Life Sciences Mass Spectrometry

School of Pharmaceutical SciencesUniversity of Geneva

University of Lausanne

Geneva, Switzerland

1

Acknowledgments

S. Le Faucheur and V. SlaveykovaInstitute F.A. Forel, University of Geneva, Switzerland

K. Watanabe Shimadzu Corp., Kyoto, Japan

CTC Analytics AG

Zwingen, Switzerland

2

1

1

1

2

2

Instrumentation

Introduction

Conclusions

Overview• Automated Bligh and Dyer extraction for metabolomic studies.

Integrated Platform including Automated Bligh and Dyer Extraction and Dual-Column UHPLC-MS/MS Separations for Metabolomic Analyses of Tissues and Cells

1 Sample preparation workflows for metabolomic studies of tissues or cells require most of the time a Bligh and Dyer extraction or one of its variant (e.g. the Folch extraction). This step is cumbersome and generally perfor-med manually in order to separate the aqueous fraction containing polar endogenous metabolites from the organic fraction containing apolar compounds like lipids. Proteins remain at the interface of the two solvents.

Here we propose to integrate an automated Bligh and Dyer extraction on a robotic system including a dual-column UHPLC-MS/MS platform for the metabolomic analysis of tissues or cells. The aqueous fraction is split and analysed sequentially at two different mobile phase pH values, whereas the lipidic fraction is ana-lysed alternately with an extended gradient.

1 mL of algae culture

Add 1 mL of cold aq. MeOH 80%

(≈ 3.5 x 106 cells)

(-70 °C, metabolic activity quench)

(H2O-MeOH fraction) (CHCl3 fraction)

(0.35 bar, 35°C,10 min.)(30 s, 30 Hz, 0.25 mL glass beads Ø = 1.0-1.5 mm)

• Dual-column UHPLC setup for the analysis of the polar and lipidic fractions.

• Alternating acidic and basic mobile phase for the separation of the polar fraction.

• Identification of unknown compounds by SWATH HR MS2 spectra acquisition.

1

Automated Sample Preparation Platform

Dual-column UHPLC-MS Instrumentation

SoftwareThe RTC robot was controlled by PAL Sample Control v. 2.1 (CTC Analytics).

The two low-pressure gradient UHPLC systems were controlled by LabSolutions v. 5.6 SP2 (Shimadzu) by means of two independent hardware configurations.

The TripleTOF 5600 MS was controlled by Analyst TF 1.6 (AB Sciex) and data processing was done with PeakView v. 2.0 including the MasterView plug-in as well as MultiQuant v. 2.1 (AB Sciex).

Plots were done with Tableau Desktop Professional v. 8.1 (Tableau Software) or Excel 2010 (Microsoft Office Professional Plus 2010).

For chromatographic separation of the B&D fractions, two quaternary low-pressure Nexera LC30AD UHPLC pumps (Shimadzu) were used. The polar fractions were diluted on-line by the RTC robot and injected onto a 100 x 2.1mm XBridge BEH C18 XP column (Waters) running alternately with an acidic or a basic mobile phase. The organic fractions were evaporated to dryness, reconstituted and injected onto a 150 x 2.1 mm XBridge BEH C8 XP column. Both columns were kept at a temperature of 40 °C in a HotDog XL5090 oven (Prolab).

Mobile phases for C18 separations were: A) 5 mM NH4FA + 0.1% FA (pH 3.0), B) ACN + 0.1% FA, C) 0.025% NH4OH (pH 8.3 adj. w/ FA), D) ACN + 0.0125% NH4OH.Mobile phases for C8 separation were: A) 5 mM NH4Ac + 0.1% AA (pH 4.2), B) ACN + 0.1% AA.Gradient was linear from 0-100% in 10 min. for C18 separations (both pH) and in 15 min. for C8 separation with washing steps of 3 and 4.5 min., respectively.

The dual-column UHPLC platform was hyphenated to a TripleTOF 5600 MS (AB Sciex) by means of a Valco-VICI switching valve controlled by contact closure from the respective pumps. MS and MS/MS data were acquired in positive ESI polarity using a Turbo V ion source equipped with an APCI probe for automated calibration (CDS device, AB Sciex). SWATH MS/MS acquisition mode was performed with 24 variable Q1 windows (15-50 u) and a TOF mass range of 50-800 u (polar fractions) and 50-950u (lipidic fractions) with accumulation time of 30 ms. Collision energy was ramped from 20 to 60 V. MS duty cycle was of ca. 0.85 s.

The RTC robot (CTC Analytics) was equipped with several modules as shown on the picture above.

Sample preparation workflow for Bligh and Dyer extraction

2On-line sample dilution with the RTC robot

a) Manual off-line preliminary steps b) Automated on-line sample preparation with RTC platform Automated evaporation to dryness Distribution of the compounds after B&D extraction

3

2

Analysis of the lower organic fraction (UHPLC system 2)

• Reducing the MeOH content in AQ fractions by on-line dilution enables better chromatographic resolution and peak shape.

• Alternating mobile phase pH is essential for improved retention and peak shape for more basic compounds.

• Automated Bligh & Dyer extraction shows similar efficiency, but improved repetability compared with the manual approach.

4Comparison of automated vs. manual Bligh and Dyer extraction

5Unknowns identification with data independent MS/MS acquisition

Emmanuel VaresioGuenter Boehm

Sandra JahnSandrine CudréRenzo Picenoni

Gérard Hopfgartner

Cooled stacks for samples

and fractions

Park stations with the different

transfer tools

Reconstitution solvents rack

Injection valves

Centrifuge

Injector wash station

Low-pressure gradient UHPLC

systems

Vortex-mixer

Evaporating device with its exhaust line

Column oven

Analysis of the upper H2O-MeOH fraction (UHPLC system 1)

Spin down & flash freeze the pellet (N2 liq.)

Cells disruption by cryogenic grinding

(30 s, 30 Hz, mixing with Mixer Mill MM 400)

(855 µL into a 2 mL glass vial)

Add 1’140 µL H2O:MeOH:CHCl3 (0.8:2:1,v:v)

Evaporate to drynesswith N2 gas

(≈ 10% MeOH in samples)

Dilution 5-fold by adding 60 µLto 240 µL of dilution solvent

2) Aspirate 250 µL of lower fraction1) Aspirate 500 µL of upper fraction

Add 225 µL H2O + 225 µL CHCl3Vortex for 10 sCentrifuge for 5 min. (4000 rpm, ≈ 900 x g)

Transfer supernatant

b) Automated B&D extraction with RTC platform

Reconstitute in 150 µL MeOHAdd chromatographic standards

Add chromatographic standards

(UHPLC system 2)Inject 5 µL on C8 column

(UHPLC system 1)Alternately inject 25 µL on C18 column

pH 3 pH 8

Unicellular green algae(Chlamydomonas reinhardtii)

Aspirate upper fraction Aspirate lower fraction

Fig. 4. | Evaporation plots by a gentle N2 flow at 35°C (n=3).

The automated evaporation of 400 µL in 1.2 mL vase vials was achieved in less than 7 min. or 14 min. for CHCl3 or MeOH, respectively.

0 1 2 3 4 5 6 7 8 9 10

E vaporation time [min]

0

50

100

150

200

250

300

350

400

So

lven

t vo

lum

e [µ

L]

CHCl3 - 0.35 bars

0 2 4 6 8 10 12 14 16 18

E vaporation time [min]

0

50

100

150

200

250

300

350

400

So

lven

t vo

lum

e [µ

L]

MeOH - 0.70 bars

Evaporation device

aden

ine

nico

tine

mel

anos

tatin

eac

etam

inop

hen

caffe

ine

aspa

rtam

e

met

opro

lol

coca

ine

aldo

ster

one

corti

sol

test

oste

rone

anan

dam

ide

17-H

O-p

roge

ster

one

dicl

ofen

ac

Δ9 -t

etra

hydr

ocan

nabi

nol

biot

in

theo

brom

ine

4-py

ridox

ic a

cid

3-m

etho

xyty

ram

ine

a) b) c)

Fig. 1 | XIC traces (20 mmu) of selected metabolites (SST 0.2 µg/mL in 10% MeOH - 5 µL). Selected compounds eluting throughout the gradient (a-c) with plots showing the in�uence of MeOH content in sample solvent and injection volume (n=6).

For polar compounds (e.g. plot a), peak width increases when injecting more than 5 µL, but area ratio remains ade-quate up to 10% MeOH in the sample solvent even with an injection volume of 45 µL.

For lipophilic compounds (e.g. plot c), peak width remains at 0.1 min. with large injection (45 µL), however higher or-ganic content in the sample solvent is required (solubility).

Fig. 2 | XIC traces (20 mmu) of selected metabolites (SST 0.2 µg/mL in 50% MeOH). A) No dilution - 5 µL injected, B) 2-fold dilution - 10 µL injected, C) 5-fold dilution - 25 µL injected, D) 10-fold dilution - 50 µL injected. Plots on the right show the on-line dilution results for the selected compounds labeled on the XIC chromatograms (n=6).

The water-methanol Bligh and Dyer fraction contains ≈ 50% MeOH, thus an on-line dilution is necessary to lower the organic content in the sample solvent. Different dilution factors, with their adapted injection volumes to keep the same amount injected on-column, were tested with the RTC robot.

A) No dilution (50% MeOH - 5 µL) B) 2x on-line dilution (25% MeOH - 10 µL)

D) 10x on-line dilution (5% MeOH - 50 µL)C) 5x on-line dilution (10% MeOH - 25 µL)

C omponent Name Dilution F ac tor

0 2 4 6 8 10 12R etention T ime [min]

0.0 0.5 1.0 1.5 2.0A rea ratio relative to 2x dilution

0.0 0.5 1.0 1.5 2.0Height ratio relative to 2x dilution

0.0 0.1 0.2 0.3 0.4Width [min] at 5% peak height

Melanos tatin

None

2x

5x

10x

Metoprolol

None

2x

5x

10x

T es tos terone

None

2x

5x

10x

A nandamide

None

2x

5x

10x

T HC

None

2x

5x

10x

∆

§

#

†

*

For highly polar compounds (e.g. adenine, nicotine), these experimental conditions (i.e. high organic content or large injection volume) were detrimental to their peak shape.

For lipophilic compounds (e.g. ∆, †), losses were observed due to their poor solubility in mostly aqueous solvent (i.e. 5 or 10% MeOH).

For other compounds (e.g. *, §, #), consistent results were observed across the dilution factors.

Alternating pH of mobile phasesThe upper B&D fraction is analyzed at two different mobile phase pH (i.e. pH 3.0 and pH 8.3) to improve the retention of certain polar compounds (e.g. adenine, nicotine). When alternating the mobile phase pH, the column reconditioning time for the basic pH is critical to have a stable system.Thus, injection scheme B (i.e. alternating every injection) was preferred.

1 2 3 4 5 6 7 8 9 10 11 12 13 min.0.0e0

5.0e3

1.0e4

1.5e4

2.0e4

2.5e4

3.0e4

3.5e4

Inte

nsity

x0.25

*

§

∆

#

†

1 2 3 4 5 6 7 8 9 10 11 12 13 min.0.0e0

1.0e4

2.0e4

3.0e4

4.0e4

5.0e4

6.0e4

7.0e4

8.0e4

9.0e4

1.0e5

Inte

nsity

0.25x

Ơ

#

§

*

1 2 3 4 5 6 7 8 9 10 11 12 13 min.0.0e0

1.0e4

2.0e4

3.0e4

4.0e4

5.0e4

6.0e4

7.0e4

8.0e4

9.0e4

1.0e5

Inte

nsity

x0.25

∆

§

#

†*

1 2 3 4 5 6 7 8 9 10 11 12 13 min.0.0e0

1.0e4

2.0e4

3.0e4

4.0e4

5.0e4

6.0e4

7.0e4 0.25

8.0e4

9.0e4

1.0e5

Inte

nsity

x

∆

#

†

§

*

1 2 3 4 5 6 7 8 9 10 11 12 13 min.0.0e0

1.0e4

2.0e4

3.0e4

4.0e4

5.0e4

6.0e4

7.0e4

8.0e4

9.0e4

1.0e50.25

Inte

nsity

x

a) % MeOH Inj. V olume

0 5 10R etention T ime [min]

0.2 0.5 1 2 5 10A rea ratio relative to 10 µL injec tion

0.2 0.5 1 2 5 10Height ratio relative to 10 µL injec tion

0.0 0.2 0.4 0.6 0.8 1.0Width [min] at 5% peak height

5 %

5 µL

10 µL

25 µL

45 µL

10 %

5 µL

10 µL

25 µL

45 µL

25 %

5 µL

10 µL

25 µL

45 µL

50 %

5 µL

10 µL

25 µL

45 µL

C ompound Name - A denine

b) % MeOH Inj. V olume

0 5 10R etention T ime [min]

0.2 0.5 1 2 5 10A rea ratio relative to 10 µL injec tion

0.2 0.5 1 2 5 10Height ratio relative to 10 µL injec tion

0.0 0.2 0.4 0.6 0.8 1.0Width [min] at 5% peak height

5 %

5 µL

10 µL

25 µL

45 µL

10 %

5 µL

10 µL

25 µL

45 µL

25 %

5 µL

10 µL

25 µL

45 µL

50 %

5 µL

10 µL

25 µL

45 µL

C ompound Name - C affeine

c) % MeOH Inj. V olume

0 5 10R etention T ime [min]

0.2 0.5 1 2 5 10A rea ratio relative to 10 µL injec tion

0.2 0.5 1 2 5 10Height ratio relative to 10 µL injec tion

0.0 0.2 0.4 0.6 0.8 1.0Width [min] at 5% peak height

5 %

5 µL

10 µL

25 µL

45 µL

10 %

5 µL

10 µL

25 µL

45 µL

25 %

5 µL

10 µL

25 µL

45 µL

50 %

5 µL

10 µL

25 µL

45 µL

C ompound Name - A nandamide

Fig. 3 | Retention times of selected compounds (SST 0.2 µg/mL) for three injection schemes: A) Only pH 8.3 mobile phase, B) Mobile phase pH was alternated every injection,C) Mobile phase pH was alternated every second injection.

Fig. 6 | Column diagrams showing the peak areas of selected variables (de�ned by m/z and retention time, RT) as well as their coe�cients of variation obtained after analysis of the aqueous (AQ) and organic (ORG) B&D fractions from the automated or the manual extraction procedure (n=5): a) AQ fractions at pH 3 - C18 column, b) AQ fractions at pH 8 - C18 column, c) ORG fractions (pH 4) - C8 column.

To compare variation and repeatability of the automated Bligh & Dyer extraction via the RTC platform with the common manual procedure, n=5 extractions of Chlamydomonas reinhardtii algae were performed, respectively. From the analyses of aqueous (AQ) and organic (ORG) fractions, ca. 20 variables were selected randomly by means of Marker View and/or Peak View software (only monoisotopic peaks with S/N > 30 were considered) and coefficients of variation (CV) were calculated for the corresponding peak areas.

Lower variation (CV < 22%) and, thus, better repeatability was obtained with the automated approach throughout all experiments as shown in figure 6. This result was espe-cially pronounced for the LC-MS runs of AQ fractions at pH 8 and of the ORG fractions.

S tudy pH

0 1 2 3 4 5 6 7 8 9 10 11 12 13

R etention T ime [min]

A

pH 8.3

pH 8.3

pH 8.3

pH 8.3

B

pH 3.0

pH 8.3

pH 3.0

pH 8.3

pH 3.0

pH 8.3

C

pH 3.0

pH 8.3

pH 8.3

pH 3.0

pH 3.0

pH 8.3

A denine Nic otine B iotin C oc aine Dic lofenac 17-HOproges t.. A nandamide T HC

Fig. 5 | Representative XIC traces (20 mmu) of selected metabolites (SST 0.2 µg/mL) sepa-rated by automated Bligh & Dyer extraction (n=3). a) Analysis of the organic fraction recon in 150 µL MeOH on C8 column (inj. 5 µL)b) Analysis of the 5x diluted aqueous fraction at pH 3 on C18 column (inj. 25 µL)

After evaporation, different reconstitution volumes (i.e. 100, 150, 200 µL) of MeOH were tested for the organic fraction (injecting 5 µL on the C8 column). 150 µL were found to sufficiently concentrate the sample and chosen for further experiments.

The automated B&D extraction lead to efficient separa-tion of polar from unpolar metabolites (figure 5).

Only few compounds with amphiphilic properties were re-trieved in both fractions (e.g. metoprolol). These results are in accordance with comparative analyses of manual ex-tracted SST test mix.

a)

b)

17-H

O-p

roge

ster

one

Δ9 -t

etra

hydr

ocan

nabi

nol

coca

ine

test

oste

rone

anan

dam

ide

corti

sol

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17Time, min

0.0e0

5.0e4

1.0e5

1.5e5

2.0e5

2.5e5

3.0e5

Inte

nsity m

etop

rolo

l

aden

ine

aspa

rtam

ebi

otin

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17Time, min

0.0e0

5.0e4

1.0e5

1.5e5

2.0e5

2.5e5

3.0e5

Inte

nsity

acet

amin

ophe

nm

elan

osta

tine

met

opro

lol

LC-MS/MS data from automated and manual B&D extractions of algae were acquired in SWATH mode enabling the generation of data independent MS/MS information due to dynamic Q1 windows.

The results could, thus, directly be subjected to a library search without the need of performing further targeted MS2 experiments. Using both an in-house SWATH-based library and other commonly known spectral libraries (e.g. MassBank), certain unknowns could quickly be identified.

As an example figure 7 shows the experimental and theoretical LC-MS data of adenosine which was identified from analyses of AQ fractions at pH 8 (automated approach).

Fig. 7 | Exemplary result after searching the in-house SWATH library for un-knowns. Adenosine was identi�ed in pH 8 analyses of the AQ algae fractions. In a) the corresponding XIC trace is shown. Valida-tion of the isotopic pattern (Formula Finder) and ob-tained SWATH MS2 spec-trum at 4.76 min. are de-picted in b) and c), respec-tively.

a)

0.0

5.0

10.0

15.0

20.0

25.0

0.00E+00

5.00E+04

1.00E+05

1.50E+05

2.00E+05

2.50E+05

3.00E+05

226.

2/7.

622

8.2/

8.2

256.

2/8.

726

7.2/

10.3

283.

2/5.

130

2.3/

9.9

309.

2/6.

731

7.1/

11.1

322.

2/6.

434

4.2/

5.3

354.

3/9.

736

3.2/

7.1

381.

2/5.

738

8.3/

5.5

432.

3/5.

645

3.3/

6.0

476.

3/5.

765

1.6/

13.5

761.

4/5.

7

CV

[%]

Aver

age

Area

[-]

Variable [m/z/RT]

AQ pH3 automated AQ pH3 manual CV automated CV manual b)

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

0.00E+00

1.00E+05

2.00E+05

3.00E+05

4.00E+05

5.00E+05

6.00E+05

7.00E+05

8.00E+05

122.

1/7.

119

5.1/

4.9

218.

2/8.

122

6.2/

6.2

228.

2/8.

423

4.2/

6.9

239.

2/5.

124

6.2/

9.3

252.

1/4.

925

6.2/

8.7

322.

2/5.

434

4.2/

5.5

354.

3/9.

838

1.2/

6.0

388.

3/5.

643

2.3/

5.8

453.

3/6.

146

6.2/

6.5

476.

3/5.

872

4.3/

7.7

810.

4/7.

9

CV

[%]

Aver

age

Area

[-]

Variable [m/z/RT]

AQ pH8 automated AQ pH8 manual CV automated CV manual c)

0.0

5.0

10.0

15.0

20.0

25.0

30.0

0.00E+00

5.00E+05

1.00E+06

1.50E+06

2.00E+06

2.50E+06

3.00E+06

502.

3/11

.760

0.5/

18.9

606.

5/20

.160

8.5/

20.7

609.

3/15

.462

3.3/

16.4

636.

6/21

.570

4.5/

18.6

710.

6/19

.773

2.6/

19.0

734.

6/19

.573

8.6/

20.4

754.

6/18

.375

6.6/

18.6

758.

6/18

.976

0.6/

19.4

762.

5/17

.976

2.6/

20.2

764.

5/18

.492

4.6/

17.2

926.

6/17

.492

8.6/

17.9

934.

6/19

.2

CV

[%]

Aver

age

Area

[-]

Variable [m/z/RT]

ORG automated ORG manual CV automated CV manual

b)

268.0 268.5 269.0 269m/z

.5 270.0 270.5 271.00

500

1000

1500

2000

2500

Inte

nsity

268.1038

269.1063

270.1083

C10H13N5O4 [M+H]+

N

NN

NNH2

O

OHOH

HO

c)

60 80 100 120 140 160 180 200 220 240 260m/z

-100%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

100%

%In

tens

ity(o

f7.3

e5)

136.06

268.10119.04

136.0619

119.0348137.0635 268.1038

94.039373.0273

Adenosine

Library MS2 Spectrum, CE = 32.5 ± 55 V

SWATH MS2 Spectrum, CE = 20-60 V

• Identification of unknowns is readily feasible due to SWATH-based acquisition following library search.

a)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 170

0.4e3

0.6e3

0.8e4

1.0e4

1.2e4

1.4e4

0.2e3

1.6e4 4.76

min

Inte

nsity