Biosynthesis of Yersinia enterocolitica serotype O:3 lipopolysaccharide outer core

Lipopolysacharide (LPS) present on the outer leaflet of Gram-negative bacteria is important for the adaptation of the bacteria to the environment. Structurally, LPS can be divided into three parts: lipid A, core and O-polysaccharide (OPS). OPS is the outermost and also the most diverse moiety. When OPS is composed of identical sugar residues it is called homopolymeric and when it is composed of repeating units of oligosaccharides it is called heteropolymeric. Bacteria synthesize LPS at the inner membrane via two separate pathways, Lipid A-core via one and OPS via the other. These are ligated together in the periplasmic space and the completed LPS molecule is translocated to the surface of the bacteria. The genes directing the OPS biosynthesis are often clustered and the clusters directing the biosynthesis of heteropolymeric OPS often contain genes for i) the biosynthesis of required NDP-sugar precursors, ii) glycosyltransferases needed to build up the repeating unit, iii) translocation of the completed O-unit to the periplasmic side of the inner membrane (flippase) and iv) polymerization of the repeating units to complete OPS. The aim of this thesis was to characterize the biosynthesis of the outer core (OC) of Yersinia enterocolitica serotype O:3 (YeO3). Y. enterocolitica is a member of the Gram-negative Yersinia genus and it causes diarrhea followed sometimes by reactive arthritis. The chemical structure of the OC and the nucleotide sequence of the gene cluster directing its biosynthesis were already known; however, no experimental evidence had been provided for the predicted functions of the gene products. The hypothesis was that the OC biosynthesis would follow the pathway described for heteropolymeric OPS, i.e. a Wzy-dependent pathway. In this work the biochemical activities of two enzymes involved in the NDP-sugar biosynthesis was established. Gne was determined to be a UDP-N-acetylglucosamine-4-epimerase catalyzing the conversion of UDP-GlcNAc to UDP-GalNAc and WbcP was shown to be a UDP-GlcNAc- 4,6-dehydratase catalyzing the reaction that converts UDP-GlcNAc to a rare UDP-2-acetamido- 2,6-dideoxy-d-xylo-hex-4-ulopyranose (UDP-Sugp). In this work, the linkage specificities and the order in which the different glycosyltransferases build up the OC onto the lipid carrier were also investigated. In addition, by using a site-directed mutagenesis approach the catalytically important amino acids of Gne and two of the characterized glycosyltranferases were identified. Also evidence to show the enzymes involved in the ligations of OC and OPS to the lipid A inner core was provided. The importance of the OC to the physiology of Y. enterocolitica O:3 was defined by determining the minimum requirements for the OC to be recognized by a bacteriophage, bacteriocin and monoclonal antibody. The biological importance of the rare keto sugar (Sugp) was also shown. As a conclusion this work provides an extensive overview of the biosynthesis of YeO3 OC as it provides a substantial amount of information of the stepwise and coordinated synthesis of the Ye O:3 OC hexasaccharide and detailed information of its properties as a receptor.

Biosynthesis of pyranonaphthoquinone polyketides:characterization of the alnumycin pathway

Alnumycin A is an aromatic pyranonaphthoquinone (PNQ) polyketide closely related to the model compound actinorhodin. While some PNQ polyketides are glycosylated, alnumycin A contains a unique sugar-like dioxane moiety. This unusual structural feature made alnumycin A an interesting research target, since no information was available about its biosynthesis. Thus, the main objective of the thesis work became to identify the steps and the enzymes responsible for the biosynthesis of the dioxane moiety. Cloning, sequencing and heterologous expression of the complete alnumycin gene cluster from Streptomyces sp. CM020 enabled the inactivation of several alnumycin biosynthetic genes and preliminary identification of the gene products responsible for pyran ring formation, quinone formation and dioxane biosynthesis. The individual deletions of the genes resulted in the production of several novel metabolites, which in many cases turned out to be pathway intermediates and could be used for stepwise enzymatic reconstruction of the complete dioxane biosynthetic pathway in vitro. Furthermore, the in vitro reactions with purified alnumycin biosynthetic enzymes resulted in the production of other novel compounds, both pathway intermediates and side products. Identification and molecular level studies of the enzymes AlnA and AlnB catalyzing the first step of dioxane biosynthesis – an unusual C-ribosylation step – led to a mechanistic proposal for the C-ribosylation of the polyketide aglycone. The next step on the dioxane biosynthetic pathway was found to be the oxidative conversion of the attached ribose into a highly unusual dioxolane unit by Aln6 belonging to an uncharacterized protein family, which unexpectedly occurred without any apparent cofactors. Finally, the last step of the pathway was found to be catalyzed by the NADPH-dependent reductase Aln4, which is able to catalyze the conversion of the formed dioxolane into a dioxane moiety. The work presented here and the knowledge gained of the enzymes involved in dioxane biosynthesis enables their use in the rational design of novel compounds containing C–C bound ribose, dioxolane and dioxane moieties.

Optimization and Applications of Large-Scale Biomedical Event Networks

The overwhelming amount and unprecedented speed of publication in the biomedical domain make it difficult for life science researchers to acquire and maintain a broad view of the field and gather all information that would be relevant for their research. As a response to this problem, the BioNLP (Biomedical Natural Language Processing) community of researches has emerged and strives to assist life science researchers by developing modern natural language processing (NLP), information extraction (IE) and information retrieval (IR) methods that can be applied at large-scale, to scan the whole publicly available biomedical literature and extract and aggregate the information found within, while automatically normalizing the variability of natural language statements. Among different tasks, biomedical event extraction has received much attention within BioNLP community recently. Biomedical event extraction constitutes the identification of biological processes and interactions described in biomedical literature, and their representation as a set of recursive event structures. The 2009–2013 series of BioNLP Shared Tasks on Event Extraction have given raise to a number of event extraction systems, several of which have been applied at a large scale (the full set of PubMed abstracts and PubMed Central Open Access full text articles), leading to creation of massive biomedical event databases, each of which containing millions of events. Sinece top-ranking event extraction systems are based on machine-learning approach and are trained on the narrow-domain, carefully selected Shared Task training data, their performance drops when being faced with the topically highly varied PubMed and PubMed Central documents. Specifically, false-positive predictions by these systems lead to generation of incorrect biomolecular events which are spotted by the end-users. This thesis proposes a novel post-processing approach, utilizing a combination of supervised and unsupervised learning techniques, that can automatically identify and filter out a considerable proportion of incorrect events from large-scale event databases, thus increasing the general credibility of those databases. The second part of this thesis is dedicated to a system we developed for hypothesis generation from large-scale event databases, which is able to discover novel biomolecular interactions among genes/gene-products. We cast the hypothesis generation problem as a supervised network topology prediction, i.e predicting new edges in the network, as well as types and directions for these edges, utilizing a set of features that can be extracted from large biomedical event networks. Routine machine learning evaluation results, as well as manual evaluation results suggest that the problem is indeed learnable. This work won the Best Paper Award in The 5th International Symposium on Languages in Biology and Medicine (LBM 2013).