Synthetic biology aims to engineer cells as miniature biological devices to sense, process, and respond to exogenous stimuli. Protein switches are designed to sense and respond to various molecular queues in a fast and specific manner, which fits the requirements of mammalian cell engineering. Especially, small-molecule responsive protein switches are particularly suitable for remotely controlling engineered cell functions, and could be significantly important in the context of therapeutic applications. My thesis leverages computational tools for the design of chemically responsive protein switches. We developed a strategy to repurpose drug-inhibited protein-protein interactions into OFF- and ON-switches controlled by preclinical or clinically approved drugs. The designed binders and drug-receptors form chemically-disruptable heterodimers (CDH) which dissociate in the presence of small molecules. Moreover, we converted the CDH into a multi-domain architecture which we refer to as activation by inhibitor release switches (AIR) that incorporate a rationally designed drug-insensitive receptor protein. CDHs and AIRs have been applied in regulating gene expression, protein degradation and signal transduction, and they showed excellent performance as drug responsive switches to control combinations of synthetic circuits in mammalian cells. Moreover, we anticipate that the CDHs and AIRs are highly modular for many other proximity-dependent applications. I further elaborated these designs to create a bifunctional switch to sense and respond to two chemicals, which is referred to as small-molecule controlled activation and repression integrator (sm-ART). sm-ART consists of a bifunctional scaffold that binds to two drug-receptor proteins in a mutually exclusive manner, representing two steady and opposing states (e.g., ON vs OFF) that can be switched by the corresponding drug-receptor inhibitor. The sm-ART regulated by ON- and OFF- signals have been engineered with cell surface receptors to reversibly control the signaling transduction. In a different line of research we discovered that the human Bcl2 protein homodimerize in the presence of its inhibitors, which led to the construction of three chemically inducible gene expression circuits assembling different chimeric transcription systems. Furthermore, the truncated human Bcl2 proteins were incorporated into a split heterodimeric CAR construct (ON-CAR) which can be switched ON in the presence of Bcl2's FDA-approved inhibitor, Venetoclax. The ON-CAR showed comparable killing activities to the standard second-generation CARs in the presence of Venetoclax in vitro and in vivo. We demonstrated that truncated human Bcl2 dimerize under the effect of its clinically approved drug, which opens up exciting possibilities for the engineering of safe and controllable CARs to meet many of the persistent challenges in clinical development of cell-based therapies. Altogether, my thesis presented a computational protein design blueprint to rationally design chemically responsive protein switches controlled by preclinical or clinically validated drugs. These protein switches are also useful tools in synthetic biology and cancer immunotherapy.