Proteomic approach to characterize differential protein expression in toxic and harmful algal bloom species (HABs)

Abstract

This thesis reports on my work using the proteomic approach to study causative agents of HABs. Four sets of studies were conducted. Firstly, a comprehensive study to find the optimal sample preparation conditions for 2DE analysis of Prorocentrum triestinum, a model causative agent of harmful algal blooms (HABs) was carried out. Our results showed that sonication is easy to perform and resulted in a higher protein yield. Taken overall, a combination of sequential extraction and desalting by Biospin chromatography for sample treatment before the 1st dimension of 2DE gave the best results based on its simplicity and minimal protein loss. Subsequently, the sample preparation procedures established for P. triestinum were adapted to cover both thecate and athecate dinoflagellates. Optimized technical procedures were developed and used to generate proteome reference maps for eight other local causative species of harmful algal blooms (HABs): P. micans, P. minimum, P. sigmoides, P. dentatum, Scrippsiella trochoidea, Karenia longicanalis, K. digitata and K. mildmotoi; together with one American species K. brevis (Florida, USA). These proteome maps were used for species recognition. Species-specific 2-DE protein profiles were observed in all ten species and it was possible to distinguish between even closely related species within the same family. Further, to demonstrate the extent of reproducibility and usefulness of these 2-DE- reference maps, 2-DE has been used to analyse three geographically distinct isolates of P. dentatum, and to distinguish species composition in a mixed culture. Application of 2D-PAGE analysis to differentiate between taxonomically confused strains of a single species could be a powerful taxonomic tool. Thirdly, simultaneous comparison of differentially expressed protein profiles of P. triestinum grown under different growth phases and growth conditions were performed aiming to find phase-specific and stress-responsive proteins respectively. Correlation studies on these proteins in relation to cell division phasing patterns and to models of phytoplankton growth inferred possible functions of these proteins. Most notable among these proteins were proteins thought to trigger or mediate the cells through specific phases of division of this alga, e.g. BP1, BP2, PB1, PB2 and PB3. Other proteins (e.g. Group 1 proteins) thought to be responsible for maintaining and supporting cell concentration under adverse conditions were found. Furthermore, another groups of proteins (Group 2 proteins) thought to be stress-responsive was also detected. Taken overall, these differentially expressed proteins provided important information for uncovering unique, or widely conserved, protective and adaptive mechanisms in the dinoflagellate's life cycle and monitoring the presence of these phase-specific proteins could be an important biomarker for bloom prediction. Fourthly, differential protein expression profiles were compared between different clonal culture strains of Alexandrium minutum isolated from two separate sites in Taiwan. HPLC analysis demonstrated that the toxin components of toxic strains of A. minutum were composed mainly of gonyautoxins (GTX1-4) and their proportions varied largely between different clones. Interestingly, some culture strains from these proximate sites were found to be non-toxic. Variation in morphological features between clones was minimal and not significant. Also, variation in differential protein expression within either toxic or non-toxic strains was low, but pronounced differences were detected between toxic and non-toxic strains, the most notable of which were several abundant proteins with PIs ranging from 4.8 to 5.3 and apparent molecular masses between 17.5 and 21.5 kDa. NTl, NT2, NT3 and NT4 were consistently found in all non-toxic strains while Tl and T2 were prominent in the toxic strains. The expressions of these strain-specific proteins have been further investigated under different growth phases and under different environmental stresses. Our results showed that these "fingerprints" were a stable property and can be regarded as "taxonomic markers" to distinguish toxic and non-toxic strains within A. minutum. The expression pattern of Tl in relation to different phases of the growth cycle and physiological conditions led to the hypothesis that Tl might be directly related to toxin biosynthesis in this test alga and it can be regarded as a "toxin biomarker" to study the toxin biosynthetic mechanism in toxic dinoflagellate cells. The method described here may serve as a model for developing similar tools in other toxic species of phytoplankton. We believe this is the first proteomic approach to study the blooming mechanism and toxin biosynthetic pathway of causative agents of HAB and it could be an important new tool for studying the physiological ecology of harmful and toxic dinoflagellates in their natural environment. The approach here, of identifying toxin-related proteins or "indicator proteins" which can be used to rapidly assess the nutritional or metabolic status of these phytoplankton cells using 2-DE, followed by physiological verification, may also serve as a blueprint for similar work with other toxic species in the future.