Search

As a structural analogue of adenine (6-aminopurine), 2-aminopurine (2AP) is a powerful fluorescent probe, when substituted in DNA in place of the natural adenine. Time-resolved fluorescence measurements of 2AP-labeled oligonucleotides, together with steady-state spectroscopy give us an in-depth view of DNA-enzyme interactions, especially the conformational dynamics in solution phase. Herein, this technique has been extended to the study of the transient unzipping of DNA bases, to investigate the structure of three-way junction (3WJ), and the role of base unzipping in the mechanism of human flap endonuclease (FEN). Seven 2AP-labelled 3WJs were investigated, each containing only one 2AP base in place of adenine. In four of the 3WJs, 2AP was placed in the long duplex region of an arm; while in the other three 3WJs, 2AP was placed near or in the branch point. Comparative time-resolved fluorescence measurements on the 3WJs and corresponding ssDNA and dsDNA controls were made to study the base dynamics, in particular the possibility of unzipping in the vicinity of the branch point. In combination with single-molecule FRET measurements and molecular dynamics simulations, the local and global structure of a DNA 3WJ in solution could be unraveled. It was found to adopt a Y-shaped, pyramidal structure, in which the bases adjacent to the branch point are unzipped, despite the full Watson-Crick complementarity of the molecule. Human flap endonuclease (hFEN) is divalent metal ion-dependent phosphodiesterase. hFEN carries out structure-specific hydrolysis of 5’ bifurcated DNA endonucleolytically. Cleavage occurs at a position one nucleotide into the downstream duplex region. Previous structural, biochemical and modeling studies suggested a double-nucleotide unzipping mechanism at single/double strand junctions for scissile phosphate placement. To confirm this mechanism, 2AP time-resolved fluorescence spectroscopy was used to investigate nucleotide unzipping in hFEN substrates. 2AP was substituted at positions +1 and -1 (relative to the scissile phosphodiester) respectively, in double flap substrates. A series of hFEN mutants including Y40A, R100A, K93A, were used in this study. In the experiments, ssDNA, dsDNA substrates, DNA substrate-enzyme complexes were investigated in order to elucidate the enzyme-induced distortion of the substrate at the +1 and -1 positions. TseI is a thermophilic type II restriction enzyme which has ideal activity at an elevated temperature. It is able to recognise and cut the 5 bp palindromic sequence of 5’-GCWGC-3’ (W=A or T). A range of biophysical methods have been applied to investigate this enzyme, including size-exclusion chromatography; fluorescence anisotropy (Kd value determination); denaturing HPLC for DNA cleavage analysis on matched and mismatched substrates; fluorescence-based activity assay (KM, Vmax, kcat, specificity constant values determination); steady-state fluorescence measurements (DNA-enzyme interaction study). The DNA cleavage characteristics of TseI were fully studied and it was found that it cuts A:A and T:T mismatches in CAG and CTG repeats. This potentially makes it a useful tool for exploring unusual DNA structures containing super-long CAG and CTG repeats which are involved in the aetiology of some neurodegenerative diseases, such as Huntington’s disease (HD). EcoP15I is a type III restriction-modification enzyme whose recognition sequence is 5-CAGCAG-3’. Methyltransferase EcoP15I (M.EcoP15I) adds a methyl group to the second adenine, in the presence of cofactor S-adenosyl methionine (SAM). SDS-PAGE, densitometry and size-exclusion HPLC were applied to confirm that EcoP15I adopts a Res1Mod2 stoichiometry in solution. The large structural distortion of its substrate (base flipping) by M.EcoP15I was investigated by both steady-state and time-resolved fluorescence. Also, nine 120 mer DNA duplexes, each containing two reversely oriented recognition sites were used to study matched and mismatched sequence cleavage by R.EcoP15 and a cleavage pattern was revealed.

Proliferative vitreoretinopathy (PVR) is an anomalous cell proliferation condition that occurs after surgery on the retina. 5-Fluorouracil (5-FU) has been shown to be effective in reducing the incidence ofPVR in animal models and recently been used as a clinic adjunctive treatment of PVR. However, the results of such treatment have been disappointing, largely because these drugs are quickly metabolized within the eye minimizing their long-term effectiveness; high doses ofthese drugs may also be toxic to the retina. Agaricus bisporus lectin (ABL) is obtained from t~e edible common mushroom and has been shown to have reversible, dose-dependent and specific anti- proliferation activity in vitro. The aim ofthis research project is to conjugate 5-FU and ABL together by using chemical conjugation methods to give potentially potent and specific anti-proliferative activity for the treatment ofPVR. The first part of this project is to study the protein structure of the ABL, which is to find the suitable residues as conjugation targets. ABL is found as a homotetramer protein and each ofthe four subunits has a molecular mass of 16,053.8 Da. There are one cysteine and five lysines for each of the lectin subunits. The free -SH of cysteine, free -NH2 of lysine and the N-terminal amino groups are conjugation targets. The second part of the project is to modify the 5-FU with either amino or azido functional groups preparing derivatives of 5-fluorouridine which can then be conjugated to ABL. The third part of the work describes the conjugation of the modified 5-FU with ABL by using SMCC (succinirnidyI4-(N-maleimidomethyl) cyclohexane-l-carboxylate) cross linker, aconitic acid anhydrate linker, 'click' chemistry as well as the photochemistry method separately, followed by characterization by SDS-PAGE (sodium dodecyl sulphate-polyacrylamide gel electrophoresis) and the biological test. MALDI-MS (matrix-assisted laser desorption/ionization mass spectrometry) is also attempted for further characterization.

Text classification, the task of metadata to documents, needs a person to take significant time and effort. Since online-generated contents are explosively growing, it becomes a challenge for manually annotating with large scale and unstructured data. Recently, various state-or-art text mining methods have been applied to classification process based on the keywords extraction. However, when using these keywords as features in the classification task, it is common that the number of feature dimensions is large. In addition, how to select keywords from documents as features in the classification task is a big challenge. Especially, when using traditional machine learning algorithms in big data, the computation time is very long. On the other hand, about 80% of real data is unstructured and non-labeled in the real world. The conventional supervised feature selection methods cannot be directly used in selecting entities from massive data. Usually, statistical strategies are utilized to extract features from unlabeled data for classification tasks according to their importance scores. We propose a novel method to extract key features effectively before feeding them into the classification assignment. Another challenge in the text classification is the multi-label problem, the assignment of multiple non-exclusive labels to documents. This problem makes text classification more complicated compared with a single label classification. For the above issues, we develop a framework for extracting data and reducing data dimension to solve the multi-label problem on labeled and unlabeled datasets. In order to reduce data dimension, we develop a hybrid feature selection method that extracts meaningful features according to the importance of each feature. The Word2Vec is applied to represent each document by a feature vector for the document categorization for the big dataset. The unsupervised approach is used to extract features from real online-generated data for text classification. Our unsupervised feature selection method is applied to extract depression symptoms from social media such as Twitter. In the future, these depression symptoms will be used for depression self-screening and diagnosis.