No abstract

No abstract

To investigate the mechanism for the production of paralytic shellfish toxins (PST) in toxic dinoflagellates, with a 2D-gel based approach, we had made two sets of proteomic comparisons: (a) between a toxic Alexandrium catenella (AC-T) and a phylogenetically closely related non-toxic strain (AC-N), (b) between toxic AC-T grown in a medium with 10% normal amount of phosphate (AC-T-10%P) known to induce higher toxicity and AC-T grown in normal medium. We found that photosynthesis and energy production related proteins were up-regulated in AC-T when compared to AC-N. However, the same group of proteins was down-regulated in AC-T-10%P when compared to normal AC-T. Examining the relationship of photosynthesis and toxin content of AC-T upon continuous photoperiod experiment revealed that while growth and associated toxin content increased after 8 days of continuous light, toxin content maintained constant when cells were shifted from continuous light to continuous dark for 3 days. This emphasized the cruciality of light availability on toxin biosynthesis in AC-T, while another light-independent mechanism may be responsible for higher toxicity in AC-T-10%P compared to normal AC-T. Taken all together, it is believed that the interplay between “illumination”, “photosynthesis”, “phosphate availability”, and “toxin production” is much more complicated than what we had previously anticipated.

No abstract

Protein sequence identification by tandem mass spectroscopy (LC-MS/MS) identifies thousands of protein sequences even in complex mixtures, and provides valuable insight into the biological functions of different cells. For non-model organisms, transcriptomes are generally used to allow peptide identification, an important addition to their use as a gene catalog allowing the potential metabolic activities of cells to be determined. We used LC-MS/MS data to identify which of the six possible reading frames in the transcriptome was actually used by the cell to make protein, and asked whether this would have an impact on downstream analyses using the dataset. We combined results from several LC-MS/MS experiments designed to identify peptide sequences in extracts from the dinoflagellate Lingulodinium polyedra using a 74 655-sequence transcriptome. We compiled a list of 6628 translated nucleic acid sequences that contained the ensemble of peptide matches (termed MS-validated sequences) and assessed the similarity in downstream analyses between this data set and the 6628 nucleic acid sequences from which they were derived. When compared with BLASTx analyses of the DNA sequences, the MS-validated protein-sequences-analyzed using BLASTp showed differences in gene ontology, had more identified BLAST hits, and contained more KEGG pathway enzymes. The MS-validated protein sequences also differ from datasets containing longest open reading frame (ORF) protein sequences. We also note a poor correlation between the levels of protein and mRNA abundance, a comparison not previously performed for dinoflagellates. The differences observed between analyses of MS-validated protein sequence and nucleic acid sequence datasets suggest use of the former may provide a more accurate representation of cellular capacity than the latter. Developing MS-validated protein sequence datasets may also speed interpretation of MS-MS spectra in bottom up proteomics experiments.