for a pyruvate carboxylase gene.

Figure S7
Schematic reconstruction of acetate activation, the TCA cycle, the non-oxidative pentose phosphate pathway and glycolysis/gluconeogenesis in the partial genome of 'Ca. Syntrophonatronum acetixodidans'. Red names and red lines indicate absence of these genes or these conversions, respectively.

Figure S8
Partial sequence alignments of sodium or proton-dependent F1F0-type ATP synthase c-subunits with the one of 'Ca.
Syntrophonatronum acetixodidans'. Boxed sequences show sodium dependent F1F0-type ATP synthase c-subunits whereas all others are proton-dependent. Thermoacetogenium phaeum has both amino acids found in sodium-and proton dependent F1F0-type ATP synthase c-subunits.