Are you missing key metabolites with Data Dependent Analysis?

Data dependent analysis (DDA) is popular as it greatly simplifies MS/MS experiments. For metabolite identification, DDA targets the most abundant peaks, as well as expected metabolites for further MS/MS analysis. However, if you rely on DDA alone, you might be missing a low-level, unexpected metabolite that is hiding underneath a more abundant peak. Thorough examination of your full-scan LC/MS data is necessary to make sure that you don't miss a key trace metabolite, but this can be difficult and time-consuming, particularly if you performing this task manually, or are dealing with complex sample matrices.

One ACD/Labs software user recently demonstrated how he can quickly extract trace metabolites from complex full-scan LC/MS data sets, even for sample matrices such as bile, hepatocytes, and liver microsomes. He noted that the software was particularly effective for identifying "peaks under peaks"—low level metabolites which might not be picked up by DDA. The software is then used to help interpret the MS/MS spectrum for each metabolite. Assigned metabolite spectra are added to a database for future reference.

Baughman, Todd, 'How ACD/IntelliXtract Aids Discovery Metabolite Identification at GSK-RTP', ACD/Labs ASMS 2007 Seminar, Indianapolis, IN, June 3, 2007.

In one example, hepatocytes were incubated with a drug of interest for 0, 4, and 24 hours. The molecule has low intrinsic clearance, but a glucuronide metabolite had been observed in vivo. Samples were analyzed in negative and positive ionization modes. The negative ion mode LC/MS data was examined first. Two metabolites, much less abundant than the parent (<1%) were observed and were reliably extracted by the software. These metabolites were not found in positive ion mode.

In a second example, major metabolites of a different drug in rat, dog, monkey, and human hepatosomes were incubated. Chromatography was not able to fully separate three metabolites; however, because the chromatographic profiles for each metabolite were unique, the software was able to deconvolute the peaks. The mass spectral information associated with each component was then used to confirm the component and assign the protonated molecule, [M+H]⁺.

A final example illustrates the identification of metabolites in rat bile samples. A pooled sample is compared with a control. The total ion chromatograms (TIC) are complex and, at first glance, don't appear to be significantly different; however, a handful of key metabolites are successfully extracted when ACD/IntelliXtract software is used to compare the two samples.