Comprehensive proteomics of Methylobacterium extorquens AM1 metabolism under single carbon and nonmethylotrophic conditions

Abstract

In order to validate a gel free quantitative proteomics assay for the model methylotrophic bacterium Methylobacterium extorquens AM1, we examined the M. extorquens AM1 proteome under single carbon (methanol) and multicarbon (succinate) growth, conditions that have been studied for decades and for which extensive corroborative data have been compiled. In total, 4447 proteins from a database containing 7556 putative ORFs from M. extorquens AM1 could be identified with two or more peptide sequences, corresponding to a qualitative proteome coverage of 58%. Statistically significant nonzero (log(2) scale) differential abundance ratios of methanol/succinate could be detected for 317 proteins using summed ion intensity measurements and 585 proteins using spectral counting, at a q-value cut-off of 0.01, a measure of false discovery rate. The results were compared to recent microarray studies performed under equivalent chemostat conditions. The M. extorquens AM1 studies demonstrated the feasibility of scaling up the multidimensional capillary HPLC MS/MS approach to a prokaryotic organism with a proteome more than three times the size of microbes we have investigated previously, while maintaining a high degree of proteome coverage and reliable quantitative abundance ratios.

Figure 1

Project overview, showing (A), the experimental design; (B), an outline of the analytical procedure; and (C), the number of protein relative abundances called as significantly changed by the two calculation methods and the observed overlap of 129 proteins called as changed by both approaches.

Figure 2

Scatter plot and linear regression showing the reproducibility of spectral counts for the biological replicates, i.e. separate chemostat runs, of methanol and succinate. (A), Plot of 2,959 proteins common to both replicates for methanol. (B), plot of 3,115 proteins common to both replicates for succinate.

Figure 3

Scatter plot and linear regression showing the reproducibility of summed signal intensities for the biological replicates of methanol and succinate. (A), Plot of 2,959 proteins common to both replicates for methanol. (B), plot of 3,115 proteins common to both replicates for succinate.

Figure 4

Scatter plots and linear regression for (A), the log2 protein abundance ratios calculated by summed signal intensity versus spectral counting for 762 proteins judged to be significantly changed by either method; (B), the log₂ mRNA abundance ratios [16] versus the log2 spectral count proteomic abundance ratios for 585 ORFs showing significant change in protein abundance; and (C), the log₂ mRNA abundance ratios [16] versus the log₂ summed signal intensity proteomic abundance ratios for 317 ORFs showing significant change in protein abundance.