Metabolome, as the downstream of the genome, the entire set of metabolites in a biological system, describes and reflects the cellular activities more vividly and specifically at the functional level. The procedures of sample preparation for metabolomics studies need to assure reliable and reproducible results. These are particularly challenging, especially for microbial samples, which are sensitive to manipulation and limited in quantity to some extent. Therefore, this study aimed to establish reliable and reproducible sample preparation methods for microbial species allowing metabolomics study in bacteria via mass spectrometric analysis. A gram-negative bacterium, C. marina, was chosen for optimizing a protocol from the quenching step to the final derivatization. By applying this, more than 170 intensive signals were detected as a “snapshot” of its intracellular state with acceptable a relative standard deviation (RSD) for most analytes. Based on this, the metabolic response of C. marina could be investigated under the artificial oxidative stress by ozone treatment and famishing stress by nutritional starvation. With the comparison in contour plots and the visualization in GasPedal, intracellular pathways involved metabolites such as fatty acids (oleic acid), amino acids (L-cystine) were found to be the potential biomarkers of bacterial intracellular state varying once encountering extreme conditions.

The second goal was to overcome the limitation of bacteria quantity. To this scope, a sample introduction system combined with an inkjet technique was set up and then optimized for the MALDI-MS spotting method. To reduce the “coffee-ring” effect, the inkjet technique was further optimized involved multi-layer introduction. The feasibility of the method was tested by analyzing E. coli, which demonstrated that the inkjet printing provides spots homogeneity and possibility to analyze very small amount of bacteria. By inkjet printing, significant signals of E. coli metabolites could be obtained with approximately 1000 bacteria. Thus, an alternative introduction method was proposed for bacterial samples on MALDI-MS.