Membrane changes during syntrophic interactions of an archaeal-bacterial coculture

Abstract

Syntrophic interactions between bacteria and archaea are vital for anaerobic processes, relying on close cell-to-cell contact for efficient metabolite and electron transfer. Membrane-associated proteins and lipids likely play key roles in stabilizing these contacts, though little is known about membrane changes during syntrophy. These interactions are also central to theories of eukaryogenesis, where a symbiosis between an archaeal host - likely an Asgard archaeon - and a bacterial partner may have arisen from prior syntrophic interactions. Model systems of syntrophic microbes provide valuable insights into how such intimate associations could have led to the emergence of eukaryotic life. Here, we used syntrophic cocultures of the sulfate-reducing bacterium Desulfovibrio vulgaris and the methanogenic archaeon Methanococcus maripaludis to investigate membrane changes during a syntrophic interaction involving cell-to-cell contact. Evolved cocultures after several generations under syntrophic conditions were analyzed by proteomics and transcriptomics to identify differentially expressed proteins connected to cell-to-cell interactions, as well as by lipid analyses to determine changes in the cell membrane of both syntrophic partners. These data suggest a higher impact on the archaeon M. maripaludis, affecting transmembrane, signaling, and lipid biosynthesis proteins. To investigate the impact of evolutionary adaptation, both partners were re-isolated from a non-evolved ancestral coculture (coculture after mixing species), as well as from evolved (several generations) cocultures. While lipid profiles had changed in the coculture due to evolutionary adaptation, isolates were found to revert their lipid composition to the wildtype profile when growing independent again. This in-depth analysis of a model syntrophic coculture provides clues on how interdomain cell-to-cell interactions might have led to membrane changes during early eukaryogenesis.

Keywords: Desulfovibrio vulgaris; Methanococcus maripaludis; Archaea; Eukaryogenesis; Intact polar lipids; Membrane changes; Membrane lipids; Syntrophy.

Fig. 1

Scheme of experimental set up and visualization of the Methanococcus maripaludis (M. maripaludis; archaeon) and Desulfovibrio vulgaris (D. vulgaris; bacterium) cocultures. A Monocultures of both wild type (WT) species and cocultures obtained from different generations during adaptation to syntrophic growth have been analyzed for their lipids, proteome and transcriptome (green). From all three cocultures, the differentially adapted strains of both species were isolated into monocultures and checked for occurred mutations (orange). Monocultures of WT and isolated strains were used to start initial cocultures using different combinations and analyzed for their lipids after grown once to stationary phase (blue). B Microscopy image of M. maripaludis (cocci, blue) and D. vulgaris (vibrio shaped) in aggregates of Anc-Co, 300-Co and 1000-Co cocultures during early stationary phase. WT: wild type monoculture, Mmp: M. maripaludis, Dvu: D. vulgaris, Anc-Co: non-evolved/ancestral coculture, 300-Co: evolved 300 generations coculture, 1000-Co: evolved 1,000 generations coculture

Fig. 2

Analysis of the proteomics data for M. maripaludis (A & C) and D. vulgaris B & D). A & B Relative numbers of proteins with higher (up) and lower (down) abundances of all protein coding genes between different two-sample t-test combinations of cultures. The numbers indicated in the figures show the absolute number of proteins affected (cutoff FDR 0.05, S0 0.1), (C & D) Venn diagrams showing relations between proteome of all three cocultures in comparison with WT monoculture. Numbers indicated represent absolute numbers of detected proteins whose abundance varies significantly (n = 5 for each culture group). WT: wild type monoculture, Anc-Co: non-evolved/ancestral coculture, 300-Co: evolved 300 generations coculture, 1000-Co: evolved 1,000 generations coculture

Fig. 3

Functional characterization of proteins with transmembrane domains (TM) with increased abundance. Dot plots show the absolute number of detected and identified proteins (n= 5 for each culture group) and the identified gene ontology of biological processes (GOBP) for M. maripaludis (A) and D. vulgaris (B) found to be more abundant in the different t-test comparisons (FDR 0.05, S0 0.1). GOBP annotation was performed using UniProtKB database. Unclassified proteins: no GO annotation available, WT: wild type monoculture, Anc-Co: non-evolved/ancestral coculture, 300-Co: evolved 300 generations coculture, 1000-Co: evolved 1,000 generations coculture, only Anc-Co: only found in Anc-Co coculture not 300-Co and 1,000-Co, evolved Co: only found in evolved 300-Co and 1,000-Co cocultures not in Anc-Co, all Co: found in all three cocultures

Fig. 4

Changes in the membrane lipid biosynthetic proteins and selected membrane-associated proteins in the archaeon M. maripaludis. A Scheme of archaeal lipid biosynthesis involving the mevalonate pathway for isoprenoid synthesis followed by prenylations to form ether bonds with glycerol-1-phosphate (G1P) and subsequent modifications to add polar head groups and saturation of isoprenoid side chains [38]. Shown are corresponding genes found in M. maripaludis and heatmaps of protein and transcript changes of involved proteins/genes. B Heatmap of additional proteins and transcripts putatively involved in lipid biosynthesis based on function and gene ontology. Protein heatmaps show Student’s t-test difference; transcript heatmaps show log2FC. White: no significant change (proteomics: FDR 0.05, S0 0.1 & transcriptomics Padj ≤ 0.05), grey: not detected, ???: gene not known, WT: wild type monoculture, HMGS: 3-hydroxy-3-methylglutary-CoA synthase; HMG-CoA: 3-hydroxy-3-methylgutaryl-CoA; HMGR: 3-hydroxy-3-methylglutary-CoA reductase; MVA: mevalonate; MVK: mevalonate kinase; MVA-5-P: mevalonate 5-phosphate; IP: isopentenyl phosphate; IPK: isopentenylphosphate kinase; IPP: isopentenyl pyrophosphate; IDI: isopentenyl-diphosphate delta-isomerase; DMAPP: dimethylallyl pyrophosphate; GGPP: geranylgeranyl diphosphate; GGGP: 3-O-geranylgeranyl-sn-glyceryl-1-phosphate; DGGGP: 2,3-digeranylgeranyl-sn-glycerol-1-phosphate; G1P: sn-glycerol-1-phosphate; DHAP: dihydroxyacetone-phosphate; CDP: cytidine diphosphate diglyceride-glyceride precursor. Arrow on the X-axis depicts the direction of the evolutionary process in the cocultures

Fig. 5

Changes in the membrane lipid biosynthetic proteins and selected membrane-associated proteins in the bacterium D. vulgaris. A Scheme of the bacterial membrane lipid biosynthesis including fatty acid synthesis (FAS) and subsequent modifications to add the polar head groups. Shown are corresponding genes found in D. vulgaris and heatmaps of protein and transcript changes of involved proteins/genes. B Heatmap of additional proteins and transcripts putatively involved in lipid, lipopolysaccharide and exopolysaccharide biosynthesis based on KEGG pathway and gene ontology. Protein heatmaps show Student’s t-test difference, transcript heatmaps show log2FC, white: no significant change (proteomics: FDR 0.05, S0 0.1 & transcriptomics Padj ≤ 0.05), grey: not detected. DHAP: dihydroxyacetone phosphate, ACP: acyl carrier protein, WT: wild type monoculture. Arrow on the X-axis depicts the direction of the evolutionary process in the cocultures

Fig. 6

Membrane lipid composition in M. maripaludis (A, B) and D. vulgaris (C, D) growing in wild type monocultures (WT) and cocultures (Anc-Co, 300-Co, 1000-Co). A-C Composition of intact polar lipids: core lipid archaeol (AR), polar head groups, glycolipid-based: monohexose (MH) & dihexose (DH), phosphate-based: phosphatidic acid (PA), phosphatidylserine (PS), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), cardiolipin (CL) and aminolipid polar head group: ornithine lipid (OL). B Abundance of hydroxylation in core archaeol: no hydroxylation (AR), monohydroxylated (AR-OH) and dihydroxylated (AR-diOH). D Relative abundance of unsaturations in the fatty acids of D. vulgaris core lipids. FA saturation: fully saturated (SAT), one unsaturation (UNSAT-1) and two unsaturations (UNSAT-2). Relative numbers in tabular form are provided in Supplementary Material 4

Fig. 7

Lipid analysis of isolates and impact of adaptation level on lipid rearrangement (A) IPL compositions and (B) hydroxylation level of archaeol membrane lipid in M. maripaludis and (C) IPL polar head group composition and (D) fatty acid saturation levels of the membrane lipids of D. vulgaris. Shown are results for non-syntrophic wild type (WT) monocultures and isolates, initial cocultures mixed of isolates with same adaptation level (isolated from same coculture), impact of different combinations on WT cells, Ancestor isolate (Anc) and evolved 1000 generations (1000) isolate. IPLs: Archaeol (AR), monohexose (MH), dihexose (DH), phosphatidic acid (PA), phosphatidylserine (PS), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), cardiolipin (CL) and ornithine lipid (OL). Hydroxylation of AR: No hydroxylation (AR), monohydroxylated (AR-OH) and dihydroxylated (AR-diOH). FA saturation: fully saturated (SAT), one unsaturation (UNSAT-1) and two unsaturations (UNSAT-2). Relative numbers in tabular form are provided in Supplementary Material 4