The central dogma of molecular biology states that information flows from DNA to protein via RNA (Crick, 1970). This model is central to our understanding of biology but can lead to the assumption that changes in transcription and transcripts will inevitably lead to changes in protein levels, and so directly impact the metabolic and biosynthetic state of the cell. To test this assumption, we used a biological system characterised by genome-wide, cyclical changes in transcription, to assess whether changes in transcription are reflected in changes at the level of protein. We reveal that despite large changes in transcription at the majority of genes, there is little change in protein. This decoupling results from the slow rate of protein turnover. The changes protein activity we did observe were instead a reflection of the metabolic state of the cell, resulting from post-transcriptional modifications such as acetylation and phosphorylation, that in turn drive the cycling of processes such as transcription and ribosome biogenesis. Thus, transcriptional and transcript cycling reflects rather than drives the metabolic and biosynthetic changes during biological rhythms. We suggest that caution is needed when inferring the activity of biological processes from transcript data, as this reflects but does not predict a cell's state.

Isomerization reactions are fundamental in biology. Lactate racemase, which isomerizes L- and D-lactate, is composed of the LarA protein and a nickel-containing cofactor, the nickel-pincer nucleotide (NPN). In this study, we show that LarA is part of a superfamily containing many different enzymes. We overexpressed and purified 13 lactate racemase homologs, incorporated the NPN cofactor, and assayed the isomerization of different substrates guided by gene context analysis. We discovered two malate racemases, one phenyllactate racemase, one α-hydroxyglutarate racemase, two D-gluconate 2-epimerases, and one short-chain aliphatic α-hydroxyacid racemase among the tested enzymes. We solved the structure of a malate racemase apoprotein and used it, along with the previously described structures of lactate racemase holoprotein and D-gluconate epimerase apoprotein, to identify key residues involved in substrate binding. This study demonstrates that the NPN cofactor is used by a diverse superfamily of α-hydroxyacid racemases and epimerases, widely expanding the scope of NPN-dependent enzymes.

No abstract