Chromatographic Fingerprinting and Comparative Analysis for Chemical Profiling by Mass Spectrometry

Norm Glassbrook1 and Scott J Campbell2

1NC State University, Raleigh, NC, USA; 2SpectralWorks Ltd, The Heath Business & Technical Park, Runcorn, UK

First Published: ASMS 2005

Abstract

Simple plotting techniques were developed for visualization of complex chromatographic data sets. The techniques were applied to metabolite profiling data from microbial samples analyzed by gas chromatography coupled with mass spectrometry (GC/MS). Over 1000 single ion peaks, corresponding to several hundred chemical components are typically detected in a given sample matrix. Chromatographic ‘fingerprinting’ yielded a quick graphical overview of the chromatographic data that was much easier to review than the total ion chromatogram (TIC). However, due to the complexity of the data and the relatively unrefined nature of the peak data, it was difficult to use the fingerprint plots for sample-to-sample comparisons. Some normalization for sample loading and internal standards, along with filtering to remove peaks that did not meet acceptance criteria for reproducibility and linearity were required for comparative analyses. This data refinement removed as many as half the peaks detected and presented in the fingerprints, but the resulting comparative analysis plots were extremely valuable for highlighting differences in component concentrations between samples. The general plotting technique may be useful for displaying other types of comparisons, such as time course data, or display of compounds unique to specific samples or treatments

Introduction

Chemical profiling techniques typically yield very complex data sets that can be difficult to process and review efficiently. Here we report on the development of a relatively simple technique for presenting a visual overview of profiling data generated by chromatography coupled with mass spectrometry.

These fingerprinting and comparative analysis techniques have been applied to LC/MS and GC/MS analyses of biological samples conducted in support of metabolism studies with the fungus Aspergillus flavus. These studies are intended to better characterize primary and secondary metabolism associated with aflatoxin production.

The GC/MS data presented are the results of preliminary experiments under growth conditions conducive to aflatoxin production (28°C) and those not conducive to production (37°C).

Materials and Methods

Summary

Aspergillus flavus NRRL 3357 samples were lyophilized and treated to produce trimethylsilyl (TMS) derivatives. The supernatant was analyzed by GC/MS.

Chromatographic components were catalogued by retention index (Kovats), and mass spectrum of each component.

Peak areas for various components were plotted versus sample weight to verify linearity over the range of sample loading.

Figure 1. Summary of Materials and Methods

Test Samples and Reference Materials

Aspergillus flavus, finely ground, 3-10mg dw/tube
Fungus cultured in shake flasks at 28°C and 37°C
Alkanes, C10-C40 even, as retention STDs

GC/MS Conditions

Agilent 6890 Gas Chromatograph
Agilent 5973N Mass Selective Detector
SCAN MODE, m/z 50-550
SGE BPX50 Column: 30m X 0.25mm, 0.25µm df
Helium Carrier Gas at ~40cm/sec
Programmed Oven Temperature 50°C to 350°C
Inlet 280°C, 1µL Injection, Splitless

Data analysis

ChemStation, Agilent Technologies
AnalyzerPro, SpectralWorks Ltd
Excel, Microsoft

Results

 

Figure 2. Total Ion Chromatogram of Aspergillus Composite Sample

 

Approximately 90 peaks detected by autointegrate function.

Figure 3. Chromatographic Fingerprint (Contour Plot)

raw peak data3470 single ion peaks detected by AnalyzerPro are saved in a target list as part of data processing method for test samples and exported to Excel for further processing. Peak data is recoded and plotted in several overlays.

Figure 4. Linearity Checks

Sample loading (mg) and abundance (peak area) are used calculating linear regressions. Ions not present in all replicate samples or failing linearity were not used for comparisons. 1890 peaks passed for 28C; 1417 passed for 37C; 916 single ion peaks pass linearity and are common to both test conditions.

Figure 5. Total Ion Chromatograms

Large differences between samples readily observed by overlaying plots.

Figure 6. Comparative Analysis Plot

All original raw peak data from the composite sample are marked in grey. Peaks that pass linearity and are detected in all samples are overlaid in appropriate color.- red for levels higher by comparison, blue for lower. ‘Live’ data points yield retention time and m/z values for quick review (lysine at 8.67 min and m/z 156 shown).

Conclusions

Profiling methods yield very complex data sets that contain high quality analytical data along with data for trace level or overloaded components, background contamination, and spurious peaks. As a result, a large proportion of the profiling data may not be suitable for relative quantitation. However, profiling data can be refined with procedures comparable to those used in classical quantitative methods to yield subsets of data suitable for comparative analyses.
A relatively simple plotting technique can be applied to complex data sets to allow quick visualization of chromatographic data and comparative analyses.

Our Products

AnalyzerPro

AnalyzerPro

AnalyzerPro® is a productivity software application for LC-MS and GC-MS data with support for multiple vendors’ data. This comprehensive post-processing utility provides optimized workflows for sample-to-sample comparison, target component analysis, quantitation and library searching for data generated from any LC-MS and GC-MS platform. Using its proprietary algorithms to detect obscured components that existing software packages are unable to find without .....

AnalyzerPro XD

AnalyzerPro XD

AnalyzerPro® XD is the latest version of our productivity software application now available with support for two dimensional chromatography. Two-dimensional chromatography is a powerful analy­tical tool that has evolved from technology used mainly in the R&D laboratory to a robust commercially available soluti­on from several manufacturers. The technology continues to evolve to address challenges in the analysis of complex samples...

RemoteAnalyzer

RemoteAnalyzer

RemoteAnalyzer is the only open access software solution for today’s fast-paced laboratory. It is designed to operate and optimize the management of multiple locations, multiple types of analyses, multiple scientists of the entire skill set range as well as multiple instrument types from a variety of different vendors. See a case study from the University of Durham regarding their implementation. The walk-up ......

NIST MS Library

NIST MS Library

The NIST/EPA/NIH Mass Spectral Library with Search Program is the standard MS spectra reference database. This library is available with version 2.0g of the full-featured NIST MS Search Program for Windows and includes ....