RAPID COMMUNICATIONS IN MASS SPECTROMETRY, VOL. 6,105-108 (1992)

Peptide Sequencing by Matrix-assisted Laser-desorption Mass Spectrometry Bernhard Spengler,* Dieter Kirsch and Raimund Kaufmann Institute of Laser Medicine, University of Diisseldorf, Moorenstr. 5, D-4000 Diisseldorf, Germany

Ernst Jaeger Max-Planck-Institute of Biochemistry, D-8033 Martinsried, Germany SPONSOR REFEREE: Dr P. Roepstorff, Department of Molecular Biology, University of Odense. Odense, Denmark

A novel method of peptide sequencing by mass spectrometry is described. Metastable decay of laser-desorbed ions, taking place in the first field-free drift region of a reflectron time-of-flight mass spectrometer, has been monitored to get structural information from larger peptides. Fragment ions from metastable decay are mass analysed by adjusting the potentials of the ion reflectron according to the kinetic energies of the ions. The features of the technique and its significance for future applications are outlined.

Matrix-assisted laser-desorption is known to have several advantages over other methods of ion formation, like the extension of the mass range to over 100 000 Da,’ high sensitivity,2fast analysis, insensitivity with respect to solvent i m p ~ r i t i e s ~and . ~ others. Structural information cannot be achieved directly since the molecules do not undergo prompt fragmentation (prior to desorption). However, intensive metastable decay, occurring on a longer time-scale in the field-free drift region of a time-of-flight mass spectrometer, can be used to analyse the structure of pep tide^.',^ In particular, side chain fragmentations, known to ease sequence analysis dramatically, can be observed at high intensities. Both unimolecular and bimolecular (by collision with residual gas molecules) decay of laser-desorbed peptide and protein ions has been found to take place in the field-free drift region of a time-of-flight mass spectrometer to a great extent, depending on various instrumental parameter^.^ Sensitivity in matrix-assisted laser-desorption is known to be extremely high (down to the low femtomolar range)2 and can be assumed to be that high even in the sequencing mode of operation. Further data suggest that sequence information by metastable decay should be achievable even for peptides or proteins with molecular weights exceeding 5000 Da.

Stable ions were reflected in a two-stage r e f l e ~ t r o n . ~ Ions were detected by a dual channel-plate detector. Metastable ions behave differently, since their kinetic energy is dependent on the mass fraction they keep of the parent mass (velocity of products basically stays constant after fragmentation). Thus, large metastable fragment ions appear as normal peaks in the mass spectrum (the reflectron compensates for their energy loss). Smaller and smaller metastable fragment ions, on the other hand, eventually have less kinetic energy than is needed to reach the second stage of the reflectron. These ions get reflected in the first stage and appear as a broad peak in the mass spectrum. The total flight time of such species is more or less independent of their fraction of the parent ion mass, being determined by the velocity of the parent ion. Fragment ions from metastable decay thus can be mass analyzed, if their kinetic energies lie within the range of kinetic energy deviations compensated by the two-stage reflectron. To get the complete mass spectrum of metastable fragment ions, the reflectron potentials have to be stepped down (leaving the acceleration voltage constant) in order to compensate for the lower kinetic energies of the smaller fragment ions. Fragment ion masses then have to be calculated from the instrumental parameters.

EXPERIMENTAL The instrument used is described schematically in Fig. 1. Samples were prepared using sinapinic acid as a matrix.8 Sinapinic acid was dissolved in a 1:l mixture of ethanol+water. Peptides were dissolved in a 1 : l mixture of ethanol + 0.1% trifluoroacetic acid. Matrix and analyte solutions were mixed to a molar ratio of =300:1.Ten pL of the mixture was blow-dried on a polished aluminum target. Ions formed by a pulsed UV laser beam (nitrogen laser, h = 337 nm) were accelerated to 5-20 keV kinetic energy. An irradiance slightly above the threshold of ion detection was used (“5 x lo6W/cm2). The residual mBar. gas pressure was held at 9 X

RESULTS AND DISCUSSION Structural information by metastable decay has been obtained from various peptides and proteins up to MW 12 360 (cytochrome c ) . A complete structural analysis with rather good mass resolution has been performed for several smaller peptides in the range 1000 to 2000Da, used as test substances. Figure 2 shows a typical spectrum of a small range of metastable fragment ions from a synthetic peptide (MW 1084.6). The settings of the potentials for this spectrum were: acceleration voltage 8.6 kV, first-stage reflectron potential 4.44 kV, second-stage reflectron potential 6.6 kV. The mass range of metastable fragment ions accepted by the reflectron in this case is 560 to 832 Da. The broad intense peak on the left edge is due to fragment ions reflected in the first stage of the reflec-

Author to whom correspondence should be addressed. 0951-4198/92/020105-04 $05.00 @ 1992 by John Wiley & Sons, Ltd.

Received 4 December 1991 Accepted 4 December 1991

PEPTIDE SEQUENCING BY MATRIX-ASSISTED LASER-DESORPTION MASS SPECTROMETRY

106

100 1621

I349

z .c:

75

2848

I

I

I

4568

I

3

. a 9

4

2 50 small

*-

3

25

*-

..* ... .. .*i ..!"stable

channel plate detector

.... "'.::::::.,.. i .......:':::::;....::...* ***

ions

&large . . metastable i. .i fragments

tastable i lments i sample , acceleration

4000 mlz

.. ... .. .. ... ...

-*

--"*

I W I I

ll, I

Figure 1. Schematic diagram of the reflectron time-of-flight mass spectrometer for matrix-assisted laserdesorption. The complete spectrum of metastable fragment ions can be analyzed by stepping down the reflectron potentials 17, and U2. The mass spectrum of a peptide mixture of substance P (MW 1348), bombesin (MW 1620), melittin (MW 2847) and ACTH (MW 4567) demonstrates the appearance of the different types of ions. See text for further explanation.

tron (kinetic energy lower than 4.44 keV). Larger fragments (kinetic energy higher than 6.6 keV) were not reflected but transmitted through the third grid of the reflectron. Varying the reflectron potentials systematically, leads to a series of spectra from which the full fragmentation pattern of the peptide can be derived. mail fragments

I I I I I I I I I I I I I I I I I I I I I I I I i 90 100 110 t [PI

Figure2. Typical spectrum of a small range of metastable fragment ions from the synthetic peptide Gly-Ala-Lys-Ala-Val-Gly-Glu-Ala-Lys-Ala-Ala-Leu (MW 1084.6). Acceleration voltage was 8.6 kV, first stage reflectron potential was 4.44 kV and second stage reflectron potential was 6.6 kV. The residual gas pressure was 9 x IW'mBar. The broad peak on the left edge is due to fragment ions reflected in the first stage.

The metastable fragmentation patterns of Tyr-bombesin (MW 1668.8 Da) and a synthetic peptide (MW 1084.6) are summarized in Figs 3 and 4. The observed fragment ions are labelled according to the nomenclature of Roepstorff and Fohlmann'" and of Johnson et al." Several properties of the spectra and the observed fragmentation pattern are important to note: (i) Almost all of the observed peaks were directly interpretable as expected fragmentations. (ii) Almost all of the expected fragmentations were observed. (iii) Side chain fragmentations ("D", "V", "W"), which are usually rather weak with other methods of peptide fragmentation, are quite abundant. (iv) Intense metastable-ion signals are observed while the abundance of parent molecular ions is still high. (v) The degree of fragmentation is dependent on the instrumental parameters (residual gas pressure, acceleration voltage, matrix-to-analyte ratio, laser irradiance. . .). Additional information on weaker or stronger bonds can thus be achieved by varying the amount of fragmentation. (vi) A fairly good mass resolution has been observed for the metastable ions. With parent molecular mass in the range of 1000 to 2000 Da, the accuracy of metastable ion mass determination was usually better than f l u. (vii) A complete fragmentation pattern of an unknown peptide can be obtained very rapidly. Using automated mass calculation, the total analysis time is in the range of a few min.

PEPTIDE SEQUENCING B Y MATRIX-ASSISTED LASER-DESORPTION MASS SPECTROMETRY

......

.............................

......

......

:.....

....

.

iw

.....

......

:.__._

iV11

-

.........

4...

....

.........

...

....

.........

%

x3 ....

Jy, ...

I

...

yo .... I/. . . . . . . . . . . 8. . . . . . . . . . . . .;w6 :.....

:...... I

....

Li

.........................

:.....

...

.

.

.

107

.....

$

;/ ......... ;/

I

....

i

H-Gly iAla Lys Ala V a l . Gly Glu Ala LYS Ala Ala Leu- OH i

a4j

......j

...........

.....;i B...9;' ...B5;i

;

.......

~~i

'

-1

......j

..........

.....j

/

.................

c4;

i

V7

: strong

VT

: medium

.......

.C$

..... .:

Di ...si

6.;

......j

AS

a 9....

.....

B7;

..................

c; c ; c 6: ...7: ..8. 1; D ; /

:

1;

i

:

a i

a3

...........

1.:

.....

...

11; ......

B ;

List of interferences:

...... I1

b?..

..... 10

....

.....

C....... ll

D9 ....

.....

I

......

...

.....!

A !

a 10 .....

/ :

:weak "/" : Fragmentation impossible v7

Figure 3. Fragmentation pattern of a synthetic peptide (MW 1084.6) observed by metastable decay of UV-laser-desorbed ions. = 337 nm, matrix: sinapinic acid. Nomenclature according to Refs 10 and 11. The 'list of interferences' names pairs of fragment ions having identical nominal masses ..............................................................................

1..... :..... ;..... :

i.....

i.....

i.....

:

;.....

.....

;

%1

iT.2

1T.l

;y;;. 1 ;. . . . . . . .iy;. . . . . . .iyr .................................... ..................... jXl0 ; w, g7 /% j ; 1x3 jX2 ;XI ............ .......................................... . . . . . . , . . . .. . . . . . . , . . . . . :/ ;/ j i/ ; ia"s ; .............................................................................. ;/ !V$ ;v7 ; ;v5 ;v4 ;v3 :vz .....

;y;' ;3f