High Performance Gas Chromatography Mass Spectrometry in Addressing the Challenges of Metabolomic Studies Separation in Time and Mass

Transcription

1 High Performance Gas Chromatography Mass Spectrometry in Addressing the Challenges of Metabolomic Studies Separation in Time and Mass Jeffrey S. Patrick Director of Marketed Technology LECO Separation Science

2 Outline of Presentation The Metabolomics Problem Technologies and Challenges The Zucker Rat Biological Problem and Other Metabolomic Data The Outcome of the Study What is Expressionist? GC-HRMS Technology and Data GCxGC-TOF-HRMS Rats and Breath

3 Objectives GCMS provides capabilities to define modulated analytes in populations and phenotypes which complements LCMS HRMS enables identification of unknowns and confident identification of knowns Accurate m/z for fragments Isotopic Abundance for knowns and unknown Mass accuracy and Isotopic abundance confirm formulae for m/z Chemical Ionization with accurate m/z enables unknown ID Provides linearity and sensitivity needed for metabolomics analysis Deconvolution enables the ability to: detect and quantify metabolites provide searchable spectra from difficult peak pairs provide interpretable spectra from difficult peak pairs GCxGC TOF MS Separation of additional analytes Differential Analysis and enhanced Sensitivity Genedata enables an HRMS-optimized tool for differential analysis of phenotypes and populations.

4 The Metabolomic Problem

5 Analytical Challenges in Metabolomics Accurate Differential Analysis of Biochemistry Comprehensive Analysis (vs. Targeted) > 60% Unknowns! Range of Analyte Concentrations Complex Samples and Analytes Diverse samples (plants, phys fluids, insects, tissue, etc) Matrix Effects Spectral Dynamic Range Isomer Differentiation Reproducibility Peak Capacity (1000s of Analytes/hour) Data Interpretation Systems Biology and Contextual Information

6 Can t cover needed analytes with GC or LC Alone R. Goodacre 13 th Annual Fera/JIFSAN 2012

7 Fast and easy to adopt Variety of MS Opportunities (Nominal, HRMS, MSMS) Universal Application for M < 600 (w/ Deriv) Why GCMS? Linear Response and Good Dynamic Range Over 50 yrs of Application and Largest Presence of ANY MS System Sensitive, Reliable, Robust and Quantitative HUGE Wellestablished Database (>250k spectra) Highest Peak Capacity Chromatography

8 System Under Study Zucker Rats 3 Phenotypes/Strains w/ Animals bred to be Lean (n=12), Fatty (n=12), Obese (n=12) 7-9 Weeks old (terminal bleed) Objectives Disodium EDTA as anti-coagulant 0.1 µm filtered Identify analytes which are up or down regulated with phenotype using high performance MS Fit Test the capabilities of HRTs Husky??

9 Zucker Rat Study: Sample Preparation 1)Plasma (100 µl) 2) MeOH (400 µl) 3) Vortex 4) Centrifuge & Remove Protein Pellet 7) MeONH 2 6) Lyophilize (Overnight) 5) Dry (2 hrs) Pegasus GC-HRT 8) MSTFA 9) FAMEs

10 Why High Resolution and Accurate Mass?

11 Why?? 63% What are these values? 51% Corrupt or Impure Spectra 48% 68% Uncertainty Percentage in m/z of Unidentified EI only with no M Analytes in a Metabolomic Study Limited MS/MS or accurate mass databases Limited Libraries to match Derivatives or Analytes Inadequate Chemical 56% Analysis 71% Tools

12 What is the real value of Accurate Mass? Rachel L. Sleighter and Patrick G. Hatcher (2011). Fourier Transform Mass Spectrometry for the Molecular Level Characterization of Natural Organic Matter: Instrument Capabilities, Applications, and Limitations, Fourier Transforms - Approach to Scientific Principles, Prof. Goran Nikolic (Ed.), ISBN: , InTech, DOI: / Available from:

13 Zucker Rat Study: Workflow Sample Preparation Pegasus GC-HRT (EI and CI-MS) Separation & Detection ChromaTOF HRT Untargeted and/or Targeted Peak Find Metabolite Identification GeneData (Differential Analysis) Alignment, Normalization & Statistics

14 Zucker Rat Study: Instrument Parameters GC Column Carrier Gas, Flow Injection/Volume Temp. Program Agilent 7890 with 7693 Auto Sampler Restek Rxi-5Sil MS (30m x 0.25mm x 0.25mm) & 5m Guard He, 1.0 ml/min Constant Flow Splitless, 1 µl (CI 2 µl) 70 C (4 min) to 300 C at 20 C/min (6 min) MS Transfer Line Temp. Ion Source Temp. Ionization LECO Pegasus GC-HRT 300 C 250 C (CI 200 C) EI (70 ev); CI (140 ev) Mass Range (CI , Reagent Gas = 5% NH 3 in CH 4 ) Acquisition Rate Mass Calibration 6 sps PFTBA (Internal)

15 General Findings 6.0e6 5.0e6 4.0e6 3.0e6 2.0e6 1.0e6 0.0e0 6.0e6 5.0e6 4.0e6 3.0e6 2.0e6 1.0e6 0.0e0 6.0e6 5.0e6 4.0e6 3.0e6 2.0e6 1.0e6 0.0e0 Time (s) AIC Fatty 6 AIC Lean AIC Obese 6 Total Average Features Found (S/N > 100) 662 (+/- 57) Analytes having ID Match > (+/- 34) Analytes at > 600 and M, 2ppm 266 (+/- 26) (N = 36)

16 ChromaTOF HRT: Deconvolution A Methionine (2TMS) 5-Oxo-Proline (2TMS) Methyl Decanoate 5.0e7 B XIC( ±0.0005) XIC( ±0.0005) XIC( ±0.0005) Obese 4.0e7 3.0e7 2.0e7 1.0e7 0.0e0 Time (s) AIC

17 Peak True - sample "Obese 9", L-Proline, 5-oxo-1-(trimethylsilyl)-, trimethylsilyl ester (CAS), at s, Height (Counts) O Si e5 N 4.0e5 O 3.0e5 Si 2.0e5 O 1.0e5 835/ e0 Library Hit - Similarity: Library: Wiley9 - L-Proline, 5-oxo-1-(trimethylsilyl)-, trimethylsilyl ester (CAS), Abundance M/Z EI-HRT: Accuracy & Spectral Similarity ppm ppm Peak True - sample "Obese 9", Dodecanoic acid, methyl ester (CAS), at s, Height (Counts) 5.0e5 4.0e5 3.0e5 2.0e5 1.0e5 0.0e Library Hit - Similarity: Library: Wiley9 - Dodecanoic acid, methyl ester (CAS), Abundance O O 750/ M/Z

18 ChromaTOF HRT: Representative Compounds (0.9 ppm mass error) Name Formula R.T. (s) Area LM (1000) Ion Observed Ion m/z Mass Accuracy (ppm) Alanine (3TMS) C 9 H 23 NO 2 Si [M-CH 3 ] Oxalic Acid (2TMS) C 8 H 18 O 4 Si [M-CH 3 ] Valine (2TMS) C 11 H 27 NO 2 Si [M-C 3 H 7 ] Serine (3TMS) C 12 H 31 NO 3 Si [M-CH 3 ] Threonine (3TMS) C 13 H 33 NO 3 Si [M-CH 3 ] L-Proline, 5-oxo- (2TMS) C 11 H 23 NO 3 Si M [M-CH 3 ] Citric Acid (4TMS) C 18 H 40 O 7 Si [M-CH 3 ] Galactose (MEOX, 5TMS) C 22 H 55 NO 6 Si [M-C 10 H 27 O 2 Si 3 ] Glucose (MEOX, 5TMS) C 22 H 55 NO 6 Si [M-C 11 H 31 O 3 Si 3 ] Inositol (6TMS) C 24 H 60 O 6 Si [M-C 6 H 20 O 2 Si 2 ] Octadecanoic acid (TMS) C 21 H 44 O 2 Si M Arachidonic acid (TMS) C 23 H 40 O 2 Si M Cholestadiene C 27 H M Cholesterol TMS C 30 H 54 OSi M Campesterol, TMS C 31 H 56 OSi M Ave = 0.90 ppm

19 Genedata Expressionist MSX Integrated MS data processing and analysis MS Data Integrated Data Processing Statistical Analysis Software truly optimized to handle GC-HRMS data

20 GeneData Results Fatty Obese Fold Change > 1.5 P values < 0.05 Lean

21 Pathways Implicated in GCMS BCAAs which Feed fatty acids and lipids Statistics and Pathways are important but individual metabolites are important as well

22 Serotonin Obesity, Sleep and Appetite Serotonin Hydroxytryptamine Implicated Pathways and Physiology Tryptophan Biosynthesis Neurotransmission Appetite control Depression Obesity Fatty Lean Obese Total

23 Branched Amino Acids Isoleucine Fatty Lean Obese Total Leucine and Valine show similar trends Branched Chain AA, Fatty Acid and other Metabolism Figures of merit Average m/z = Target m/z Avg. Mass Error = 0.72 ppm (Abs) (N = 36 over 2 days)

24 Modulated Fatty Acid Hexadecanoic Acid Fatty Lean Obese Total Figures of merit Average m/z = Target m/z Avg. Mass Error = 0.97 ppm (Abs) (N = 36 over 2 days)

25 Other Modulated Metabolites Myoinositol (6TMS) m/z m/z Fatty Lean Obese Total Linked to selected lipid metabolism Monitored by each of 2 accurate m/z values one is more consistent/selective than the other

26 Metabolite ID in Tobacco Leaf

27 Cholesterol in Green Tobacco Leaf 1.6e7 1.4e7 1.2e7 1.0e7 8.0e6 6.0e6 4.0e6 2.0e6 0.0e0 Time (s) TIC

28 7.0e6 6.0e6 5.0e6 4.0e6 3.0e6 2.0e6 Time (s) TIC 4.5e5 4.0e5 3.5e5 3.0e5 2.5e5 Tocopherol Cholesterol 2.0e5 1.5e5 1.0e5 5.0e4 0.0e0 Time (s) XIC( ±3ppm) XIC( ±3ppm)

29 dl-à-tocopherol Cholesterol (CAS) Green Leaf : Co-eluting Compounds Area (Abundance) XIC( ) Peak True - sample"gl 9aT Splitless", Cholesterol (CAS), at s M/Z Library Hit - Library: Wiley9 - Cholesterol (CAS) Area (Abundance) Peak True - sample"gl 9aT Splitless", dl-à-tocopherol, at s Area (Abundance) 1.2e3 1.0e3 8.0e2 6.0e2 4.0e2 2.0e2 0.0e0 M/Z * 0.0e0 M/Z Library Hit - Library: mainlib - dl-à-tocopherol Area (Abundance) 1.2e3 1.0e3 8.0e2 6.0e2 4.0e2 2.0e2 1.2e3 1.0e3 8.0e2 6.0e2 4.0e2 2.0e2 0.0e0 M/Z Peak True LM = 833/100 Peak True LM = 880/ HO O HO

30 Interpretative Power of Accurate Mass Cholesterol (in Tobacco Leaf Extract NOT TMS) + H 2 C Area (Abundance), Peak True - sample"gl 9aT Splitless", Cholesterol, at s C H CH 3 H 3 C 0.9 ppm Match = C H 3 CH CH 3 CH 3 H 3 C 1.1 ppm + CH 3 H 3 C M ppm M/Z Structures and fragmentation with ACD/MS Workbook Suite

31 GCxGC TOF MS

32 Sample Preparation Freeze; Pulverize; Extract with MeOH Centrifuge Pool by Condition SpeedVac BSTFA GCxGC TOF - MS Control Ethanol

33 Peak True - sample "POOLED ETOH ", peak 866, at s Library Hit - similarity 811, "Octadecanoic acid, trimeth ylsilyl ester" Octadecanoic acid, trimethylsilyl ester Peak True - sample "POOLED CTRL REF", peak 989, at s Library Hit - similarity 869, "Octadecanoic acid, trimeth ylsilyl ester (CAS)" Octadecanoic Acid in Mouse Liver extracts arrow indicates up regulation DISEASED ETHANOLIC NORMAL CONTROL Time (s) POOLED ETOH 132 POOLED CTRL REF Similarity > 800

34 How can we identify differences between similar samples? Normal Control Pooled (Reference) Disease Ethanolic Pooled (Sample) ChromaTOF s Reference Feature computes the relative concentration of peaks in a sample with respect to a user specified reference. Type tags are assigned to each analyte: Match Out of Tolerance Not found Unknown Present in reference and sample within user-specified concentration tolerance Present in reference and sample, but not within user-specified concentration tolerance Present in reference, but not in sample Present in sample, but not in reference

35 Match: Proline, di-tms Representative Examples Type Concentration Name Similarity R.T. (s) Quant S/N Area Height Quant Masses Match Proline, di-tms , Out of Tolerance 199 Linoleic Acid, TMS , Not Found Ala-Gly, N-TMS-, TMS ester , Unknown Tryptophan, tri-tms , Out of Tolerance: Linoleic Acid, TMS Control Disease Control Disease Not Found: Ala-Gly, N-TMS-, TMS ester Unknown: Tryptophan, tri-tms Control Disease Control Disease

36 Breath Analysis Library ID: Undecane

37 Breath tests using 1D GC MS Demonstrated proof of principle in: Lung cancer Breast cancer Heart transplant rejection Pulmonary tuberculosis Environmental toxicology Influenza

38 All of these studies were performed with 1D GC MS and then 2D GC MS came along

39 Human breath VOCs Two-dimensional gas chromatography and time-of-flight mass spectrometry Breath Breath minus air

40 The limitations of 1D GC MS Poor selectivity ~ 200 VOCs in breath Co-elution +++ Biomarker ID not consistent The advantages of GCxGC TOF MS Excellent selectivity ~ 2,000 VOCs in breath Co-elution minimal Biomarker ID highly consistent

41 Conclusions Metabolomics Studies using GCMS provide capabilities to define modulated analytes in populations and phenotypes which complements LCMS HRMS provides an ability to identify unknowns and to have confident identification of knowns Accurate m/z for fragments Isotopic Abundance for knowns and unknown Mass accuracy and isotopic Abundance confirm formulae for m/z CI enables molecular ion m/z Provides linearity and sensitivity needed for metabolomics analysis Deconvolution enables the ability to: detect and quantify metabolites provide searchable spectra from difficult peak pairs provide interpretable spectra from difficult peak pairs GCxGC TOF MS Separation of additional analytes Differential Analysis and enhanced Sensitivity Genedata enables an HRMS-optimized tool for differential analysis of phenotypes and populations.

42 People Doing the Work

43 For More Information Contact LECO at: World Headquarters/United States In United States: or Outside U.S.A.: