Use of RapidFire/MS for Metabolomics. - from untargeted to targeted and Pathway driven analysis. RapidFire/MS. Moritz Wagner Agilent Technologies

Transcription

1 Use of RapidFire/MS for Metabolomics RapidFire/MS - from untargeted to targeted and Pathway driven analysis Moritz Wagner Agilent Technologies 1

2 Agilent RapidFire TM High-throughput Mass Spectrometry System Fast autosampler, SPE and Sample Quantitation Complementary to LC-MS Integrated, automated, micro-scale solid-phase extraction Integrates with standard ESI MS instruments Compatible with most biological matrices cycle time: 5 10 seconds / sample 2

3 RapidFire 365 plate handling Capacity 63 plates samples over 60 hours 3 September 17, 2013

4 RapidFire 365 cartridge changer Capacity 12 cartridges matching the Benchbot 4

5 RapidFire cartridges Up to 12 re-usable SPE cartridges Same type or mix injections per cartridge 5 September 17, 2013

6 RapidFire 365 solvent delivery 3 pumps with 4 channels each up to 12 different solvents 6 September 17, 2013

7 Bottlenecks in a typical core facility/academia lab? Accelerate all steps from method development to data processing Faster method development: Use different chemistries Swap solvents Browse sample prep conditions Run more projects Run larger cohorts Efficient use of mass spec MassHunter/ MPP 7

8 Where to apply RapidFire in an Omics lab? Fast method development/ early discovery studies Software tools Candidate biomarker discovery + validation Fast validation 8

9 Finding the sweet spot(s) of your samples Optimization - an act, process, or methodology of making something (as a design, system, or decision) as fully perfect, functional, or effective as possible Simple View 1 parameter Real View many parameters 9

10 Screening of methods and parameters Polar, ESI- Polar, ESI+ Apolar, ESI+ Apolar, ESI- Typical landscape of heterogenously distributed metabolites Ideally, each compound or at least compound class should be analyzed using a dedicated method. Finding the best parameter set can be cumbersome 10

11 Untargeted profiling Polar, ESI- Polar, ESI+ Apolar, ESI+ Apolar, ESI- Case study: 58 kidney stone formers vs. 38 healthy individuals Analyzed using UHPLC/QTOF-MS C18 and ESI+ only 20 hours run time 11

12 Untargeted profiling C4, ESI+ Graphitized carbon, ESI+ HILIC, ESI+ Case study: 58 kidney stone formers vs. 38 healthy individuals Analyzed using 3x using different chemistries 1 hour run time, 60-fold improvement vs LCMS 12

13 Untargeted profiling Total ion chromatograms of 96 injections on C4 cartridge 13

14 Untargeted profiling Individual samples 12 seconds/sample 19 minutes / well plate 14

15 Untargeted profiling Individual spectra for each injection 15

16 Untargeted profiling Superimposed spectra for all 96 injections Mass accuracy/reproducibility is crucial 16

17 Untargeted profiling 1. Convert/split data 2a. Create data analysis method 3. Statistical analysis 2b. Run worklist/sequence for data processing 17

18 Untargeted profiling Step 3 Statistics For human urine matrix and compounds ionizing in ESI+, graphitized carbon was best suited Hundreds and thousands of data points can be quickly browsed and evaluated 18

19 Untargeted profiling Step 3 Statistics Control Stone formers 2nd pass PCA showing only graphitized carbon data 19

20 Untargeted profiling Step 3 Statistics Extracted ion chromatogram for one statistically marked compound Stone formers Controls Validation by isotope pattern, MS/MS and retention time for isomers needs LC/MS 20

21 Untargeted profiling Mouse plasma extracts 3 different cell lines (n = 3 each, visualized by shape) DB-S + H 2 S donor C57-S DB-unt2 C57-N DB-N no H 2 S donor 21

22 Untargeted profiling Normal Cancer Adenoma PCA plot of fecal matter extracts 3 groups (n = 10 each) 22

23 Targeted profiling and quantitation ESI + C4 PGC 0 nm 10 nm 100 nm 1 µm 10 µm Study design: 5 injections at different concentrations: Blank, 10 nm, 100 nm, 1 µm and 10 µm ESI positive and negative ESI - HILIC C4 PGC HILIC 3 different cartridge chemistries: C4, graphitized carbon (GC) and HILIC 71 different metabolites Injection count: 5 (concentrations) x 2 (polarities) x 3 (cartridges) = (injections) x 71 (compounds) = 2130 data points Acquisition time for 2130 data points: 7 minutes 1 71 Each column is a different metabolite 23

24 / Glyoxylate / Pyruvate / L-Lactate / L-Serine / L-Proline / Fumarate / 3-Methyl-2-oxobutanoate / L-Valine / Succinate / L-Threonine / Nicotinate / 4-Methyl-2-oxopentanoate / 5-Amino-4-oxopentanoate / 5-Amino-4-oxopentanoate / L-Leucine / Ornithine / Ornithine / Ornithine / L-Aspartate / L-Malate / Adenine / L-Histidinol / 2-Oxoglutarate / L-Lysine / L-Glutamate / L-Methionine / D-Ribose / Guanine / L-Histidine / Phenylpyruvate / Phenylpyruvate / L-Phenylalanine / Dihydroxyacetone phosphate / Glycerol 3-phosphate / cis-aconitate / cis-aconitate / Shikimate / L-Arginine / D-Glucose / L-Tyrosine / L-Tyrosine / 3-Phospho-D-glycerate / 3-Phospho-D-glycerate / N2-Acetylysine / Citrate / D-Glucuronate / D-Erythrose 4-phosphate / D-Erythrose 4-phosphate / L-Tryptophan / L-Tryptophan / 2-Methylcitrate / Pantothenate / Pantothenate / N-Acetyl-D-glucosamine / D-Xylulose 5-phosphate / D-Glucosamine 6-phosphate / D-Glucosamine 6-phosphate / D-Glucose 6-phosphate / 6-Phospho-D-gluconate / Sedoheptulose 7-phosphate / camp / D-Fructose 1,6-bisphosphate / Maltose / 2',3'-Cyclic GMP / AMP / GMP / ADP / GDP / ATP HILIC GC C4 Targeted profiling and quantitation MM C4 / 0 MM C4 / 10 MM C4 / 100 MM C4 / 1000 MM C4 / MM GC / 0 MM GC / 10 MM GC / 100 MM GC / 1000 MM GC / MM HILIC / 0 MM HILIC / 10 MM HILIC / 100 MM HILIC / 1000 MM HILIC / ESI negative data, color codes for intensity (blue: low brown: high) 24

25 / D-Glucuronate / D-Erythrose 4-phosphate / D-Erythrose 4-phosphate / L-Tryptophan / L-Tryptophan / 2-Methylcitrate / Pantothenate / Pantothenate / N-Acetyl-D-glucosamine / D-Xylulose 5-phosphate / D-Glucosamine 6-phosphate / D-Glucosamine 6-phosphate / D-Glucose 6-phosphate / 6-Phospho-D-gluconate / Sedoheptulose 7-phosphate Targeted profiling and quantitation / Glyoxylate / Pyruvate / L-Lactate / L-Serine / L-Proline / Fumarate / 3-Methyl-2-oxobutanoate / camp / D-Fructose 1,6-bisphosphate / L-Valine / Succinate / Maltose / 2',3'-Cyclic GMP / L-Threonine / Nicotinate / 4-Methyl-2-oxopentanoate / 5-Amino-4-oxopentanoate / 5-Amino-4-oxopentanoate / L-Leucine / Ornithine / Ornithine / Ornithine / L-Aspartate / L-Malate / Adenine / L-Histidinol / 2-Oxoglutarate / L-Lysine / L-Glutamate / L-Methionine / D-Ribose / Guanine / L-Histidine / Phenylpyruvate / Phenylpyruvate / L-Phenylalanine / Dihydroxyacetone phosphate / Glycerol 3-phosphate / cis-aconitate / cis-aconitate / Shikimate / L-Arginine / D-Glucose / L-Tyrosine / L-Tyrosine / 3-Phospho-D-glycerate / 3-Phospho-D-glycerate / N2-Acetylysine / Citrate / AMP / GMP / ADP / GDP / ATP / D-Glucuronate / D-Erythrose 4-phosphate / D-Erythrose 4-phosphate / L-Tryptophan MM C4 / 0 MM C4 / 10 MM C4 / 100 MM C4 / 1000 MM C4 / MM GC / 0 MM GC / 10 MM GC / 100 MM GC / 1000 MM GC / MM HILIC / 0 MM HILIC / 10 MM HILIC / 100 MM HILIC / 1000 MM HILIC / Low intensities for all compounds on C4 Good intensities and concentration based correlation for nucleotides using GC Good intensities and concentration based correlation for sugar phosphates using HILIC Details visualized on a compound class base 25

26 Targeted profiling and quantitation AMP Calibration for nucleotides using GC 10 nm 10 µm (AMP,GMP,ADP,ATP) 100 nm 10 µm (camp,cgmp,gdp) GMP All R 2 > 0.999, quadratic fit ATP camp ADP cgmp GDP 26

27 Targeted profiling and quantitation Enlarged hierarchical cluster analysis of all metabolites in E.coli extract using METLIN database search Combine targeted and untargeted data for Pathway based analysis HILIC 5 HILIC 4 HILIC 3 HILIC 2 HILIC 1 HILIC 0 GC 5 GC 4 GC 3 GC 2 GC 1 GC 0 C4 5 C4 4 C4 3 C4 2 C4 1 C4 0 27

28 Pathway driven analysis Focussing on small section, e.g. One-carbon metabolism Human pathway map from: 28

29 Pathway driven analysis Tetrahydrofolate 29

30 Pathway driven analysis Folates Superimposed ion chromatograms of THF, MTHF, CHO-THF and CH2-THF Triplicate injections 10 µm 2 µm Blank 16 nm 80 nm 400 nm 30

31 Conclusions RF/MS is very well suited for all Metabolomics stages Time-to-answer is greatly reduced compared to LCMS, injection times are typically between seconds per sample Semi-automated workflow with existing and proven tools including MPP workflow 31

32 Contributions Thank you to: Rebecca Konietzny, CCMP Oxford Nicola Zamboni, ETH Zürich Qiuying Chen and Steven Gross, Weill Cornell Medical College, NY 32