Chiba, G.; Kamei, K.-i. F.; Oda, A. H.; Ohta, K.; Wakamoto, Y.

Understanding how specific genetic changes lead organisms to different phenotypes is the cornerstone of genetics. However, even a single gene perturbation can cause global changes in gene expression profiles. Consequently, predicting which genes will be affected by a particular genetic perturbation remains a challenge. Here, we analyze changes in genome-wide gene expression profiles caused by single-gene perturbations in bacterial, worm, and human cells and show a simple rule that genes that intrinsically maintain the ratio of expression levels (stoichiometry) with a perturbed gene are more likely to be affected when it is deleted or inactivated. Furthermore, deleting a gene that maintains their stoichiometry with more genes tend to affect the expression levels of other genes genome-wide. Stoichiometry-conserving gene clusters can be found beyond local transcriptional regulatory units and metabolic pathways. Thus, the effects of genetic perturbations are propagated to other genes through a multi-level stoichiometry-conserving architecture in cells.