Metabolites cross the membrane, the transcriptome does not follow: co-localization is required for the cyanobacterial response to a yeast partner

Interactions between phototrophs and heterotrophs are usually explained by cross-feeding, yet whether that exchange alone accounts for the phototroph's response has rarely been tested. We grew the sucrose-secreting cyanobacterium Synechococcus elongatus with the yeast Rhodotorula toruloides as a synthetic community, either in mixed co-culture or in membrane-separated co-culture (0.22 μm pore size), each with a matched axenic control. Metabolite exchange occurred in both configurations: sucrose reached the yeast and was consumed, and yeast-derived metabolites reached the cyanobacterium. The yeast reached a higher final density behind the membrane, whereas the cyanobacterium gained about 9% over its axenic control only in mixed co-culture. Yet the cyanobacterial transcriptome responded only with co-localization: expression of 51 genes changed consistently across time points (day-averaged q < 0.05), against one gene behind the membrane, and the number of differentially expressed genes rose from 20 on day 4 to 91 and 127 on days 6 and 8. Nine regulatory programs (iModulons), mainly nitrogen and photosynthesis, responded in mixed co-culture and none behind the membrane. Reactor configuration alone changed 40 genes in axenic culture. These results suggest that co-localization, in addition to metabolite exchange, was required for the heterotrophic partner to trigger a response in the phototroph. Co-culture studies that infer interactions from shared metabolites may miss effects that depend on spatial structure. Importance Partnerships between light-powered microbes and the heterotrophs living alongside them move large amounts of carbon in oceans and soils, and are now built on purpose to turn carbon dioxide into fuels and chemicals. Such partnerships are usually described by the chemicals the partners trade, so we asked whether those chemicals are enough. The cyanobacterium Synechococcus elongatus and the yeast Rhodotorula toruloides were grown together in one vessel, or in two chambers divided by a membrane that molecules cross but cells cannot. The yeast fed on cyanobacterial sugar in both, yet the cyanobacterium changed its gene activity only when the two shared a space. Partners in nature are usually clustered rather than evenly mixed, so closeness may contribute as much as the chemicals themselves. Keeping partners apart in engineered cultures can still deliver carbon to the heterotroph, but may not deliver the benefits of partnership to the phototroph.