Proteins without a well-defined tertiary structure are intrinsically unstable and are prone to degradation by the 20S proteasome. In my thesis, I investigate the protection mechanisms of intrinsically disordered (ID) protein regions via protein interactions using the AP-1 complex as a model system. AP-1 is composed of c-Fos and c-Jun proteins, out of which c-Fos has a shorter half-life than c-Jun. Interactions by c-Jun were shown to prolong the lifetime of c-Fos, leading to the proposal of the nanny model. This mechanism, where weak protein interactions protect unstructured regions without an induced folding, however, has never been probed directly. Here I investigate the nature of the interactions of c-Fos with c-Jun and how changes in disordered regions contribute to changes in half-life. I use mutational analysis to provide insight into changes in degradation rate as a function of the binding affinity in the bound form with c-Jun. I designed five mutants at the structured regions of c-Fos affecting specific contact sites (L165V, L172V) or charge separation (E175D, E189D, K190R) with c-Jun of which both modulate c-Fos turnover, proportionally to their impact on binding affinity. Interestingly, removal of the disordered region in the complex beyond the structured domain is observed to decrease c-Fos half-life indicating their role in the stability of the complex. The finding suggests that the protein turnover by the 20S proteasome can be fine-tuned by both structured and unstructured regions between c-Fos and c-Jun, consistent with the proposed 'nanny' model. These results highlight a novel aspect of disordered regions present in the bound form (fuzziness) in regulating protein half-life via fine-tuning the association rates between the two proteins. First, it demonstrates that the protection of disordered regions from degradation could be achieved without inducing a stable structure as confirmed by ECD spectroscopy. Binding to a partner generates a fuzzy complex, where fuzzy regions in protein complexes can serve as a nonspecific transient anchor. Second, the protection of disordered regions can be achieved with many binding configurations in the bound state without decreasing the conformational entropy. Thus, the protective role of fuzzy interactions from the 20S proteasome could also provide a possible explanation for how low-complexity sequence motifs involved in higher-order protein structures might serve as selective inhibitors of proteolysis.