The phototrophic bacterium Chlorojlexus aurantiacus does not use any of the known autotrophic CO, fixation pathways. There is evidence for a new cyclic autotrophic pathway in which acetylCoA is converted to 3-hydroxypropionate and further to succinate and malate. This hypothesis was tested by feeding growing cultures during several generations with 3-hydroxy [ l-13C]propionate, [ l -I3C]acetate, or [2-13C]acetate, in addition to unlabeled CO,. The relative 13C content of individual carbon atoms in biosynthetic amino acids and nucleosides was determined by 'H-and I3C-NMR spectroscopy. I3C coupling patterns were analyzed by two-dimensional 13C-TOCSY experiments which were optimized for the analysis of multiply I3C-labeled biosynthetic samples. From the 13C enrichments of amino acids and nucleosides, the labeling patterns of central metabolic intermediates were evaluated by a retrobiosynthetic approach. Both 3-hydroxypropionate and acetate were incorporated into all central metabolic pools. The I3C labeling and coupling patterns suggest a novel carbon fixation pathway via 3-hydroxypropionate. Specifically, we propose that acetyl-CoA is carboxylated to malonyl-CoA which is reduced under formation of 3-hydroxypropionyl-CoA. Dehydration and reduction yield propionyl-CoA which is converted to succinate by a second carboxylation reaction. The net product of autotrophic carbon fixation appears to be glyoxylate. However, it is not yet known how glyoxylate is channeled into anabolic metabolism. Assimilation of acetate can proceed via the CO, fixation pathway, but also via the glyoxylate pathway.In bacteria, three pathways of autotrophic CO, fixation have been evaluated: the reductive pentose phosphate cycle (Calvin cycle) found in aerobic eubacteria, the reductive citric acid cycle, and the reductive acetyl-CoNcarbon monoxide dehydrogenase pathway found in anaerobic eubacteria and archaebacteria [ 11.Evidence for a fourth autotrophic CO, fixation pathway has recently been presented for Chloroflexus aurantiacus [2, 31, an anaerobic thermophilic phototrophic eubacterium [4, 51. This organism excretes substantial concentrations (5 mM) of 3-hydroxypropionate at the end of autotrophic growth on CO, plus H, [2]; this proves that the bacterium is able to synthesize 3-hydroxypropionate from CO, and H, alone. The question is whether 3-hydroxypropionate is a dead-end product formed from CO, in a side path (e.g. by a peculiar fermentation of storage polyglucose during light limitation in the late growth stage) or whether it is a true intermediate of the novel autotrophic carbon cycle which has not yet been investigated in detail. In the first case, exogenous 3-hydroxy- Traditionally, isotope incorporation studies are interpreted in the forward metabolic direction. In this case, the essential question is whether a given precursor can or can not serve as a precursor for the down-stream product of interest. A different approach was used in the present study. The fed I3C-labeled compounds were incorporated into all amino acids and nu...