A NOVEL APPROACH TO MALDI-TOF-MS SAMPLE PREPARATION

Post

erRE

PRIN

T

Ed Bouvier2, Jeff Brown1, Emmanuelle Claude1, John L. Gebler2, Weibin Chen2, *Dominic Gostick1, Kevin Howes1,James Langridge1, and John Peter Lee2

1Waters Corporation, Manchester, UK. 2Waters Corporation, 34 Maple Street, Milford, MA 01757-3696, USA

Introduction

MALDI-TOF-MS has become a well establishedtechnique for the analysis of biological samples. Thishas been mainly due to its ease of use and relativeinsensitivity to biological matrixes which are used inthe preparation of most biological samples.However, it has been previously demonstrated thatremoving these contaminants can significantlyimprove the quality of the resulting spectra.

Furthermore MALDI-TOF-MS is now a technique whichis routinely automated for both the samplepreparation and analysis of many biological samples,which include those resulting from a proteomicsexperiment. Clearly, any method of samplepreparation developed for MALDI-TOF-MSapplications must be amenable to full automation.

Presented in this poster is a new plate design forsample preparation of biological samples (patentsapplied for). The plate design allows for theconcentration of very dilute samples and the removalof inorganic salt contamination. The whole samplepreparation procedure is performed in full upon theMALDI-TOF-MS sample stage. Presented in the posteris the optimised plate design and preparationmethods to significantly improve the quality of theresulting MS spectrum compared with conventionalMALDI sample preparation.

The MALDI sample preparation of protein digestsresulting from in gel digestion of gel spots from 2Dgel electrophoresis are evaluated. It is shown in thisposter that the sensitivity of the mass spectrometer isdramatically improved for faint gel spots resultingfrom both Coomassie blue and silver stained gelpieces. The method is compared with alternativemethods of sample preparation and its amenabilityfor automation is demonstrated.

Methods

The sample stage or plate of a MALDI-TOF massspectrometer has been modified such that an increasedsample volumes (upto 10 µL rather than typically 1 µL)can be loaded. The increased volume is allowed tofocus onto a small region of an adsorbent surface.

In addition, the deposited sample containing bio-molecules such as peptides can be washed in situ withwater to remove contaminants such as inorganic lowmolecular salts (for example NaCl), detergents (forexample SDS and CHAPS ), organic buffers (forexample TRIS). Therefore the resulting mass spectraare significantly enhanced by the focussing and clean-up of the sample. The sample preparation process is asimple procedure and is amenable to full automationby a liquid handling robot (MassPREP Micromass).

The standard stainless steel MALDI sample plate (Figure1a) is coated with a thin layer of PTFE film. The centreof the sample well is then etched to remove the PTFEfilm to produce a small clearing with a diameter of0.4-0.6 mm (Figure 1b). The entire plate surface is thencoated with a thin film of a hydrophobic polymer thathas adsorbent properties (Figure 2).

Figure 1. The MALDI-TOF-MS sample stage (a) Thestandard stainless steel plate (b) The modifiedMALDI sample stage

Presented at ABRF 2002, Austin, Texas, USA, 9th - 12th March 2002

A:

B:

A:

B:

Post

erRE

PRIN

T

Figure 2a. An aerial view of the modified targetplate

Figure 2b. A cross sectional view of the modifiedtarget plate

The sample preparation method is described asfollows. The adsorbent phase surface is activatedby adding a small volume (approximately 1µL) ofan organic solvent such as acetonitrile. Theorganic solvent allows the hydrophobic adsorbentphase to be wetted. An equal volume of water isthen added to the organic solvent to reduce therate of evaporation. This conditioning solution isthen removed just prior to the addition of sample.The sample is deposited in a large volume of 5-10µL. The sample solution contains 20-30%acetonitrile to further aid wetting of the adsorbantphase surface. The large volume sample loading ispossible because the hydrophobic surface providesan increased contact angle with the sample solutioncompared to the stainless steal sample plate. Inaddition the sample moat geometry maintains thehigh contact angle and acts as a barrier to thedroplet perimeter. The combination of both thehydrophobic surface and sample moat gives anapproximate 5-10 fold improvement in samplevolume retention (Figure 3).

Figure 3. The sample volume capacity on themodified target plate

Figure 4. Focussing and sample localisation oftarget plate macro-structure

The analyte solution is allowed to evaporate.During the evaporation the analyte bio-moleculesare immobilised onto the central portion of theadsorbent phase surface. The bio-molecules areforced to the surface by non-specific hydrophobicinteractions. Although the adsorbent phasepolymer is highly hydrophobic, significantfocussing of the sample to the central portion of theadsorbent phase surface is observed (Figure 4).Once the sample solution has completelyevaporated the analyte bio-molecules becomeimmobilised to the adsorbent phase surface. Thesample positions can then be washed individuallywith the use of a pipette to remove the inorganicsalts (Figure 3).

The washed sample can then be analysed directlyon the sample plate by the addition of a smallvolume of MALDI matrix such as CHCA inapproximately 1µL. The matrix solvent has a highorganic content typically 70-90%.

Sample Moat

PTFE film coatedwith adsorbentpolymer

Central clearing coated withadsorbent polymer

PTFE filmStainless Steeltarget plate

Adsorbent Phase

Sample MoatAdsorbentpolymer

Post

erRE

PRIN

T

The solvent will interact with the bio-molecules fromthe adsorbent phase surface so allowing the co-crystallisation of analyte and matrix whilst beingfocussed into a small area. The sample plateswere analysed by MALDI-TOF-MS (M@LDIMicromass)

Results

ResolutionA solution of Alcohol dehydrogenase (ADH; 100fmol) was deposited onto the modified target plate.The sample was washed in TFA (0.1% v/v) beforethe addition of MALDI matrix CHCA (0.1 mg/ml).The sample was then analysed by MALDI-TOF-MSand the full width at half maximum resolution wasestimated from the (M+H)+ 968 and 2212 ions(Figure 5). The FWHM resolution of 13000 and13500 is typical for the routine resolution obtainedfrom the standard MALDI-TOF-MS target plate (datanot shown)

Figure 5. MS resolution on the modified targetplate

Mass AccuracyThe mass measurement accuracy was investigatedon the surface of the modified plate. A solutiondigested of ADH was deposited onto the targetplate at four concentrations (1, 10, 100 and 1000fmol) and a calibrant was added to the near pointsample well at a concentration of 100 fmol. Thesample wells were analysed by MALDI-TOF-MS andthe resulting spectra were lock mass correctedtwice, once from a single internal peptide andonce from the near point calibrant peptide.

The mass measurement accuracy can be seen inFigure 6 are the overall RMS mass measurementaccuracy for the internal and near point methodare 7 and 37, respectively.

Figure 6. Mass measurement accuracy from themodified target plate

Sensitivity and Sample FocussingThe sample focussing effects and limit of detectionof the modified target plate were demonstratedfrom a solution digest of BSA. A serial dilution ofintact BSA was made to give two solutions at 1fmoland 0.2fmol prior to digestion. After digestion withtrypsin equal quantities of digest sample weredeposited onto the plate (1 µl of 1fmol/µl and 5µlof 0.2fmol/µl). The samples were washed beforeanalysis by MALDI-TOF-MS. The resulting spectraare shown in Figure 7.

Figure 7. The Sensitivity and Sample Focussing ofthe Modified Target Plate

900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700m/z0

100

%

968.511

1618.898

1251.716

969.515

970.498

1252.714

1312.720

1313.719

1447.830

1619.899

2313.219

2312.217

1620.900

1669.906

1671.912

2314.223

2315.226

965 966 967 968 969 970 971 972 973 974 975m/z0

100

%

968.511

969.515

970.498

FWHM = 13000

2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321m/z0

100

%

2313.219

2312.217

2314.223

2315.226

2316.229

FWHM = 13500

Sample Conc 1 pmol

Well No A1 A5 A9 B1 B5 B9Internal (PPM) 3 5 7 6 5 11Near point (PPM) 3 25 17 12 20 71Sample Conc 0.1 pmol

Well No A2 A6 A10 B2 B6 B10Internal (PPM) 4 6 9 3 8 8Near point (PPM) 28 18 31 7 45 43Sample Conc 0.01 pmol

Well No A3 A7 A11 B3 B7 B11Internal (PPM) 3 11 6 7 7 7Near point (PPM) 29 60 34 42 16 40Sample Conc 0.001 pmol

Well No A4 A8 A12 B4 B8 B12Internal (PPM) 5 8 10 7 11 11Near point (PPM) 44 36 60 18 30 47RMS (PPM) Near point Internal

32 7

1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200m/z0

100

%

0

100

%

EC020129_18 6 (1.469) TOF LD+ 1.35e3927.463

1044.004

938.009

938.401

975.432

1479.8061044.985

1046.032

1439.8491170.6141072.043

1079.365

1300.0531249.593 1305.727

1306.653

1307.605

1882.9081481.806

1639.976

1567.745

1673.928

1779.7811883.992

EC020129_14 4 (0.612) TOF LD+ 1.23e31479.806

927.454

1044.018

928.440

1017.556

938.433

1044.999

1439.8731046.024

1170.585

1163.580

1061.966

1062.956

1301.0221171.601

1301.996

1401.624

1480.792

1881.9481481.857

1639.951

1537.805

1779.757

1641.864 1781.779

1883.910

1884.818

1 fmol/µL x 1 µL = 1 fmol onto target

0.2 fmol/µL x 5 µL = 1 fmol onto target

Post

erRE

PRIN

T

Figure 8. A flow diagram summarising theautomate protocol for in gel digestion and sampledeposition onto the modified target plate

Salt toleranceThe modified target plate was demonstrated tofacilitate the successful removal of a range ofcommon contaminants. Good quality spectra couldbe obtained after washing from samplescontaminate with the following salts and detergents(NaCl 250mM; SDS, 1.0%; Tris, 25mM; Glycine,100mM). After washing no significant differencecould be observed between the contaminated andnon-contaminated spectra (data not shown)

AutomationThe standard digestion protocol of the liquidhandling system (MassPREP Station; Micromass)was modified to incorporate sample deposition andwashing on the modified target plate. A flowdiagram summarising the modified protocol can beseen in Figure 8.

Analysis of sample from SDS-gel electrophoresisSamples of BSA (250 fmol) were ran on a 1 DSDS-PAGE (Bio-Rad). The gel was stained withCoomassie Brilliant Blue (Bio-Rad) and bands atapproximate 70 KDa were excised and place in a96 well microtitre plate. The resulting gel pieceswas automatically destained, reduced andalkylated and then digested with trypsin use therobotic liquid handling system (MassPrep:Micromass UK Ltd). An aliquot (2µL) of theextracted peptides was deposited onto a standardstainless steel MALDI target plate before theremaining extracted peptide (6µL) were depositedand wash on the modified target plate. The twoplates were analysed by MALDI-TOF-MS (Figure 9).The resulting spectrum from the modified targetplate shows an improved signal to noise andpeptide coverage compared to the standardspotting.

A second set of samples of BSA (250 fmol) wereran a SDS-PAGE and were stained with silver stain(Bio-Rad) and bands were excised and digested asabove. The resulting extracted peptides wereprepared on either a modified target plate orprepared using a micro reverse phase pipette tip(ZipTip; Millipore) and spotting to a standard targetplate. The resulting MALDI-TOF-MS spectrum fromthe modified target plate shows an significantimprovement in both signal to noise and peptidecoverage compared to the standard plate (Figure10).

Figure 9. A comparison of the modified target plateand the standard target plate

Start

Digest6 ng/µL trypsin in 50 mM ammoniumbicarbonate (25 µL), 5 hours at 37ºC

Extract60% aq. 1% formic acid /40% acetonitrile (30 µL)

Gel pieces in 96-wellmicrotiter plate

Extract50% aq. 1% formic acid /50% acetonitrile (24 µL)

End

Reduce10 mM dithiothreitol in 100 mM

ammonium bicarbonate,30 minutes

Alkylate55 mM iodoacetamide in

100 mM ammonium bicarbonate,20 minutes

Wash100 mM ammonium

bicarbonate / acetonitrile

Dehydrateacetonitrile

Coomassie de-stain100 mM ammonium bicarbonate

/ acetonitrile

Digest SPP1.0.MPT

No

De-stain?

Silver de-stain15 mM potassium ferricyanide/

50 mM sodium thiosulphate

Coomassie Silver

Gel type? 1D

no

2D

Spot8 µL extracted

peptides

TransferRemaining extracted peptides

to PCR plate for CapLC24 µL

Secondextraction?

TransferPCR plate for

CapLC15 µLYes

Thirdextraction? YesNo

Extract50% aq. 1% formic acid/50% acetonitrile (24 µL)

TransferPCR plate for

CapLC15 µL

Condition3 µL Acetonitrile

Wash4 µL 0.1 % TFA

Wash4 µL 0.1 % TFA

Wash4 µL 0.1 % TFA

MALDISPP target

Spot Matrixcyano 0.5 mg/mL in90% acetonitrile /10% aq. 1% TFA

1000 1050 1100 1150 1200 1250 1300 1350 1400 1450 1500m/z0

100

%

0

100

%

DG010719_10 4 (0.130) TOF LD+9.54e31479.93

1179.69

1163.721045.64

1121.671107.62

1046.64

1475.90

1180.71

1444.86

1365.771445.89

1480.91

1481.95

BSA_026 4 (0.131) TOF LD+4241060.11

1044.13

993.79

1011.37

1026.34

1480.901479.89

1066.16

1163.71

1081.381107.61 1121.61

1399.781305.80

1184.67 1300.611202.85 1234.71 1306.801367.831329.59

1400.76

1475.861444.83

1445.77

1481.93

1502.761504.71

2 l spotted on Standard Plateµ

6 µl spotted on a Modified Sample Plate

Post

erRE

PRIN

T

Figure 10. A comparison of the modified targetplate and a micro reverse phase pipette tip

Conclusion

● A modified target plate has been developedwhich allows large volumes (1-10 µL) ofsample to be deposited onto the MALDI-TOF-MS sample stage. The modified target platecan be washed to effectively remove samplecontaminants. Biomolecules of interest remainattached to the hydrophobic surface.

● The modified target plate does not impede theresolution and mass measurement accuracy ofthe MALDI-TOF-MS. Furthermore the modifiedtarget plate significantly enhances thesensitivity of the MALDI-TOF-MS allowing limitsof detection in the atomolar range.

● The deposition and subsequent washing isamenable to automation can be incorporatedinto a full digestion protocol

● The modified target plate is shown to givesignificant improvements when analysingsamples prepared from an in gel digestion.The peptide coverage and signal-to-noise issignificantly enhanced compared withstandard preparation methods.

1100 1150 1200 1250 1300 1350 1400 1450 1500 1550 1600 1650 1700 1750 1800m/z0

100

%

0

100

%

dg010718_07 2 (0.064) TOF LD+3.35e31444.84

1163.65

1107.54

1109.51

1129.62

1164.65

1179.62

1180.63

1193.64

1305.761194.66 1235.60

1234.66 1265.68 1365.73 1390.76

1445.84

1479.89

1446.851480.90

1511.931512.94

1640.051513.94

1674.97 1794.94

ZipTip552 3 (0.097) TOF LD+4661235.57

1107.57

1112.63

1113.63

1142.611143.65

1195.691232.65

1236.57

1479.86

1308.721237.59

1321.751373.75

1412.81 1477.82

1640.00

1480.86

1481.86

1482.891569.811484.79

1519.831591.89

1791.81

1641.00

1642.01

1741.921774.79

1792.82

1793.81

Modified Sample Plate

ZipTip

Post

erRE

PRIN

T

Author to whom all correspondence

should be addressed:

Dominic Gostick

Waters Corporation

(Micromass UK Limited)

Floats Road, Wythenshawe

Manchester, M23 9LZ

Tel: + 44 (0) 161 946 2400

Fax: + 44 (0) 161 946 2480

e-mail: dominic.gostick@micromass.co.uk

©2002 Waters Corporation November 2002 / WMP189

For research use only. Not for use in diagnostic procedures.

WATERS CORPORATION34 Maple St.Milford, MA 01757 U.S.A.T: 508 478 2000F: 508 872 1990www.waters.com

Made in the United Kingdom

Certificate No: 951387