Logo
DeutschClear Cookie - decide language by browser settings
Gong, Jing (2010): Databases and computational interaction models of Toll-like receptors. Dissertation, LMU München: Faculty of Geosciences
[img]
Preview
PDF
Gong_Jing.pdf

9Mb

Abstract

Toll-like receptors (TLRs) recognize pathogen-associated molecular patterns (PAMPs) on invading organisms and are the first line of defense in innate immunity. To date, much has been learned about TLRs and their roles in autoimmune diseases are being unraveled. The autoimmune disease systemic lupus erythematosus (SLE) progresses as a consequence of the inappropriate recognition of self nucleic acids by TLRs. For the development of therapeutic approaches of SLE it is necessary to understand possible negative regulation mechanisms of TLR. Single immunoglobulin interleukin-1 receptor-related molecule (SIGIRR) is the best characterized TLR signaling inhibitor. It can interfere with the receptor complexes and attenuate the recruitment of downstream adaptors to the receptors. So far, the mechanisms of structural interactions between SIGIRR, TLRs and adaptor molecules are unknown. To develop a working hypothesis for these interactions, we constructed three- dimensional models for these single molecules based on computational predictions. Then, models of essential complexes involved in the TLR signaling and the SIGIRR inhibiting processes were yielded through protein-protein docking analysis. With the high-throughput genome sequencing projects, a central repository for the growing amount of TLR sequence information has been created. However, subsequent annotations for these TLR sequences are incomplete. For example, the indicated numbers and positions of leucine-rich repeat (LRR) motifs contained in individual TLR ectodomains are greatly distinct or missing in established databases. In this vein, we have developed a database of TLR structural motifs called TollML (http://tollml.lrz.de). It integrates all TLR protein sequences that have been identified to date. These sequences were semi-automatically partitioned into three levels of structural motif categories. The manual motif identification procedure provided TollML with the most complete and accurate database of LRR motifs compared with other databases that contain TLR data. LRR motifs are present not only in TLRs, but also in many other proteins. To date, more than 6,000 LRR protein sequences and more than 130 crystal structures of them have been determined. This knowledge has increased our ability to use individual LRR structures extracted from the crystal structures as building blocks to model LRR proteins with unknown structures. Because the individual LRR structures are not directly available from any protein structure database, we have developed a conformational LRR database called LRRML (http://lrrml.lrz.de). It collects three- dimensional LRR structures manually identified from all determined crystal structures of LRR-containing proteins and thus provides a source for the structural modeling and analysis of LRR proteins. With the help of TollML and LRRML, we constructed models of the human/mouse TLR5-13 ectodomains and suggested some potential receptor-ligand interaction residues based on these models.