T280 is required for Dig1mediated Ste12 activity: cells expressing Dig1-YFPT280A displayed a severely diminished transcriptional response, phenocopying the full deletion of Dig1. Strikingly, the closely situated residues T277 and S279 had the opposite effect in that they inhibit signal output: mutation of either residue leads to an increased pheromone response. Since these sites lack consensus MAPK sequence context, the mechanistic and architectural details of how they might exert their effects could be constrained through computational modeling, but not determined unequivocally. Similarly, since in this sensitized strain background, CFP-Ste12 required Dig1-YFP for full activity, the significance of the phenotypes of the Dig1 mutations for the native pheromone pathway remains to be determined. However, our current findings that closely situated mutations exert opposite effects are sufficient to suggest that multiple layers of Mepiroxol post-translational regulation can be superimposed to fine-tune quantitative signal output. Our results indicate that many putative sites of phosphorylation contribute to and adjust the input-output relationship of this model eukaryotic signaling system. We propose that multiple small influences of such individual phosphorylation events can endow signaling systems with plasticity and evolvability. Consistent with this view, in the motifs that we studied on Ste12 and Ste50, the presence of a minimum consensus requirement for a MAPK substrate arose relatively recently and is Lomitapide Mesylate conserved in closely related yeast species but not in more distantly related yeast species. In 
Dig1, where residues T277, S279 and T280 are not likely to be MAPK substrates, the presence of serine or threonine residues at similar positions in orthologs is conserved in closely related yeast species but not in more distantly related yeast species. In all cases, the putative phosphorylation events that we studied fall in regions predicted to be unstructured in Ste12, Ste50 and Dig1. Such unstructured regions are well suited to accommodate amino acid changes to generate new sites for post-translational events like phosphorylation. This plasticity and evolvability may be advantageous for finetuning the input/output relationship of the pathway. By contrast, the core activation mechanism of the pathway is highly conserved during evolution. Activation of the MAPK Fus3 is governed by a dual phosphorylation event and mutation of either site results in a complete loss of function. This dual phosphorylation motif in the kinase activation loop is conserved in MAPKs all the way from yeast to mammalian MAPKs such as Erk1 and Erk2. In conclusion, we suggest that new layers of post-translational regulation can be gained and lost to rapidly adapt quantitative system output in the face of changing selective pressure without compromising the core structural and functional integrity of key signaling proteins. We note that this idea suggests means to systematically design and alter signaling pathway components to introduce novel regulatory loops or sever existing ones, and to confer new regulatory properties to pathway specific kinases and phosphatases. An approach based on iterating the design and construction of such re-engineered signaling pathways guided by quantitative experimentation and interpreted via appropriate models, should facilitate design-based alteration of signaling systems to bring about desired cellular behaviors.