Massachusetts Institute of Technology1, Cambridge, MA 02139 Argonne National Laboratory2, Lemont, IL 60439
While genomic methods of bacterial analysis are limited to strain identification, proteomic methods detect levels of gene expression, potentially revealing information about the bacterium’s growth conditions. Such characterization is desirable for defense against bioterrorism, as it would enable the determination of the source of a highly infectious agent such as Bacillus anthracis. Protein microarrays can circumvent the need for gel technology by testing hundreds of protein fractions on a small biochip; however, currently there are no universal methods to analyze results from these microarrays. Our lab has written MATLAB executables to aid in automated statistical analysis and visualization of microarray results. In this study, we employed these programs to analyze results from protein microarray studies of
B. anthracis samples grown under varying growth conditions. After obtaining bacterial spore coat protein fractions separated by isoelectric point and hydrophobicity, we printed fractions on microarray slides and tested for protein glycosylation using lectins. We compared samples grown in solid and liquid media, visualizing assay results through heatmaps. Finally, we successfully implemented hierarchical clustering to distinguish between samples grown in different media and at different temperatures. The results bode well for future applications of protein microarray analysis.
[Abstract (DOC)]