Functional classification and structural characterization of immune regulatory proteins

Abstract

Proteins participate in, and are central to, many processes associated with cellular function. Among the many different roles that proteins fulfill are the catalysis of chemical reactions and the mediation of complex processes including the immune response and cell motility. Currently, there are approximately ten million known protein sequences, a number that is exponentially growing. However, only a fraction of the sequenced proteins have had their three-dimensional structure determined or have an experimentally validated functional annotation. The current project attempts to bridge the gap between structurally/functionally characterized and uncharacterized proteins by clustering the proteins into evolutionary and, in particular, functionally related protein families. This type of the family classification can be utilized to support the selection and prioritization of "interesting" protein targets for which structure determination is most likely to be informative. Specifically, amino acid sequence signatures derived from a family can be used to predict the presence of unique structural features that are directly responsible for a specialized biological function. These sequence signatures can also be used to identify and classify new family members, thus expanding our understanding of sequence-structure-function relationships.;Currently, the most common approach for identifying evolutionary and functional relationships between proteins is through sequence similarity. This approach is limited as it is prone to false positive errors and requires a relatively high percent of identical residues between the compared sequences, resulting in many unclassified proteins. In the first part of my thesis, I focused on the development and application of novel computational tools that use current genomic sequencing data to classify proteins into families and to predict biologically important protein structural and functional features. Chapter II describes a computational method we developed to analyze correlated mutation patterns in multiple sequence alignments in order to predict disulfide bond connectivity, which is important for protein stability, folding rate, and in some cases has a direct functional role. Therefore, disulfide bond connectivity patterns can be used to facilitate structural and functional annotation of proteins.;Chapter III introduces two computational approaches, one that utilizes physical genome map information to search for gene duplication events as a mean to identify evolutionary relationship without sequence consideration. The other approach detects similarities between pairs of sequence profiles instead of directly comparing the sequences themselves. These methods aim to overcome the limitations of existing approaches, which require high sequence identity when classifying proteins. We applied both of these methods to explore functional classes within the immunoglobulin superfamily (IgSF), which makes up the majority of immune costimulatory and cell adhesion proteins. These proteins are critical regulators of the immune response and thus are valid targets for protein-based therapeutics to treat a wide range of human diseases.;Although all IgSF proteins share similar structural features, they often have less than 30% sequence identity, resulting in a large variety of structures and functions that are difficult to properly classify. Remarkably, clustering with our scoring schemes produced groups that are highly correlated with experimentally validated functional subfamilies within the IgSF. Our method also identified new members of these families, immediately providing functional hypotheses. Chapter IV describes the 2.4 A resolution X-ray crystal structure of the class-I MHC-restricted T-cell-associated molecule (CRTAM) - a newly defined member of the nectin/nectin-like family. Consistent with the computational analysis, the structure of CRTAM showed a high similarity to another structure from the nectin-like family thereby validating its placement within this family. The X-ray structure of CRTAM also provided important insights into the ligand recognition mechanism of this important co-stimulatory protein and allowed for the construction of a model describing its binding interactions.;The methods developed in this thesis represent new tools to cluster immune regulatory proteins into evolutionary and functional families. This classification supports the selection of protein targets for structural determination by prioritizing proteins that are expected to reveal new structural features, functional mechanisms, and evolutionary insights.