ABSTRACT
Post-translational modifications (PTMs) increase the diversity of the proteome and are vital to organismal life and therapeutic strategies. Deep learning has been used to predict PTM locations. Still, limitations in datasets and their analyses compromise success. Here we evaluate the use of known PTM sites in prediction via sequence-based deep learning algorithms. Specifically, PTM locations were encoded as a separate amino acid before sequences were encoded via word embedding and passed into a convolutional neural network that predicts the probability of a modification at a given site. Without labeling known PTMs, our model is on par with others. With labeling, however, we improved significantly upon extant models. Moreover, knowing PTM locations can increase the predictability of a different PTM. Our findings highlight the importance of PTMs for the installation of additional PTMs. We anticipate that including known PTM locations will enhance the performance of other proteomic machine learning algorithms.
ABSTRACT
Tools for on-demand protein activation enable impactful gain-of-function studies in biological settings. Thus far, however, proteins have been chemically caged at primarily Lys, Tyr, and Sec, typically through the genetic encoding of unnatural amino acids. Herein, we report that the preferential reactivity of diazo compounds with protonated acids can be used to expand this toolbox to solvent-accessible carboxyl groups with an elevated pKa value. As a model protein, we employed lysozyme (Lyz), which has an active-site Glu35 residue with a pKa value of 6.2. A diazo compound with a bioorthogonal self-immolative handle esterified Glu35 selectively, inactivating Lyz. The hydrolytic activity of the caged Lyz on bacterial cell walls was restored with two small-molecule triggers. The decaging was more efficient by small molecules than by esterases. This simple chemical strategy was also applied to a hemeprotein and an aspartyl protease, setting the stage for broad applicability.
