Proteins can assume a very large number of conformations (conformational substates), all concurring to its function. We present experimental evidence for the existence, in trehalose coated carboxymyoglobin, of a structured environment of the protein, tightly coupled to the heme pocket structure, as experienced by the bound CO molecule. This was evidenced by the strict correlation observed between the thermal evolution (300-20 K) of the CO stretching and of the water association bands in samples of carboxymyoglobin embedded in trebalose matrixes of different hydration. This observation put forward the coupling between the degrees of freedom of the matrix and those of the protein. In the driest sample, in which only tightly bound, nonremovable water molecules were present, temperature induced structural variations of both the heme pocket and the external matrix were small, even at room temperature. At variance, such variations were larger in two water richer samples in which their onset was already at similar to50 K. Further, the thermal evolution of the CO stretching and of the water association bands showed a single linear correlation for the drier samples in the whole temperature range investigated. The same correlation was observed for the water richest sample up to similar to180 K. that is, the temperature at which a dynamic transition for the protein motions has been recurrently observed by experimental and computational means, in water containing systems. The data presented enable us to suggest the existence of a rigid water dipole network, which extends throughout the sample, impeding structural heme pocket rearrangements, which imply charge displacement. This in turn brings about the lack of thermally induced variations of the stretching band of the bound CO, which reflects the distribution of taxonomic (A) and lower tier conformational substates. Accordingly, in agreement with previous suggestions, we speculate that, in solution, slaving of the protein internal dynamics to the dynamics of the external solvent is brought about by the "attempts" of the protein structure to match the rapidly evolving structure of the water dipole network.