PMC

A. The three clusters of genes encoding flagellum-related and chemotaxis-related proteins located on the megaplasmid. B. The flagellar structures and related chemotaxis apparatus encoded in the T. roseum genome.

The tree was built from concatenated alignments of 31 housekeeping genes using the PHYML program. The bootstrap values are based on 100 replications. Organisms with photosynthetic capability are emphasized in green. Members from the phylum Chloroflexi are highlighted by the shaded box. Note – the organism Chlorobaculum tepidum used to be known as Chlorobium tepidum.

Hexamer frequencies were extracted from 20,000 random 4 kb fragments from the whole genome and analyzed using CompostBin, a PCA-based method. The first two principal components of the data are plotted. Red indicates chromosome data; yellow indicates megaplasmid data; orange indicates overlaps between the two; circles indicate the three outlier regions.

A. Two pathways for carbon dioxide fixation and an electronic transport chain for carbon monoxide/hydrogen oxidation in T. roseum. B. Phylogenetic tree for the catalytic domain of Types I-V ribulose bisphosphate carboxylase large subunit RbcL). The tree is based on alignments generated using Pfam domain PF00016. C. Phylogenetic tree for the molybdopterin-binding domain of the aerobic CO dehydrogenase CoxL. The tree is based on alignments generated using Pfam domain PF02738.

The circles display the following features, starting with the outermost circle: (1) forward strand genes; (2) reverse strand genes; (3) chi square deviation of local nucleotide composition from the genome average; (4) GC skew (blue bars represent positive values, red bars represent negative values); (5) tRNAs (green lines); (6) rRNAs (blue lines); (7) small RNAs (red lines). Gene color indicates the assigned role category. A gene can be included in the gene count for multiple role categories.

A. CO concentration in headspace plotted against incubation time. For the 20% CO experiment: (a) O₂ concentration (not shown) had decreased to below detection level at 400 hours; and (b) the onset of CO consumption was delayed by roughly 75 hours as compared with the 5% CO experiment. B. CO₂ concentration in headspace plotted against incubation time. For the 20% CO experiment, O₂ concentration (not shown) had decreased to below detection level at 400 hrs, whereas in the 0% CO control, it fell from an initial level of 20% down to about 0.5%. Initial CO₂ production appeared to be highest in the 5% CO experiment (which had an earlier onset of CO oxidation), followed by the 20% CO experiment. Production of CO₂ lagged in the T. roseum culture with 0% CO and plateaued at a lower level.

A. Reverse phase HPLC analysis of T. roseum pigments revealed two types of carotenoid compounds: extremely hydrophobic pigments from crude extracts (top panel, peaks 1 & 2) and a polar glycoside released by saponification (bottom panel, peak 3). The glycoside is the major form of carotenoid in T. roseum. Its HPLC retention time was similar to that of the authentic oscillaxanthin (oscillol-2,2′-difucoside) from the cyanobacterium Gloeobacter violaceus PCC7421 (see for the molecular structure). B. Comparison of the in-line spectra of compounds from peak 1, 2 & 3 () with that of the oscillaxanthin from Gloeobacter violaceus PCC 7421 indicates that the major T. roseum carotenoid core is oscillol diglycoside. The structure of oscillaxanthin is shown to the right. The sugar moieties in oscillaxanthin are fucose; the identity and exact positions of the glycosyl moieties in the T. roseum oscillol diglycoside are yet to be determined. C. Carotenoid modification pathway in T. roseum. Step 1: The introduction of two hydroxyl groups at the ends of lycopene to form (2S, 2′S)-oscillol. The absorbance spectra of T. roseum pigments () indicate that 1,1′ and 2,2′ positions have all been hydroxylated. We found a 1,1′ hydroxylase in the genome; the gene(s) encoding the 2,2′ hydroxylation function are yet to be identified. Step 2: A glycosyl transferase (CruC) adds a sugar moiety (R) to both ends of (2S,2′S)-oscillol to produce oscillol diglycoside, the dominant carotenoid (). The nature of the sugar moieties are unknown. Although the figure shows a 2,2′-oscillol diglycoside, 1,1′-oscillol diglycoside is another possibility at this step. Step 3: The acyltransferase domain of the 1,1′ hydroxylase adds acyl chains (A) to the sugar moieties (R), thus producing the extremely hydrophobic pigments seen in the crude extracts ().