Bioinformatic Studies Of The Ubiquitin Conjugation System

Ubiquitin and ubiquitin-like (Ub/UBL) conjugation is one of the most important post-translational modifications responsible for regulating a variety of cellular processes. In eukaryotes, ubiquitination and other ubiquitin-like modifications share a similar E1(ubiquitin-activating enzyme)-E2(ubiquitin-conjugating enzyme)-E3(ubiquitin-protein ligase) enzyme cascade. Although extensive experimental efforts have been taken, an comprehensive classification and data resource of ubiquitin and ubiquitin-like conjugation enzymes is still not available. From all the ubiquitin-related scientific literature,26Els,105E2s,1003E3s and148DUBs were collected and classified into1,3,19and7families, respectively. To computationally characterize potential enzymes in70eukaryotic species, we constructed1,1,15and6hidden Markov model (HMM) profiles for Els, E2s, E3s and DUBs at the family level, separately. Moreover, the ortholog searches by reciprocal BLAST were conducted for E3and DUB families without HMM profiles. All experimentally identified enzymes were taken as the benchmark dataset to evaluate the prediction performance and robustness of the HMM identifications. At last, a comprehensive database UUCD (Ubiquitin and Ubiquitin-like Conjugation Database) was developed with738Els,2937E2s,46631E3s and6647DUBs of70eukaryotic species. Then we used the identified Ub/UBL enzymes and other ubiquitin-related resoures to computationally analyse the Ub/UBL conjugations in complex diseases. We mapped all human Ub/UBL conjugation enzymes to databases of Cancer Gene Census and Drugbank, and identified only27cancer genes and16drug targets. The statistical analyses with a hypergeometric distribution demonstrated that both known cancer genes and drug targets were not significantly enriched in Ub/UBL conjugation enzymes (p-value>0.05). However, the results of KEGG analysis suggested ubiquitination enzymes plays an important role in SCLC (Small Cell Lung Cancer)-related PI3K-Akt signaling, cell cycle, apoptosis and p53signaling pathways.Based on the classification of E3s in UUCD database, we developed a tool for predicting E3-specific ubiquitination sites. E3confers ubiquitination the major specificty for substrate recognition. Identification of protein ubiquitination sites with their cognate E3s has been critical for understanding the complete ubiquitination and potential relationships between ubiquitinaton and other important cellular processes. To data, at least10prediction programs for ubiquitination sites have been developed, but none of them were able to predict ubiquitin ligases for substrates’ ubiquitination sites. Also, we collected1027ubiquitination sites with E3information from all the ubiquitin-related scientific literature, and integrated these E3specific sites into1214E3-site pairs. All E3-site pairs were classfied into9datasets in E3family level and92datasets in E3single level. For prediction of E3-specific ubiquitination sites, we adopted three sequential steps of Motif length selection (MLS), weight training (WT) and matrix mutation (MaM) in the GPS2.2(Group-based Prediction System, version2.2) algorithm to train all the datasets and maximize the improvement of the prediciton performances, and developed a useful tool GPS-PLUB1.0(GPS for Prediction of ubiquitin Ligase-based Ubiquitination sites, Version1.0) for predicting E3-specific Ub-sites for92E3s in hierarchy. Based on the tool, we performanced large scale predictions for all at least one hundred thousand experimentally verified Ub-sites.Analogous to the specificity between E3and ubiquitination site, there is also a specific relationship between the ubiquitin conjugation type and ubiquitination site. Ubiquitin conjugation types include monoubiquitination or different polyubiquitin chain linkages (K6, Kll, K27, K29, K33, K48and K63linked), which could determine different fates of substrates by degradation or various nondegradative functions. To date, there is no tool-that could computationally analyze the specificity between ubiquitin conjugation type and ubiquitination site.We collected at least1000ubiquitination sites with information of ubiquitin conjugation types from the ubiquitin-related scientific literature. Then all sites were classified two groups, two subgroups and seven families. Based on this classification, we integrated11datasets, including mono, poly, poly_proteasome, polynonojproteasome, polyK6, polyKll, polyK27, polyK29, polyK33, polyK48and polyK63. Then GPS3.0algorithm were adopted to to train all the datasets. Finally, we developed a tool GPS-PUCT (GPS for Prediction of Ubquitin Conjugation Types).Taken together, bioinformatics could play an important role in study of ubiquitin and ubiquitin-like conjugation. Using bioinformatics methods, we built a comprehensive database for ubiquitin-related enzymes and a novel tool for prediction E3-specific ubiquitination sites, and furthermore made a computational analysis of Ub/UBL conjugations in complex diseases.