TITLEPlant Metabolomics

Prof Ute Roessner

Head of School, School of BioSciences

Node Leader, Metabolomics Australia

Past Immediate President, International Metabolomics Society

Picture courtesy J. Bowne

What can happen What appears to

be happeningWhat makes it happen What is happening

and has happened

What is Metabolomics about?

This is the analytical approach of a large-scale,

non-targeted and high-throughput determination

of metabolites in a biological system

The real challenge of metabolomics is to comprehensively

cover the analysis of many compounds having

many different chemical structures

“Think metabolism...

If I were doing a PhD [in

cancer research], I'd be

doing it in metabolomics”.Prof. James Watson, Nobel Laureate

Co-discoverer of the structure of DNA

The “Other” Metabolomes(courtesy of Prof. David Wishart)

M mM M nM pM fM

Endogenous metabolites

Drugs

Food additives/Phytochemicals

Drug metabolites

Toxins/Env. Chemicals3230 (T3DB)

1240 (DrugBank)

28500 (FooDB)

1550 (DrugBank)

19700 (HMDB)

Courtesy Prof David Wishart

Biological

Question

Experimental

Design

Sample

Preparation

Data

AcquisitionStatistical

Analysis

Data

Pre-treatmentData

Processing

(Matrix)

Analytics Bioinformatics/Data analysis

Data

Visualization

and

Interpretation

Results

Biology/Analytics/

Bioinformatics

Biology/Analytics

Bioinformatics

• Untargeted

• Targeted

• Instrument

• Replicates

• Controls

• Method dev

• Sample type

• Extraction

• Quenching

• Internal stds

• PBQC

• Instrument

• GC-MS

• LC-MS

• NMR

• Data matrix

• Compound ID

• Quantification

• Normalization

•Transformation

•Multivariate

•Univariate

•Time course

•Pathway

mapping

Metabolomics Workflow

metabolites per cell ~ 1,000 to 6,000

All low molecular weight metabolites

(<1500 Daltons)

Includes, but not limited to:Sugars, sugar phosphates

Organic acids

Amino acids,

Fatty acids

Nucleosides

Lipids (i.e. sterols, phospholipids)

Small peptides

Other secondary metabolites

Drugs, insecticides, other xenobiotics

Total number of metabolites > 500,000

What types of metabolites

can be analysed?

Comprehensive workflow

shock freeze homogenizationtissue harvest extraction procedure

derivatisation

data analysisdata interpretation

Integrated Biology workflow

shock freeze homogenizationtissue harvest extraction procedure

G Lotus, J., WT, Greenhouse, flowers (R7A ), 23mg, 1/10 Acquired on 31-Jan-2002 at 06:55:04G

10.000 15.000 20.000 25.000 30.000 35.000 40.000 45.000 50.000 55.000

rt0

100

%

0

100

%

0

100

%

0

100

%

0

100

%

26.144

19.24614.82310.529

15.850

23.28622.312

38.038

27.797 32.03139.931 44.362

48.27847.513

Scan E I+

TIC

7.88e7

RT

2030BG03

26.164

19.27514.860

13.01910.52124.24519.841

38.077

31.554

30.52427.31032.037

34.627 36.805

39.933 44.35748.276

Scan E I+

TIC

9.69e7

RT

2030BG05

26.152

26.002

19.68713.02610.521 14.815 24.249

22.323

31.555

30.33227.318

33.561

38.02133.61939.935

44.36048.287

Scan E I+

TIC

7.86e7

RT

2030BG11

26.147

13.02310.519

19.85019.244 24.255

31.538

30.33227.313

33.549

38.014

33.60639.933

44.35648.271

Scan E I+

TIC

6.95e7

RT

2030BG12

26.133

19.249

13.02810.532

14.81618.138

23.02819.684

22.825

25.995

33.56331.55426.893

27.295

28.175

29.104

38.031

33.61239.934

44.35848.274

Scan E I+

TIC

6.73e7

RT

2030BG13

Genomics Transcriptomics Proteomics Metabolomics

Example: LC-MS approach

Sample

Polar extract (e.g. MeOH)

Apolar

compounds

Polar

compounds

C18RP HILIC (ANP)

+/- ESI QTOF +/- ESI QTOF

Lipid extract (e.g. chloroform)

Lipid

profile

Lipid species

quantifictaion

LC-QTOF MRM using QQQ

~ 2700 compounds

Tandem LC-QqTOF-MS

hydrophobic hydrophillic

Targeted vs Untargeted

Absolute quantification

Pool sizes

You know what you measure

Not comprehensive

More accurately

Better sensitivity

Easier to interpret

Pathway mapping

Semi-quantitative

X-fold comparison

Biomarker discovery

More comprehensive

Loads of unknowns

Multivariant analyses

“detect the expected” “detect the unexpected”

Quantitative metabolomics

36 amino acids and 30 amines 47 sugars and sugar alcohols

36 fatty acids 22 organic acids

Boughton et al. 2011 Dias et al. 2015

Identifying novel salinity tolerance

mechanisms by spatial analysis

of lipids in barley roots

Abiotic stress affects agricultural

productivity in Australia

Australian wheat belt significantly affected by salinity

In danger: Australian wheat exports worth ~ $4 billion p.a.

Rengasamy (2002) Aust. J. Exp. Agric. 42: 351-361

Effect of salt stress on plants

Roy et al, 2012

Effect of salt stress on roots

1

cm

1

cm

Control Salt

Project AIM: Understand early

tissue-specific salt responses

Study Design

Barley Cultivars: Clipper and Sahara

Growth and Treatment on Agar Plates

Root Tissue Harvest(3 Zones: Root Cap, Elongation and Maturation Zone)

Cell Wall Polysaccharide Analysis

Targeted Metabolite Profiling

RNA-Seq Analysis

Data Analysis and Statistics

Data Integration and Biological Interpretation

1. Cell Wall Polysaccharide Analysis

2. Targeted Metabolite Profiling

3. RNA-Seq Analysis

Significant differences (Sug, OA, AA)

between cultivars

CONTROL SALT

CDZ EZ MZ CDZ EZ MZ

Clip Sah Clip Sah Clip Sah Clip Sah Clip Sah Clip Sah

Significant changes (Sug, OA, AA)

following treatment

CLIPPER SAHARA

CDZCONTROL vs SALT

EZCONTROL vs SALT

MZCONTROL vs SALT

CDZCONTROL vs SALT

EZCONTROL vs SALT

MZCONTROL vs SALT

Significant differences (Sug, OA, AA)

between zonesCLIPPER SAHARA

CONTROL

CDZ vs EZ

SALT

CDZ vs EZ

CONTROL

CDZ vs MZ

SALT

CDZ vs MZ

CONTROL

CDZ vs EZ

SALT

CDZ vs EZ

CONTROL

CDZ vs MZ

SALT

CDZ vs MZ

Tissue-specific lipid analysis

Next-Generation RNA Sequencing

A clear per-zone (horizontal

axis) response is observed in

this Pearson’s correlation plot

illustrating cell differentiation

along the root zone.

Z1 Z2 Z3

Z1

Z2

Z3

Hill, C. B., et al. (2016). Sci Rep 6: 31558.

Hill, C. B., et al. (2016). Sci Rep 6: 31558.

◦Root zones are highly specialized in their biological function

◦Transcriptional differences are much larger between zones than between treatments.

◦Biochemical processes are controlled in a root-zone specific manner

Transcriptomics analysis (short-term salt stress)

Significant metabolites - Clipper

Significant metabolites - Sahara

Salt signalling

Proposed lipid based signalling

Munnik and Vermeer 2010

Total lipid analysis in

salt treated roots

Natera et al, 2016, Funct Plant Biol

Gradient Category and Source Polarity

Gradient 1/Neg Gradient 2/Pos Gradient 3/Pos Gradient 4/Pos

PC

PE

PG

PS

PI

SQDG

DGDG

MGDG

Cardiolipin

LysoGlycerolipid

PA

Ceramide

HexCer

DAG

ASG

SG

GIPC

OHC

TAG

SE

40

40

1815

1512

34

32

33

13136314887

363715

2722

521

45

60

150

Number of lipid species identifiedPC PE

PG PI

PS PA

MGDG DGDG

SQDG LysoPC

LysoPE LysoPG

LysoPI LysoPS

LysoPA LysoMGDG

LysoDGDG LysoSQDG

DAG TAG

Cardiolipin ASG

SE SG

Cer HexCer

OligoHexCer GIPC

Yu et al 2018

The barley root lipidome

Where to next?

shock freeze homogenizationtissue harvest extraction

G Lotus, J., WT, Greenhouse, flowers (R7A ), 23mg, 1/10 Acquired on 31-Jan-2002 at 06:55:04G

10.000 15.000 20.000 25.000 30.000 35.000 40.000 45.000 50.000 55.000

rt0

100

%

0

100

%

0

100

%

0

100

%

0

100

%

26.144

19.24614.82310.529

15.850

23.28622.312

38.038

27.797 32.03139.931 44.362

48.27847.513

Scan E I+

TIC

7.88e7

RT

2030BG03

26.164

19.27514.860

13.01910.52124.24519.841

38.077

31.554

30.52427.31032.037

34.627 36.805

39.933 44.35748.276

Scan E I+

TIC

9.69e7

RT

2030BG05

26.152

26.002

19.68713.02610.521 14.815 24.249

22.323

31.555

30.33227.318

33.561

38.02133.61939.935

44.36048.287

Scan E I+

TIC

7.86e7

RT

2030BG11

26.147

13.02310.519

19.85019.244 24.255

31.538

30.33227.313

33.549

38.014

33.60639.933

44.35648.271

Scan E I+

TIC

6.95e7

RT

2030BG12

26.133

19.249

13.02810.532

14.81618.138

23.02819.684

22.825

25.995

33.56331.55426.893

27.295

28.175

29.104

38.031

33.61239.934

44.35848.274

Scan E I+

TIC

6.73e7

RT

2030BG13

Transcriptomics Metabolomics Lipidomics Cell wall

Eluant Flow

Boughton

A

B

C

D

MALDI-IMS of barley roots

+ 150mM

NaCl

Ctrl

Spatial

Bulk

Analysis

MALDI

Imaging

MS-Dial

x20-

40

x3

x3

300 400 500 600 700 800 9000

5

10

15

1000m/z

Metabolite

Annotation

30 x 30 µm

~150-300,000 Res. @ 400 m/z

1-2% CMC

x20-

40

IPA/dry-ice

quench

Cryo-section

roots

Matrix coating

(DHB) Discriminative

analysis

Barley

seedlings

Dissecting

roots

Lipid

extraction

HPLC-MS

Peak detection

WorkflowWorkflow

Tentative Annotations via MSI

◦ 230 spatially distributed ions across different barley cultivars (Gairdner and Hindmarsh)

◦ Annotated using SimLipid and Metlin (5 ppm), and curated for multiple adducts & isobars

Results

Gairdner

Con

trol

15

0 m

M N

aC

l

2-hexoses PA 49:6 SQDG 26:1

Hindmarsh

Con

trol

15

0 m

M N

aC

l

2-hexoses PA 49:6 SQDG 26:1

Results

Hindmarsh

Con

trol

15

0 m

M N

aC

l

PC 36:5 LPS 27:1 PE 44:12

Gairdner

Con

trol

15

0 m

M N

aC

l

PC 36:5 LPS 27:1 PE 44:12

Results

Gairdner

Con

trol

15

0 m

M N

aC

l

PE 42:11

SHexCer t34:1 PI-Cer t34:0

Hindmarsh

Con

trol

15

0 m

M N

aC

l

PE 42:11

SHexCer t34:1 PI-Cer t34:0

GPC -

[M+H]+

GPC -

[M+K]+

GPC -

[M+Na]+

ResultsC

on

trol

15

0 m

M N

aC

l

GPC -

[M+H]+

GPC -

[M+K]+

GPC -

[M+Na]+

Gairdner Hindmarsh

Con

trol

15

0 m

M N

aC

l

◦ GPC is a degradation product of phosphatidylcholine, resulting from the removal of fatty acids by the activity of phospholipase A1, phospholipase A2 and lysophospholipase(Sahsah et al., 1998).

◦ The presence of GPC is generally interpreted as a stress-induced membrane turnover or degradation (Aubert et al., 1996).

Spatial Bulk Analysis – LC-MS

Cell Division Zone

Elongation Zone

Maturation Zone

Maturation Zone Elongation

Zone

Cell Division

Zone

Gairdner

230 Curated Features

145 Tentative ID’s SimLipid

48 Tentative ID’s Metlin

Metabolites and FA’s are discriminatory

71 Discriminatory

Salt vs Ctrl

Discriminatory Features Associated

with Zones 2 and 3

Zone 1

Zone 2

Zone 3 Hindmarsh

277 Curated Features

166 Tentative ID’s SimLipid

65 Tentative ID’s Metlin

51 Discriminatory

Salt vs Ctrl

3 Discriminatory

Ctrl vs Salt

22 Discriminatory

Salt vs Ctrl

FA’s and SL are discriminatory

Metabolites and FA’s are discriminatory

GP’s and SL’s are discriminatory

Summary

Summary

WTWT

AVP10

AVP10Low saline High saline

• Tissue type-resolved analyses demonstrate

tissue-specificity of metabolism

• Large differences between genotypes of the

same species

• Lipids are important players in salinity

response

• Imaging MS provides new means of tissue-

specific metabolite analysis

Future for spatial analysis

Con

trol

15

0 m

M N

aC

l

GPC -

[M+H]+

GPC -

[M+K]+

GPC -

[M+N

a]+

Gairdner

Microscopy Imaging MS Micro-XRF

FT-IRRoot phenotyping

Acknowledgements

Roessner Research Group

◦ William Ho

◦ Daniel Sarabia

◦ Dingyi Yu

◦ Sneha Gupta

◦ Martino Schillaci

◦ Tannaz Zare

◦ Allene Macabuhay

◦ Bo Eng Cheong

◦ Cheka Kehelpannala

◦ Holly Khoury

CollaboratorsCamilla Hill (Murdoch U)

Josh Heazlewood (Uni Melb)

Berit Ebert (Uni Melb)

Megan Shelden (Uni Adelaide)

Mark Tester (KAUST)

Sandra Schmoeckel (KAUST)

Richard Caprioli (Vanderbilt)

Jeff Spraggins (Vanderbilt)

Ivo Feussner (Goettingen Uni)

Michelle Watt (Juelich)

Uli Schurr (Juelich)

Borjana Arsova (Juelich)

Joachim Kopka (MPIMP)

Metabolomics Australia,

University of MelbourneDr Thusi Rupasinghe

Dr Berin Boughton

Dr Siria Natera

Dr Adrian Lutz