For many biomedical and biotechnological applications, modulating allosteric coupling offers unique opportunities. Such efforts can largely benefit from efficient prediction and evaluation of allostery hotspot residues that dictate the degree of co-operativity between distant sites. In several hotspot mutants revealed by recent deep mutational scanning (DMS) experiments of a bacterial transcription factor, tetracycline repressor (TetR), we demonstrate that effects of allostery hotspot mutations can be evaluated by judiciously combining extensive unbiased and enhanced sampling molecular dynamics simulations. The results recapitulate the qualitative effects of these mutations on abolishing the induction function of TetR and provide a semi-quantitative rationale for the different degrees of rescuability to restore allosteric coupling of the hotspot mutations observed in the DMS analysis. Free energy landscapes and variations of structural and energetic properties of different functional states of the mutants show that the hotspot mutations perturb inter-domain coupling and/or intra-domain conformational properties relative to the wild type. Several mutations clearly perturb multiple properties, also highlighting the complexity of realistic situations. Thus, the results support our mechanistic model that hotspot residues contribute to allostery through distinct molecular mechanisms, a feature that we propose to be broadly applicable to many allosteric systems. Overall, our study provides general computational strategies for the identification and evaluation of hotspot mutations modulating allostery in proteins.