02/07/2013

1

Mass Spectrometry analysis of Small molecules

Metabolomics- A realm of small molecules (<1000 Da)

Genome Transcriptome Proteome Metabolome Phenotype

What might be happening in a cellSnapshot of the entirephysiology

Amino acidsFatty acidsPhenolics

ProstaglandinsSteroids

Organic acidsOrganic amines

NucleosidesNucleotidesPolyaminesLipids etc.

• Profiling involves finding of all metabolites detectable to a selected analytical technique with statistically significant variations in abundance within a set of experimental and control groups.

•Identification of chemical structures of metabolites of interest after profiling

•Quantification and validation

•Interpretation of data making connections between the metabolites discovered and the biological conditions

Steps involved in metabolomic analysisPossible metabolites

Metabolomics in the context of other omics

Metabolomics is complementary to the other omics and the combinationof these three may provide important information about the status of a cell

02/07/2013

2

Application of metabolomics

• Nutrition sciences- eg. oil seed analysis/polyphenols/food adulteration/quality control

• Herbal drug evaluation, drug discovery

• Biomarker identification- eg. Cancer

• Toxicology assessment/functional genomics

Metabolomics

• Targeted (quantitative)- measurement of defined groups of chemically characterized and biochemically annotated metabolites using optimized assay

• Untargeted- comprehensive analysis of all the measurable analytes in a sample, including chemical unknowns.

02/07/2013

3

Targeted/untargeted metabolomics• Targeted metabolomics- extraction procedures can be

optimized for compounds of interest

• Optimized MRM or SRM can be used for quantitation when standards are available.

• Provide comprehensive understanding of a vast array of metabolic enzymes, their kinetics, and biochemical pathways.

• Untargeted metabolomics- single extraction method may not able to extract all compounds and important compounds may be missed during extraction. Data mining can be a problem and requires the use of metabolomic software for identification.

• Offers opportunities for discoveries of novel drug, and biomarkers.

Platform to process untargeted metabolomic data

• XCMS (developed by the Siuzdak Lab at the Scripps Research Institute) Online, is a web-based version that allows users to easily upload and process LC-MS data. It is a bioinformatics platform to identify endogenous metabolites..

• METLIN (developed by the Siuzdak Lab.) is a metabolite database for metabolomics containing over 64,000 structures and it also has comprehensive tandem mass spectrometry data on over 10,000 molecules.

02/07/2013

4

Workflow for metabolome analysis

Sample collection

Treatment orDiseased group

Control group

Urine, plasma, tissueetc.

Sample preparation Internal standard spiking for quantitative analysis Extraction (liquid-liquid, ppt with or without hydrolysis)

Sample analysisLC-MS and LC-MS/MS analysis• Q1 scan using HPLC or UPLC/-TOFMS• +ve/-ve ion mode- accurate mass measurement• MS/MS experiments using a hybrid instrument Q-Trap

Data analysisMultivariat analysis e.g. PCA

Data export

Marker identification

Points to be considered in LC-MS analysis

• Choice of ionization mode- ESI Vs APCI +ve/-ve modes

• Choice of eluting solvent- methanol Vs acetonitrile

• Additives/pH in mobile phase

• Molecular ion recognition (adduct formation)

• Chromatographic separation- stationary phase C8, C18 ..

• Evaluation of spectral quality- what to look for in a good quality spectra

02/07/2013

5

Sample preparation

Sample collection

Quenching by liquid Nitrogen or cold methanol(stops metabolism)

Extraction of metabolites(methanol, methanol-water for polar)(chloroform or hexane for less polar) (protein precipitation, supercritical fluid extraction)

Concentration(evaporation under vacuum, lyophilization, SPE)

Prepare internal standard stock solution

MS acquisition strategy

Full scan (Q1 scanning) for total profiling of metabolites(+ve and –ve ion mode) ESI/APCI

ESI- Effluent is charged and nebulized, for semi- polar or polar compounds e.g. Conjugated metabolites.

APCI- Effluent is heated but not charged- a corona discharge is needed.Good for neutral or less polar compound.

ESI is the most common ionization method

Advantage: non-selective and most ionizable ions are detectedDisadvantage: low sensitivity and detection of minor metabolites is compromised.

02/07/2013

6

MS analysis

• Direct MS analysis

- Without chromatographic separations- HTS possible.

- High resolution MS – FTICR-MS high resolution >1,000,000 and mass accuracy (<1 ppm)

• Problems- difficult to interpret the data

• LC-MS and MS/MS

• - GC-MS (volatile metabolites) and LC-MS- normal phase, reverse phase (C8/C18) and HILIC. UPLC-QTOF-MS for highly complex plant metabolomics.

Quantification

• Relative or absolute quantification.

• Relative- normalizes the metabolite signal that of an internal standard signal intensity in large scale un-targated profiling (eg. Non-naturally occurring lipid standards- Cer 17, stable isotope labeling through metabolism- AA-d8.

• Absolute quantification- based on external standards or internal isotopically labeled standards- targeted metabolomics.

• Matrix effects- signal suppression or enhancement are major issues. Stable isotope labeled standards are needed.

02/07/2013

7

Increasing metabolite coverage using +ve and –ve ion mode

Source: Nordstrom et al. Analytical Chemistry, 2007

No arachidonicAcid in +ve ion mode

Representative Q1 scans of a methanolic extract of human blood serum

Time, min

1 3 5 7 90.01.0e8

3.0e8

5.0e8

7.0e8

9.0e8

1.1e9

1.3e9

1.5e9

Intensity, cps

7.948.65

7.618.41

2.03 9.66

7.23

6.485.59

2.685.271.52

4.33 4.56

0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.50.0

2.0e8

4.0e8

6.0e8

8.0e8

1.0e9

1.2e9

1.4e9

1.6e9

1.8e9

Inte

nsity, cp

s

8.65

9.158.022.02 7.35

1.56 6.365.882.67

5.114.59

[A]

[B]

TIC obtained from grape seed extract treated urine operated in -ve Q1 [A] and +ve Q1 [B] modes

Visual inspection of the two TIC plots show that the two modes of ionization will generate different metabolomic information based on their ionization difference

02/07/2013

8

300 500 700 900 1100 1300 1500 1700m/z0

100

%

577.229

289.097

245.095

425.148

333.091

865.383

729.287675.239

796.319

1153.544

1017.455963.391

1084.991

1305.5941441.710 1730

Profiling of grape seed extract metabolitesin ESI-MS Q-TOF –ve ion mode

catechin

Dimer-

Trimer-

Tetro-

Pento- Hexo-

-gallate

-gallate

O

OH

OH

OH

OH

HO

(+)-CatechinMol wt 290

Source: Chan et al. Rapid Commun. Mass Spectrom. 2007, 21, 519-528

[A] UPLC/TOF-MS total ion current chromatogram (TIC) and HPLC-UV Chromatogram of steamed P. notoginseng

Metabolomics of raw and steamed P. notoginseng

02/07/2013

9

[A] Steamed ginseng

[B] raw ginsend

The concentration of Rg1, Re, Rb1, Rc and Rd in steam ginsengwas less than that of raw ginseng

TIC of UPLC/TOF-MS analysis of [A] steamed ginseng and [B] raw ginseng

[A] Score plot of raw and steamed groups and [B] loadings Plot obtained using pareto scaling with mean centering

Conclusion- MS based metabolomic study is able to discriminate differentially processed herbs such as raw and streamed P. notoginseng

02/07/2013

10

Lipidomics

Lipidomics- A comprehensive analysis of lipid molecules in response to cellular

pathophysiology

GlycerolipidsCholesteryl esters

Why measure lipids?

Lipidomics can perhaps best be defined as a comprehensive analysis of lipids on the systems-level scale together with their

interacting factors

Lipids are important- as a membrane bilayer- provides hydrophobic environment for protein function- reservoir of energy- signaling molecules

Membrane lipidsStorage lipids

PhospholipidsSphingolipidsSterols

02/07/2013

11

Structures of major phospholipids

Cardiolipin (diphosphatidylglycerol)

How to profile phospholipids and sphingosinesin a complex mixture using MS/MS?

PENeutral Loss scan 141

PC & SMPrecursor ion scan 184

Ceramides and sphingosins Precursor ion scan 264

PSNeutral Loss scan 185

02/07/2013

12

Tandem mass spectrometry has the ability to characterize the fatty acyl chain in -ve ion mode

Phospholipids may undergo demethylation and then the loss of the fatty acyl groups from glycerophosphocholinebackbone.

PO-

O

O

N+OO

O

R1O

R2O

PO-

O

O

NO

OO

R1O

R2O [M-15]-

CID

CH3 O

O

CH3 OCH3

O

+

O

R2O

O

R1O +

02/07/2013

13

How to extract lipids?Extraction of lipids by Bligh/Dyer method

• To a homogenized sample (1 ml containing internal standards) add methanol (2.5 ml) and chloroform (1.25 ml), sonicate by 4-5 bursts; extra 1.0 ml water and 1.25 ml chloroform added and vigorously shaken.

• Centrifuge (1,000 x g) for 2 min and separate the chloroform layer (bottom layer) and repeat the process twice.

• Combine the chloroform soluble phases and evaporate to dryness and store at -20oC until analysis.

Survey scan of metabolites (+ ion mode) for a plasma sample from lean mouse [A]; ob/ob mouse [B]. Plasma

lipidomes of obese mice are higher than lean littermates (1.5e3 vs. 863)

400 475 550 625 700 775 850 m/z0

100

%

518.318

496.335

494.326

760.570758.553544.339

546.352566.322

568.337590.321

732.558602.288

782.552

806.556810.592

811.599812.608

835.594

869

400 475 550 625 700 775 850 900m/z0

100

%

518.345496.361

494.352

758.599zz

542.347544.367

546.380

566.351568.366

590.351756.584591.359 703.605

786.632

806.604810.628

811.636812.644

835.632

[A]

02/07/2013

14

MSMS fragmentation of m/z 496 obtained from plasma of obese mouse identified to be LPC 16:0

75 150 225 300 375 450 525 m/z0

100

%

184.080

104.113

86.104 478.348

HOP

HO

O

Om/z 125

125

-H2O

50 200 350 500 650

6.5e6 184.0

703.7

685.8

MS/MS of sphingomyelin standard (2S,3R,4E)-2-acylaminooctadec-4-ene-3-hydroxy-1-Phosphocholine

Even = Odd # NitrogenOdd = Even # Nitrogen

m/z 703.7 is reported as m/z 703, instead of m/z 704odd number m/z = sphingomyelin

Even number m/z = phosphatidylcholine

How to differentiate PC and SM?

02/07/2013

15

Targeted lipidomics- Precursor ion spectra (PRE m/z184) from LNK, ObK and ObNK hepatocytes.

A 2D ESI mass spectrometric finger print for TG molecules

02/07/2013

16

300 360 420 480 540 600 660 700

8.7e4

Intensity, cps

649.0

623.1

651.0

605.0

631.1621.0594.8 636.8

647.1538.8394.7407.6 620.3602.7 633.2576.7

C32

C22:1

C20C16

-H2O

Precursor ion scan m/z 264 in +ve ion mode is specific to identify ceramides in a sample

02/07/2013

17

C20 m/z 594/264

C4, m/z 370/264

C6, m/z 398/264

C8, m/z 426/264

C18, m/z 566/264

C17, m/z 552/264IS

Time, min

0.5 1.5 2.5 3.5 4.50

700

Intensity, cps

0.5 1.5 2.5 3.5 4.50

269

Intensity, cps

0.5 2.0 3.50

299

Intensity, cps

0.5 2.0 3.50

1500

Intensity, cps

1.84

0.5 2.0 3.50

2336

Intensity, cps

0.5 2.0 3.50.0

5.0e4Intensity, cps

1.74

0.5 2.0 3.50

2998

Intensity, cps

2.28

C24 m/z 650/264

MRM chromatograms showing simultaneous determination of ceramides (C4-C24)

Conclusions• LC-MS-based metabolomic approach is promising for

the quality control of dietary supplements, and discovery of novel markers in biomedical research.

• Tandem mass spectrometry analysis of phospholipids in +ve ion mode characterizes phospholipid polar head groups, whereas –ve ion mode provide fatty acid chain structural information.

• Shotgun lipidomics can be used for rapid and reproducible global analysis of lipids in biological samples.

• Identification of metabolites (lipids or any other metabolites) at a molecular level present a great challenge due to their structural diversity (isobars and isomers) and dynamic metabolism.