Proteins perform their biological function following specific sequences of events. During these dynamical paths, highly non-trivial cooperative interactions occur. Ultimately, this is the origin of the emerging collective behavior that makes proteins the most sophisticated existing molecular machines. This complex network of processes covers a wide range of timescales, from few fs to ms, and distances, from atoms to large protein domains.Even the most recent experimental techniques generally provide ns-to-us averaged structural and dynamical information, often in non-physiological conditions. To access simultaneously atomic time and length scales would unveil the elementary conformational steps constituting a functional event and their temporal evolution.I propose to extend emerging multidimensional ultrafast optical spectroscopic techniques to the deep ultraviolet. These techniques are the analogue of multidimensional Nuclear Magnetic Resonance methods and are able to provide structural information exploiting electric dipole couplings but with fs temporal resolution. The novel extension to ultraviolet, that I shall implement, will open the possibility to exploit the optical absorption of aromatic amino-acid residues with the great advantage of studying wild type proteins. In this way, all drawbacks due to artificial labeling will be ruled out. I will use this new technique to study dynamic-assisted long range electron transfer in copper proteins and enzyme regulation in hemoglobin. These two proteins of great importance from a biological point of view have been chosen because their functions are a clear manifestation of cooperative phenomena. On a long term prospective this methodology will be a universal tool applicable to any wild type protein containing aromatic amino acids.