As a model project for the EFI which included extensive use of the EFI Microbiology Core, this work showed the strengths of combining a suite of disciplines to develop a whole-scale understanding of a not just a single enzyme, but of an entire metabolic pathway. On the basis of this and other efforts that have successfully blended methodologies, pathway elucidation has become a major driver as the EFI has evolved.
Functional assignment of uncharacterized proteins is a challenge in the era of large-scale genome sequencing. Here, we combine in extracto NMR, proteomics and transcriptomics with a newly developed (knock-out) metabolomics platform to determine a potential physiological role for a ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO)-like protein from Rhodospirillum rubrum. Our studies unraveled an unexpected link in bacterial central carbon metabolism between S-adenosylmethionine-dependent polyamine metabolism and isoprenoid biosynthesis and also provide an alternative approach to assign enzyme function at the organismic level.
Figure 1. (a) Phylogenetic tree of the RubisCO/RLP superfamily. The unrooted phylogenetic tree is based on amino acid sequence analysis of 333 representative proteins (listed in Supplementary Table 6) that were restricted to a length of 409 amino acids and aligned with ClustalW. Tree topography and evolutionary distance are given by the neighbor-joining method. Numbers at nodes represent the percentage bootstrap values for the clades of this group in 500 replications. Similar trees were obtained by using the minimum evolution and the maximum likelihood method. Extended views of each subtree are in Supplementary Figures 12 and 13. The scale bar represents a difference of 0.1 substitutions per site. RubisCOs are classified into three well-established subfamilies2, 29 (I, II and III). RLPs fall into six different subfamilies as described previously2, 29: IV-AMC, metagenomic Leptospirillum sequences from an acid mine consortium; IV-DeepYkr, R. rubrum group including mainly α- and γ-proteobacteria, some thermophilic species and Veillonellaceae; IV-YkrW, B. subtilis group including many Bacilliales, Acidithiobacillales and cyanobacteria; IV-GOS, metagenomic sequences from the global ocean sequencing program; IV-Photo, C. tepidum group, including many Chlorobiales and alphaproteobacteria; IV-NonPhoto, including many α- and some β-proteobacteria. A seventh subgroup of RLPs, established in this extended phylogenetic analysis (IV-Aful, including Clostridiales and Archaeoglobus fulgidus) was described previously as a singleton (A. fulgidus DSM 4304)2. (b) Function of the B. subtilis RLP in the classical methionine salvage pathway. MTR, methylthioribose; DKMTP-1P, 2,3-diketo-5-methylthiopentyl-1-phosphate; HKMTP-1P, 2-hydroxy-3-keto-5-methylthiopentenyl-1-phosphate; HKMTP, 1,2-dihydroxy-3-keto-5-methylthiopentene; KMTB, 2,4-keto-4-methylthiobutyrate. Amino acids are listed by their three-letter codes.
Figure 2. 1H-NMR spectra of R. rubrum cell extract incubated with MTA that show the transformation of MTA into MTR-1P and MTRu-1P over time. We incubated 0.78 mg cell extract protein of R. rubrum grown on MTA as their sole sulfur source with 0.4 mM MTA at 30 °C. Spectra were recorded at different time points as indicated. Characteristic 1H-NMR signals for MTA (green), MTR-1P (purple), and MTRu-1P (blue) are indicated by colored lines. The full array experiment is shown in Supplementary Figure 1.
Figure 3. (a) Time-dependent formation of intermediates in the MTA-isoprenoid shunt upon MTA feeding. Cell suspensions of R. rubrum (1 ml, D578 nm = 6.0) were incubated with 0.4 mM MTA and analyzed after 2 min, 10 min and 20 min by LC-FTMS metabolomics. The data are represented as extracted ion chromatograms (negative ion) at 2-p.p.m. mass accuracy for 259.00468 (MTR-1P and MTRu-1P) and 213.01696 (DXP). (b) Time-dependent formation of free thiols by R. rubrum upon MTA uptake. Cell suspensions of R. rubrum (1 ml, D578 = 4.0) were incubated with 0.4 mM MTA. The supernatant was analyzed for consumption of MTA and formation of free thiols. Free thiols formed were identified as methanethiol by HPLC and LC-FTMS (Supplementary Fig. 4). At least two independent cell batches were used in these assays. Data represent mean values ± s.d. (c) LC-FTMS metabolomics analysis of MTA-isoprenoid shunt mutants. Cell suspensions of R. rubrum wild type (WT) and different mutants were incubated with MTA (according to a) and analyzed after 10 min by LC-FTMS metabolomics (detailed analysis of the cupin mutant is in Supplementary Fig. 6). The data are represented as extracted ion chromatograms (negative ion) at 2-p.p.m. mass accuracy for 259.00468 (MTR-1P, MTRu-1P, MTXu-5P and MTRu-5P) and 213.01696 (DXP). (d) Thiol release activities by R. rubrum wild type and different mutants. Cell suspensions of R. rubrum were incubated with MTA according to b, and the formation of free thiols over time was quantified. At least two independent cell batches were used in these assays. Data represent mean values ± s.d.
Figure 4. The RLP is part of the central reaction sequence that involves the release of methanethiol from the molecule backbone. Whereas methanethiol can be recaptured as methionine via O-acetyl-L-homoserine sulfhydrylase (Rru_A0774), the rest of the molecule is converted into isoprenoid precursors. All of the intermediates of this proposed pathway that were identified by LC-FTMS metabolomics are shown in boxed bar charts, with their individual increase in metabolite level after 0 min, 10 min and 20 min feeding of MTA (+MTA, green bars) shown in comparison to control cells after 0 min, 10 min and 20 min (control, black bars). Data represent mean values ± s.d. Genes or proteins that were identified and characterized in this study are also shown and highlighted by colors.
Figure 5. Metabolite extracts prepared from R. rubrum wild-type cells and mutants that had been cultivated on minimal medium with sulfate as the sole sulfur source were analyzed for intermediates of the proposed MTA-isoprenoid shunt. The data are represented by extracted ion chromatograms (negative ion) at 2-p.p.m. mass accuracy for 296.08228 (MTA; retention time 3.5 min; green traces), 259.00468 (methylthiopentose phosphates: MTR-1P, MTRu-1P, MTXu-5P and MTRu-5P; retention time, 22 to 23 min; black traces) and 213.01696 (DXP; retention time, 24.5 min; red traces). The integrated intensities are inset and color coded for each sample. The single traces are scaled to 100% relative intensity. For comparison, integrated intensities are also summarized in Supplementary Table 7.
Figure 6. (a) Differential induced gel electrophoresis (DIGE) analysis of changes in the proteome of R. rubrum. Proteins upregulated in MTA-grown cells are shown in cyan, proteins upregulated in sulfate-grown cells are shown in red, and proteins that are not changed under both conditions are shown in white. The calculated pI (first dimension) and the molecular mass standards (second dimension) of the DIGE gel are given. More information on the five proteins circled in green that were selected for identification are in Supplementary Table 4. (b) RNA sequencing (RNAseq) analysis of changes in the transcriptome of R. rubrum. Genes annotated on the chromosome of R. rubrum are represented by dots according to their physical location on the chromosome and their fold change in mRNA level. Classical housekeeping genes are listed separately and highlighted by light gray dots in the graph. Genes of the thiol cluster shown in c are shown as cyan dots. Thiol cluster genes that were also identified by DIGE (a) are highlighted by green dots. More information on all of the transcripts upregulated more than 30-fold is in Supplementary Table 5. (c) The thiol cluster that was identified by both DIGE and RNAseq analysis. The four genes and transcripts of the thiol cluster that were identified by DIGE are numbered in green (according to their numbering in a) and highlighted in green. Transcripts of the thiol cluster that were identified by RNAseq are highlighted in cyan, and the fold upregulation of each gene is given separately in black numbers. Trsp, transporter; PLP, pyridoxal-5′-phosphate; 2-OxoGlu, 2-oxoglutarate.
Reprinted with permission from Nature Publishing Group.
© 2012, Rights Managed by Nature Publishing Group