All routine chemicals were purchased from Sigma Aldrich (Darmstadt, Germany), Carl Roth (Karlsruhe, Germany), or Merck (Darmstadt, Germany). Sulfoquinovose (SQ) and racemic 3-sulfolactate (SL) were purchased from MCAT GmbH (Donaueschingen, Germany). Racemic 2,3-dihydroxypropanesulfonate (DHPS) was synthesized as described previously (Mayer et al., 2010). Isethionate (2-hydroxyethanesulfonic acid sodium salt) was purchased from Merck Schuchardt (Hohenbrunn, Germany).

Plots to illustrate data such as growth curves were generated with Origin, Version 2021b (OriginLab Corporation, Northampton, MA, United States). Plots were aligned and figures were arranged using Inkscape 1.3.¹ Homology predictions between protein sequences or nucleotide sequences were performed using BLAST (basic local alignment search tool) (Altschul et al., 1990) with default options. For the nucleotide BLAST, uncultured bacteria and models were excluded unless otherwise noted.

The mixed culture described here was enriched with 10 mM SQ as the sole source of carbon and energy under strictly anoxic conditions, using anoxic, iron(II)-sulfide-rich (black) sediment from Lake Constance as inoculum, as sampled with a sediment gravity corer in October 2014 from Mainau Bay (Obere Güll; 47°42′00.0"N, 9°11′31.6"E); the sediment surface at approximately 22 m water depth was oxic and the sample was taken from approximately 15 cm deep in the sediment. After four transfers, SQ was confirmed to be degraded and isethionate was detected as intermediate, accompanied by the growth of bacteria visible as turbidity and microscopically. A stable, highly enriched mixed culture was achieved by repeated sub-cultivation from liquid to liquid cultures, as well as by several rounds of anoxic agar dilution series and picking of colonies, each with SQ as the sole source of carbon and energy (Pfennig, 1978); the consortium has since been routinely maintained by transfers at weekly or monthly intervals. The mineral salt medium used was the carbonate-buffered freshwater medium described by Widdel and Pfennig (1981) containing 1.0 g NaCl, 0.4 g MgCl₂ × 6 H₂O, 0.2 g KH₂PO₄, 0.25 g NH₄Cl, 0.5 g KCl, 0.15 g CaCl₂ × 2 H₂O per 1 L double distilled water (ddH₂O). As a redox indicator, 1 mL of a 0.4 mg/mL Na-resazurin solution was added. The medium was autoclaved for 30 min at 121°C. Then, 50 mM bicarbonate (final concentration) buffer was added and the medium was reduced with 1 mM titan-(III)-nitrilotriacetate [final concentration (Moench and Zeikus, 1983)]. As an additional sulfur source, 200 μM Na₂S was added (final concentration). 1 mL trace element solution SL10 (DSMZ medium 141), 1 mL selenite-tungstate solution (DSMZ medium 385), and 0.5 mL 7-vitamin solution (Widdel and Pfennig, 1981) were also added. Finally, the pH was adjusted to 7 with 1 M HCl and the medium was distributed into sterile serum flasks or culture tubes, as appropriate.

The initial enrichment culture, its sub-cultivations, and the ultimately obtained highly enriched mixed culture (consortium) were grown with SQ in anoxic Hungate-type tubes or serum bottles (up to 50 mL) sealed with butyl rubber stoppers. The consortium was later also grown in triplicate in 125 mL flasks for growth experiments and DNA extraction for metagenomics and Illumina amplicon sequencing. For total-protein and total-RNA preparations for metaproteomics and metatranscriptomics, respectively, the consortium was grown in triplicate in 500 mL flasks. All cultures were incubated at 28°C in the dark without shaking. Cultures in which the headspace (N₂/CO₂, 80%/20%) was exchanged for H₂/CO₂ (79%/21%) were incubated horizontally, in order to increase the surface area available for gas exchange.

Faecalicatena sp. DSM22707 was purchased from the Leibniz-Institute DSMZ – German Collection of Microorganisms and Cell Cultures GmbH (Braunschweig, Germany) and maintained in a complex medium based on DSMZ medium 339a, containing 5 g tryptone/peptone, 5 g yeast extract, 0.5 g NaHCO₃, 0.5 mL Na-resazurin (stock, 0.4% w/v), 10 mL salt solution (0.25 g CaCl₂ × 2 H₂O, 0.50 g MgSO₄ × 7 H₂O, 2.00 g NaCl dissolved in 500 mL ddH₂O), 5 mL phosphate buffer (1.00 g K₂HPO₄, and 1.00 g KH₂PO₄ in 500 mL ddH₂O), 10 mL L-cysteine solution (0.25 g L-cysteine-HCl x H₂O/10 mL ddH₂O), 0.5 mL vitamin K₁ solution (100 mg/L, DSMZ Medium 78) and 5 mL hemin solution (0.5 mg/mL, DSMZ Medium 78) per 1 L ddH₂O. Phosphate buffer, L-cysteine, vitamin K_1, and hemin solutions were added after autoclaving. The pH was adjusted to 7.0 by the addition of 0.5 M Na₂CO₃ or 1 M HCl. For growth experiments with 10 mM SQ as substrate, strain DSM22707 was transferred to Hungate-type tubes containing the freshwater medium as described above, supplemented by 0.05% YE, and by hemin and vitamin K₁ at concentrations as in DSMZ medium 339a. For differential proteomics experiments, 10 mM glucose served as an alternative carbon and energy source for strain DSM22707.

Growth was monitored by measuring optical density (OD) at 600 nm in samples taken from the cultures in plastic cuvettes using a Pharmacia Novaspec II spectrophotometer (LKB Biochrom, United Kingdom) or directly in Hungate tubes using a Camspec M107 spectrophotometer (Leeds, UK). Substrate turnover and product formation were monitored by HPLC, GC, or LC–MS (see below).

Cultures were examined by phase-contrast microscopy using a ZEISS Axiophot (Jena, Germany) microscope and a 1.3 EC Plan-NEOFLUAR objective at either 40× or 100× magnification; the microscope was equipped with an Axiocam 305 color camera. Images were acquired using AxioVision software release 4.7.2. Images were processed using Fiji Image J 1.52p (Schneider et al., 2012). Cells were embedded on agarose-coated slides, prepared according to the method described by Pfennig and Wagener (1986), but instead of agar, a 2% agarose solution was dispersed on clean microscope slides; the agarose slides were dried overnight in a dust-free box prior to use.

We established several bacterial enrichment cultures with chemically synthesized SQ as the sole carbon and energy source for both aerobic (e.g., Felux et al., 2015; Frommeyer et al., 2020; Denger et al., 2012) and strictly anaerobic heterotrophic growth. For one of the latter, we provided 10 mM SQ in a titanium(III)-nitrilotriacetate-reduced freshwater mineral salt medium without any additional electron acceptor (except CO₂), and inoculated it with an iron(II)-sulfide-rich (black) sediment sample from Lake Constance. The culture remained active after several transfers and, interestingly, did not convert SQ to the C₃-intermediates DHPS or SL, but to the C₂-intermediate ISE, which was new at that time. Furthermore, it produced H₂S (and no methane), suggesting that ISE was further utilized for organosulfonate respiration in the enrichment culture (see below). During subcultivation, the enrichment culture was subjected to agar-shake dilution in several stages with SQ as substrate, in an attempt to isolate the SQ-degrading community member; however, when individual colonies were picked back into SQ-liquid medium, mixed cultures were always obtained, as visible by microscopy, consisting of rod-shaped and filamentous cells (Figure 2). The rod-shaped organisms appeared to be the most abundant, and the filamentous bacterium appeared to produce terminal spores (Figure 2). Since our many attempts to isolate the primary SQ-degrading, ISE-producing strain were not successful, we proceeded to characterize the consortium by molecular methods.

The phylogenetic composition of the consortium was first examined by Illumina amplicon sequencing of the V3 to V5 region of the 16S rRNA gene, which allowed for the assignment of amplicon sequence variants (ASVs) up to the genus level and for calculation of relative abundances (Figure 3). Contributing 73.1% of the total reads was an ASV affiliated to the [Eubacterium] fissicatena_group (genus Faecalicatena) within the family Lachnospiraceae of the phylum Bacillota [Firmicutes], a SILVA 138 database (Gaspar, 2018) group with 371 sequence entries, the majority of them from uncultured bacteria, but also from some isolated strains such as from Faecalicatena contorta [Eubacterium contortum]. Its high relative abundance strongly suggested that this phylotype was the primary SQ degrader. Contributing 19.5% of the total reads was an ASV affiliated with the genus Anaerospora within the family Sporomusaceae of Bacillota. Additionally, a small proportion of reads (7.4%) were affiliated to the genus Lachnoclostridium of the family Lachnospiraceae.

We also sequenced the metagenome of the consortium (accessible via the JGI-IMG system, Taxon ID: 3300047135), resulting in three metagenome-assembled genomes (MAGs), each with an estimated completeness of 100% and a redundancy of <5%. Furthermore, two complete 16S rRNA genes could be recovered (IMG gene IDs, Ga0499732_170_207_1739 and Ga0499732_174_24_1707). One of the MAGs was classified as belonging to the genus Faecalicatena (Muricomes in the Genome Taxonomy Data Base; Parks et al., 2022). With its complete 16S rRNA gene sequence, we found through Blastn (Boratyn et al., 2013) Faecalicatena sp. DSM22707 with 99.32% identity (Supplementary Table S1) as the closest isolate that is available in a culture collection. We tested whether Faecalicatena sp. DSM22707 could grow with SQ, which it could, as described further below. For the other two MAGs, we could not find any close relatives available in culture collections for testing. The second complete 16S rRNA gene as obtained from the Negativicutes MAG, was assigned to Sporomusaceae sp. 2C, with 96.09% identity (query coverage 84%, Supplementary Table S2). However, the alignment of this 16S rRNA gene with the Anaerospora ASV gene fragment obtained by Illumina sequencing (see above), showed 99% identity. Therefore, we refer to this MAG as Anaerospora.

The growth of the consortium, and Faecalicatena sp. DSM22707 in pure culture, with SQ, was followed in detail, to confirm substrate disappearance and to identify and quantify any products formed. The consortium (Figure 4A) completely degraded SQ (8.7 ± 0.2 mM) within 7.5 d (growth rate μ_max = 0.02 h⁻¹) concomitant with biomass formation (max. OD_600nm, 0.17 ± 0.01) and ISE formation (5.4 ± 0.4 mM). In addition to ISE, small amounts of DHPS were detectable when the cultures entered the stationary phase (68 ± 5 μM; not shown in Figure 4A), but no other organosulfonates such as SL, SAc, SLA, 4-deoxy-4-sulfoerythrose or 4-deoxy-4-sulfoerythrulose (see below). Another product released into the culture fluid in high amounts was acetate (14.7 ± 0.3 mM final concentration). When the cultures entered the stationary phase, small amounts of formate (0.97 ± 0.09 mM; not shown in Figure 4A), but no other fermentation products such as butyrate or succinate (see below) were detectable. In the headspace of the cultures, molecular hydrogen (H₂) production was detected concomitant with growth (up to 3.5 ± 0.1 mM H₂) (see Supplementary Figure S1). Only small amounts of sulfide (0.23 ± 0.17 mM) were detected in the stationary phase. With continued incubation, we observed that the H₂ and the ISE concentrations decreased significantly but not completely (see Supplementary Figure S2), and we considered whether the growth of the ISE degrader in the consortium might have been limited by insufficient availability of a suitable electron donor. Therefore, the growth experiment was repeated (Figure 4B) and following the conversion of all SQ to ISE (at day 7 of incubation), we replaced the headspace from N₂/CO₂ to H₂/CO₂ (80/20%). After an additional incubation of 14 d, no ISE was detected and the sulfide concentration increased (1.6 ± 0.1 mM).

The pure culture of Faecalicatena sp. DSM22707 grew with SQ or glucose, each in Ti(III)-reduced freshwater mineral salt medium (Figure 4C), only in the presence of small amounts of yeast extract as a supplement (YE, 0.05% w/v); it therefore exhibited a first, short growth phase with YE (Supplementary Figure S3). Strain DSM22707 degraded SQ (6.9 ± 0.7 mM, n = 4) completely within 5 d (μ_max = 0.03 h⁻¹) concomitant with biomass formation (max. OD_600nm, 0.47 ± 0.01) and ISE formation (final concentration, 5.8 ± 0.4 mM). No DHPS or SAc was detectable; however, small amounts of SL (up to 0.2 ± 0.024 mM) were detected after growth, as well as trace amounts of SLA and SAA, as identified by high-resolution MS (Supplementary Figures S4, S5); for the latter compounds, we had no standards available for quantification. Other products found to be released into the culture fluid in significant amounts were acetate (final concentration, 7.8 ± 0.1 mM), butyrate (final concentration, 1.1 ± 0.01 mM), and small amounts (not shown in Figure 4C) of succinate (0.16 ± 0.05 mM) and formate (0.13 ± 0.09 mM). No H₂ was detected in the headspace after growth. For comparison, strain DSM22707 grown with glucose (6.41 ± 0.23 mM) produced acetate (7.87 ± 0.28 mM), butyrate (2.75 ± 0.12 mM), succinate (0.40 ± 0.04 mM), formate (0.48 ± 0.01 mM), but also no H₂.

In order to reproduce the sulfolytic, two-step of SQ by the consortium (Figure 4B) we used a defined co-culture model. For this, the growth experiment with Faecalicatena sp. DSM22707 was repeated (Figure 4D). After SQ was completely utilized by Faecalicatena sp. DSM22707 and ISE was produced (at day 4 of incubation), and the culture fluid was filter-sterilized and inoculated with the known ISE-degrading, sulfide-producing isolate Bilophila wadsworthia 3.1.6 (Peck et al., 2019; Baron et al., 1989). Additionally, the headspace was switched from N₂/CO₂ to H₂/CO₂. Strain 3.1.6 completely degraded ISE within 4 d, concomitant with further increases of sulfide (from 0.2 ± 0.01 mM to 2.9 ± 0.2 mM) and acetate (from 8.2 ± 0.1 mM to 10.5 ± 0.1 mM) (Figure 4D).

In order to examine which enzymes and genes could be employed for SQ and ISE degradation by the consortium and by which phylotype, we analyzed the consortium’s total proteome and transcriptome. All genes that were found to be both transcribed and translated at relevant levels, as compared to housekeeping genes, were screened for homologs of known genes for SQ and ISE degradation. In Figure 5, an excerpt of the data is shown with the identified candidates for SQ metabolism (Figure 5A), encoded by two gene clusters (contigs) each affiliated to Faecalicatena, and for ISE metabolism (Figure 5B) encoded by one gene cluster affiliated to Anaerospora, compared to representative housekeeping genes (Figure 5C). Furthermore, pure cultures of Faecalicatena sp. DSM22707 grown with SQ compared to those grown with glucose were analyzed by differential proteomics of the total proteome. Therefore, its draft genome was sequenced and annotated (accessible via the JGI-IMG system, Taxon ID 2956855026). Figure 6 shows a volcano plot for the differential proteomics dataset, highlighting all genes that were specifically induced during SQ utilization. It identified a gene cluster nearly identical to that of the consortium, comprising the same sets of candidate genes for SQ metabolism in strain DSM22707 (Figure 7A). Interestingly, these clusters included candidate genes for key enzymes of both the sulfo-TK pathway (transketolase, leading to C₂-sulfonates) and the sulfo-TAL pathway (transaldolase, leading to C₃-sulfonates) (Figure 7B), as described in the following. Because of the high identity of both gene clusters, in the text below we mostly refer to the IMG gene IDs (295685xxxx) of Faecalicatena sp. DSM22707 proteins, and their percent amino-acid sequence identities (% id-aa) relative to genes of the archetype sulfo-TK pathway of Clostridium sp. MSTE9 (Liu et al., 2021) and/or sulfo-TAL pathway of P. aryabhattai SOS1 (Frommeyer et al., 2020). For the corresponding IMG locus tags in the metagenome, see Figures 5, 7.

For liberation of the SQ headgroup from SQ glycoconjugates, a sulfoquinovosidase (SQase) candidate (IMG gene ID 2956859258) with 35% identity to the recently described NAD⁺-dependent sulfoquinovosidase of Arthrobacter sp. strain AK01 (Kaur et al., 2023) was found to be induced in Faecalicatena sp. DSM22707 during SQ growth (Figure 6). Such an SQase candidate was also detected by transcriptomics and proteomics in the Faecalicatena-phylotype of the consortium (Figure 5A). These respective SQase genes were encoded next to candidate SQ ABC transport genes for Faecalicatena sp. DSM22707 (Figure 7A) and the consortium. Furthermore, a candidate SQ mutarotase for catalyzing the anomerization between SQ anomers (Liu et al., 2021; Frommeyer et al., 2020) was produced from both Faecalicatena gene clusters (Figures 5A, 6, 7A), for strain DSM22707 (2956859254), which showed52 and 48% id-aa to SQ mutarotase candidates of the sulfo-TK pathway of Clostridium sp. MSTE9 (2524472434) and the sulfo-TAL pathway in P. aryabhattai SOS1 (2667550652), respectively (Figure 7B); these candidates showed no significant homology to the SQ mutarotase enzyme of Herbaspirillum seropedicaea (Abayakoon et al., 2018). Furthermore, SQ isomerase (2956859243) candidates, which showed 65 and 32% id-aa to the SQ isomerases of strains MSTE9 (2524472426) and SOS1 (2667550654) (Figure 7B), were produced (Figures 5A, 6) from each gene cluster, for conversion of SQ to SF. Further downstream in the clusters, two candidate genes each for transketolase subunits were expressed (Figures 5A, 6). The first subunit (2956859247) showed 63% id-aa to the SF transketolase subunit G of strain MSTE9 (2524472430), containing a thiamine diphosphate binding domain (Lindqvist et al., 1992; Nikkola et al., 1994), and the second (2956859248) showed 68% id-aa to the SF transketolase subunit H of strain MSTE9 (2524472431), containing the C-terminal domain with proposed regulatory activity (Nikkola et al., 1994) (Figure 7B). The gene (2956859249) directly downstream was also expressed (Figure 6), encoding a candidate 4-deoxy-4-sulfoerythrulose (SEu) isomerase, with 58% id-aa to the archetype enzyme of the sulfo-TK pathway (2524472432). Interestingly, also two different transaldolase candidate genes further downstream (Figure 7) were produced (Figure 5A), in strain DSM22707 specifically during growth with SQ but not with glucose (Figure 6). The first (2956859250) showed high identity (62% id-aa) to the archetype SF transaldolase enzyme of the sulfo-TAL pathway of P. aryabhattai SOS1 (2667550651), while the sequence identity for the second (2956859251) was lower (32% id-aa, 19% query cover) and its protein-family (pfam) entry was fructose-6-phosphate aldolase (see also below).

For the conversion of the SAA (sulfoacetaldehyde) produced by the SF transketolase reactions, candidate SAA reductases were encoded in both Faecalicatena gene clusters, of which that of strain DSM22707 (2956859246) was SQ-inducible produced (Figure 6) and showed 64% id-aa to that of Clostridium MSTE9 (2524472429). Candidate genes for SLA (3-sulfolactaldehyde) reductases and SLA dehydrogenases were found elsewhere in the consortium metagenome (Supplementary Table S3) or the Faecalicatena sp. DSM22707 genome (Supplementary Table S4), for the conversion of the SLA to DHPS or SL, respectively. Three candidates for SLA oxidation or reduction were found to be produced by the consortium as follows: one SLA dehydrogenase candidate with 31% id-aa to that of P. putida SQ1 (Felux et al., 2015) in the Faecalicatena-phylotype and two SLA reductase candidate in both the Anaerospora-phylotype with 37% id-aa and another one in the Faecalicatena-phylotype with 30% id-aa to E. coli (Denger et al., 2014). In the pure culture of strain DSM22707, candidates for both reduction and oxidation of SLA were produced: for SLA reduction, an annotated 2-hydroxy-3-oxopropionate reductase (2956855439) with the highest id-aa (35%) to the SLA reductase of E. coli (Denger et al., 2014), although it was produced only at very low levels and not differentially. Other proteins encoded in the clusters that were also detected to be SQ-inducible produced (Figure 6) are candidates for glycerol dehydrogenase (2956859244) and dihydroxyacetone kinase (2956859245).

Other genes in the clusters encoded transporters, that might facilitate the export of the organosulfonate intermediates, i.e., a drug/metabolite permease (2956859252), or the import of SQ, i.e., a predicted ABC-type transporter (2956859255–2956859257), for which the substrate binding protein (2956859255) was confirmed to be produced during growth with SQ (Figure 6). The substrate binding protein (2956859255) showed 32% id-aa to the characterized SQ binding protein of the ABC transporter of Agrobacterium tumefaciens (Gene Bank: AAK90108.1) (Snow et al., 2022). The permease protein (2956859257) showed 37% id-aa to the corresponding protein of Rhizobium leguminosarum (Li et al., 2020) and 36% id-aa to that of Clostridium sp. MSTE9 (2524472437). Furthermore, the repressor candidate produced (2956859253) (Figure 6) showed 47% id-aa to that of Clostridium sp. MSTE9 (2956859233).

For the degradation of ISE, an Anaerospora-phylotype gene cluster was identified due to its high expression (Figure 5B) and high homology to the ISE-utilization gene cluster known for B. wadsworthia (Figure 7C). The gene cluster encoded the candidate isethionate sulfite lyase (IslA) and its corresponding activating subunit (IslB), which showed high identity (66 and 48% id-aa, respectively) to those of B. wadsworthia 3.1.6 (Supplementary Table S5). Additionally, this gene cluster encodes for SAA reductase and CoA-acylating acetaldehyde dehydrogenase (Supplementary Table S5), such as in B. wadsworthia (Figure 7C), and these genes were also expressed (Figure 5C). Further proteins encoded in the gene cluster (Figure 7C) and expressed during growth of the consortium with SQ (Figure 5C), are candidate transporters for ISE uptake and/or sulfite/sulfonate export (Supplementary Table S5), and ethanolamine utilization (Eut) proteins EutN, EutQ and EutM. The Eut proteins are components for bacterial microcompartments (BMCs) which likely harbor the IslAB and other enzymes, as described for B. wadsworthia (Burrichter et al., 2021). The dissimilatory sulfite reductase (Dsr) complex alpha and beta and DsrC genes, which were found to be highly expressed (Figure 5C) were also affiliated with the Anaerospora-phylotype (Figure 7C and Supplementary Table S5).

AlphaFold was used to predict (Jumper et al., 2021) the structure of the two subunits G and H of the putative split gene candidate SF transketolase of Faecalicatena sp. DSM22707 (2956859247 and 9248, respectively), in multimer mode, to form an ⍺₂β₂ heterotetramer, which resembles structurally and biochemically split gene transketolase of Carboxydothermus hydrogenoformans (PDBcode: 6yak) (Supplementary Figure S6). A comparison at the residue level was performed against the single-gene transketolase of Scheffersomyces stipitis (PDBcode: 5XU2), as this transketolase contains all the necessary cofactors (see below). In addition, the SF transketolase model of Liu et al. (2021) was included for structural comparison (Supplementary Table S6). The residues were numbered according to full-length transketolase (5XU2). As expected, the predicted model revealed a similar domain arrangement to the two crystal structures compared to (6YAK and 5XU2), with two active sites within the ⍺₂β₂ heterotetramer (Figure 8).

Transketolases are thiamine pyrophosphate (TPP) and calcium (Ca²⁺) dependent enzymes, and the Faecalicatena transketolase ⍺₂β₂ heterotetramer has all the necessary residues to bind these cofactors (Figure 8). The calcium ion of 5XU2 is coordinated by the lateral groups of aspartate and asparagine and the main chain of the isoleucine. The residues are strictly conserved in the transketolase subunit G domain (2956859247) of Faecalicatena sp. DSM22707 (residues D155, N185 and I187). The TPP binding site is located at the dimer interface in 5XU2, where it forms a curved cavity that results in a high-energy, V-conformation state of the bound TPP (James et al., 2020; Leeper et al., 2005). The identical TPP coordinating residues can be found in the heterotetramere studied here indicating the same TPP binding capabilities. The difference between the enzymes lies in the sugar head group orienting residues, whereas the S. stipites and C. hydrogenoformans transketolases employ an RSFHR motif for the orientation of the phosphate moiety, while the corresponding SF transketolase subunits harbor an RAFDR motif. This alteration most likely leads to a different substrate specificity (Payongsri et al., 2012, 2015).

The AlphaFold structure predictions of the two different transaldolase candidates (Figure 9, Supplementary Figure S6) produced from the Faecalicatena sp. DSM22707 gene cluster (see above) were compared to the recently published cryo-EM structure of the SF transaldolase of P. megaterium (BmSF-TAL), which organizes into a double ring-like structure of a dimer-of-pentamers holo decamere (Snow et al., 2023). In Figure 9, the first encoded Faecalicatena transaldolase candidate (2956859250) is labeled F1, and the second candidate (2956859251) is labeled F2. As each subunit folds into a TIM barrel that harbors an active site, the comparison was only made at the monomer level to reveal potentially similar/differing catalytic capabilities. The postulated sulfonate-orienting WAR motif of BmSF-TAL was indicated to be present also in the first candidate F1 but appeared to be different in the second one, F2 (Figure 9). For transaldolase F1, we found the same residues as in BmSF-TAL for the catalytic triad (D6, K94, E66) and the sugar-orienting residues (N166, S138), while the catalytic triad in the second transaldolase candidate F2 (D23, K140, Q111) and its sugar-orienting residues (N162, T201) are different compared from BmSF-TAL (D6, K89, E61 and N111, S133, respectively).