.. E4.

2
t .
$.35“?

v.0 .. ....
2.. I

:. ch. [.2 h. 315

h.
. EZKI
:1. :7;

u... \.

 

tunklmoqdacfifi+ Ht a V

3.9.1!»

,H 9.591“? ‘

inufl a...
r: 'v Ila-‘3.

.\. 3.1.."le

3.03.: J u...

THERrg‘;

2L

 

NERSITY LIBRARI

" Illliliilllill llllll l" lllllll L

ll
3 1293 01688 5299

 

 

 

 

 

 

This is to certify that the

thesis entitled

Charge Derivatization of Peptides for Subsequent
Analysis by Matrix-Assisted Laser Desorption/
Ionization Mass Spectrometry

presented by
Rui Lu

has been accepted towards fulfillment
of the requirements for

Master Science

degree in

 

 

fag We

John Allison

 

Major professor

Date July 20, 1998

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State

3 Unlverslty

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
To AVOID FINES return on or before date due.

 

DATE DUE | DATE DUE DATE DUE

 

 

;» L13
1}? {83$

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

188 c/CIWpGG-p.“

CHARGE DERIVATIZATION OF PEPTIDES FOR SUBSEQUENT ANALYSIS
BY MATRIX-ASSISTED LASER DESORPTION/IONIZATION MASS
SPECTROMETRY

By

Rui Lu

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Chemistry

1998

ABSTRACT

CHARGE DERIVATIZATION OF PEPTIDES FOR SUBSEQUENT ANALYSIS
BY MATRIX-ASSISTED LASER DESORPTION/IONIZATION MASS
SPECTROMETRY

By

Rui Lu

Our research goals are to analyze polypeptides using 2D-polyacrylamide gel
electrophoresis with matrix-assisted laser desorption/ionization (MALDI) mass
spectrometry. After a mixture of polypeptides is separated, analytes are frequently
digested and electroblotted, resulting in complex mixtures in specific locations on a
membrane. If the membrane can be prepared for direct MALDI analysis, all components
of the mixture are usually not detected, certainly not detected with equal response.
Charge derivatization, attaching a fixed positive charge to the N-terminus of peptides, can
increase the number of components detected and can provide useful MS/MS spectra. We
are investigating formation of charged-derivatives of peptides on membranes. Our
approach is to use liquid phase or gas phase reagents to react with peptides on
membranes to form charged-derivatives. With gas phase reagents, without solvents, spots
on membranes or gels will not spread. The gas phase reagents react with single peptides,
and mixtures of up to ten peptides on a polyvinylidene difluoride membrane (PVDF). The
trimethylamine acetyl derivatives formed have increased response in MALDI. This
charge derivation method can be used to detect peptides for which a peak is not formed,

or peptides whose concentrations are too low to be accurately detected by MALDI.

Dedicated to my family

iii

ACKNOWLEDGEMENT

I still think that I am very lucky, because my first advisor and mentor in the USA. is
Professor John Allison. His courage and passion for science greatly influenced me, and
helped me over come difficulties. Many times, my spirit was ignited by his idea, and
dashed to the fulfillment. Thank you, Dr. Allison, for sharing your scientific knowledge
and experience with me. What I learned from you, like the roots, which fasten the tree on
the ground, makes me feel strong and stabilized. I will continuously get nutrition from it

in my whole life.

This thesis would not have been possible without the support of the Michigan State
University, Mass Spectrometry facility. I would like to express my appreciation for
Professor Douglas Gage, Zhi-heng Huang and other people in the facility for helping me

to operate the mass spectrometer and providing a good working environment.

I am grateful to my committee members, Professor John Allison, Professor Gary
Blanchard, Professor Robert E. Maleczka, and Professor Michael Rathke, for their

guidance through my graduate career.

Also, I would like to thank John Strahler, Gary Lavine, John Asara, Marjorie Crouse and
Niambi Wills for their friendship and valuable opinions given to me on my seminar

practices.

iv

TABLE OF CONTENTS

page
LIST OF FIGURES vii
LIST OF SCHEMES ‘ xi
LIST OF ABBREVIATIONS xii
CHAPTER ONE. INTRODUCTION 1
References 4
CHAPTER TWO. INTRODUCTION TO MATRIX-ASSISTED LASER
DESORPTION/IONIZATION TIME-OF-FLIGHT (MALDI-TOF) MASS
SPECTROMETRY 5
Introduction 5
Operation Principles 6
Sample Preparation 13
Characteristics of MALDI-MS Spectra 14
Post-source Decay 15
References 18
CHAPTER THREE. CHARGE DERIVATIZATION OF PEPTIDES 21
Introduction 21
N-terminal Charged Derivatives 31
C-terminal Charged Derivatives 41
Research Goal 43
References 45
CHAPTER FOUR. LIQUID PHASE CHARGE DERIVATIZATION 48
Derivatization of Peptides with Betaine Hydrochloride and Dicyclohexyl
Carbodiimide 48
Introduction 48
Experimental 51
Materials 5 1
Instrument 5 1
Sample preparation 52
Charge derivatization 53
Results and Discussion 53

Derivatization of Peptides with Betaine Hydrochloride and 1-(3-

dimethylaminopropyl)-3-ethyl carbodiimide and N-hydroxysuccinimide 59
Introduction 59
Experimental 61

Materials 61

Sample preparation
Charge derivatization
Result and Discussion

Derivatization of Peptides with Trimethylamine and Bromoacetic acid N-

hydroxysuccinimide ester
Introduction
Experimental
Materials
Sample preparation
Charge derivatization
Results and Discussion
References

CHAPTER FIVE. GAS PHASE CHARGE DERIVATIZATION

Derivatization of Peptides with Chloroacetyl Chloride and Trimethylamine on a

Membrane
Introduction
Experimental
Materials
Instrument
Sample preparation
Charge derivatization
Results and Discussion
References

vi

62
62
63

65
65
66
66
67
67
68
72

73

73
73
77
77
78
79
79
83
97

Figure 2.1

Figure 2.2(a)

Figure 2.2(b)

Figure 2.3

Figure 2.4

Figure 3.1
Figure 3.2

Figure 3.3

Figure 3.4
Figure 3.5

Figure 3.6

Figure 4.1

Figure 4.2

Figure 4.3

LIST OF FIGURES

MALDI-TOF mass spectrometer (linear mode).

An illustration of the initial velocity distribution of the ions in the
ion source causing the distribution on detector.

An illustration of ions with same m/z but formed at different
times in the ion source, arriving at the detector at different times.

Diagram of a MALDI-TOF mass spectrometer with a reflector.

An illustration of a PSD spectrum formed by relectron MALDI-
TOF. At high reflector voltage (11;), the precursor ions are
focused on the detector; at lower reflector voltage (uz, u3),
fragment ions are focused on the detector.

Structure of N-terminal and C-terminal peptide fragment ions.
Alternative structure of an, bn, cn and yn ions.

Charge-induced fragmentation (CIF) and charge-remote
fragmentation (CRF) mechanisms.

Structures of the fragment ions of charge-derivatized peptides.
Structures of charged groups in N-terminal derivatization.

Picture of a gas phase N-terminal charge derivatization reaction
vessel. (taken from reference [46])

MALDI-MS spectrum of trimethylamine acetyl charged-
derivative EB-L-Al (m/z 1167.4) of allatostatin 11 (Al, FW
1067.2), peptide coupling reaction product [Al+Al-H20+H]+ (m/z
2116.8) and its trimethylamine acetyl charged-derivative GB-L-
(Al+Al-H20) (m/z 2216.8).

MALDI-MS spectrum of allatostatin II protonated peak [Al+H]+
at m/z 1068.0.

MALDI-MS spectrum of a two-peptide mixture: methionine
enkephalin (Me, FW 877.0) and Tyr-adrenocorticotropic
hormone fragment 4-9 (Ty, FW 1068.2), protonated peaks

vii

10

ll

17

23

25

25

27

32

35

55

55

56

Figure 4.4

Figure 4.5

Figure 4.6

Figure 4.7

Figure 4.8

Figure 4.9

Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

[Me+H]+ (m/z 878.5) and [Ty+H]+ (m/z 1069.3), their derivative
peaks at m/z 978.5 and 1168.4 respectively. In the m/z 2000
range, there are peaks related to the coupling between peptide
molecules.

MALDI-MS spectrum of the mixture of betaine hydrochloride
and DCC.

MALDI-MS spectrum of hydrolyzed derivative [EB-L-AHHZO]
(m/z 1184.9) of allatostatin 11 (Al, FW 1067.2).

MALDI-MS spectrum of trimethylamine actyl derivative of
allatostatin II EB-L-Al (m/z 1167.5), and the peak of protonated
allatostatin II [A1+H]*(m/z 1068.5).

MALDI-MS spectrum of protonated bromoactyl allatostatin II
[Br-’9-L-A1+H]Jr (In/21189.5) and its isotopic peak [Brsl-L-A1+H]+
at m/z 1191.5.

MALDI-MS spectrum of the derivatives of methionine
enkephalin-Arg-Phe (Me, FW 877.0), Tyr-adrenocorticotropic
hormone fragment 4-9 ( Ty, FW 1068.2), at m/z 977.6 (e-L-Me)
and m/z 1168.8 (GB-L-Ty) respectively.

MALDI-MS spectrum of charged-derivatives of peptide mixture
on membrane.

Illustration of protein separation on gel followed by
electroblotting onto a membrane; separated proteins digested on
gel and then electroblotted onto a membrane.

A glass reaction vessel with two side arms which contains
membrane and reagents, is connected to a vacuum pump.

Another vessel for the reaction, the membrane was fixed on a
mold and connected with a reagent container. The reagent
molecules collided with the peptide molecules stabilized on the
membrane when the reagent vapor diffused out.

MALDI-MS spectrum of a four peptide mixture of Tyr-Ala-Gly-
Phe-Leu-Arg (TA, F W 725.8), methionine enkephalin-Arg-Phe
(Me, FW 877.0), allatostatin 11 (Al, FW 1067.2), and des-Asp-
angiotensin (AA, FW 1181.4), deposited on PVDF. Peaks at m/z
726.3, 878.0, 1068.2, and 1182.3 represent the four protonated
peptide species respectively.

MALDI-MS spectrum of chloroacetyl derivatives of allatostatin

viii

57

64

69

69

70

71

76

80

82

83

84

Figure 5.6

Figure 5.7

Figure 5.8(a)

Figure 5.8(b)

Figure 5.8(c)

Figure 5.9(a)

Figure 5.9(b)

Figure 5.10

11 (Al, FW 1067.2), neoromedin U-8 (Ne, FW 1111.3), and des-
Asp-angiotensin (AA, FW 1181.4). The m/z values shifi up by 76
for each peptide. ~

MALDI-MS spectrum of allatostatin 11 (A1, FW 1067.2) and des-
Asp-angiotensin (AA, FW 1181.4) derivatization products. The
derivatization reaction was completed after 2 hours of reaction
with chloroacetyl chloride and 2 hours with trimethylamine in the
gas phase.

MALDI-MS spectrum of peptide mixture derivatization products
on a membrane. Afier a 20-minute reaction, with first
chloroacetyl chloride, then trimethylamine, the 50% yield of the
reaction was obtained.

(a). In MALDI, 500 fmol of a peptide on a membrane is hardly
detectable.

(b). MALDI-MS spectrum of the first step derivatization product,
which is 76 m/z units higher than the original peptide, indicating
the presence of the low amount of peptide on the membrane.

(c). MALDI-MS spectrum of the two step derivatization product,
which is 100 m/z unit higher in the spectrum from the m/z value
of peptide.

(3). 50 pmol VGVAPG (FW 498.9) on a membrane is not
detectable in MALDI-MS.

(b). MALDI-MS spectrum of charged-derivative of VGVAPG.
Charge derivatization enhances the response of VGVAPG (F W

498.9), which has no response as an underivatized peptide in
MALDI-MS.

MALDI-MS spectrum of dirnethylamineacetyl derivative of
neoromedin U-8. Derivatization with chloroacetyl chloride and
dimethylamine reagent vapors of neoromedin U-8 (Ne, m/z
1111.3) produces the charged-derivative (63'-L-Ne), which in the
MALDI-MS is at m/z 1197.3.

ix

85

86

88

88

88

89

89

90

Figure 5.11

Figure 5.12

Figure 5.13

Figure 5.14(a)

Figure 5.14(b)

MALDI-MS spectrum of the result of an unsuccessful
neoromedin U-8 (Ne, F W 1111.3)derivatization with chloroacetyl

chloride and cyclohexylamine reagent vapors, only chloroacetyl
derivative [Cl-L-Ne+H]+ is observed at m/z 1187.1.

Post-source decay spectrum of trimethylamine acetyl neoromedin
U-8. The base peak is trimethylamine acetyl neoromedin U-8, at
m/z 1211. The peak at m/z 1152 is from the fragmentation and
loss of trimethylamine group from 1211.

Post-source decay spectrum of dimethylamine acetyl neoromedin
U-8. The base peak is dimethylamine acetyl neoromedin U-8, at
m/z 1197. The peak at m/z 1152 is from the fragmentation and
loss of dimethylamine group from 1197.

Electron micrograph of PVDF membrane obtained at a x4800
magnification. (taken from reference [13])

Electron micrograph of PVDF after deposition of sample and

matrix solution. The matrix crystals are embedded into the »

membrane surface. (taken from reference [13])

91

92

93

96

96

Scheme 3.1

Scheme 3.2

Scheme 3.3

Scheme 3.4

Scheme 3.5

Scheme 4.1

LIST OF SCHEMES

Mechanism of amide hydrogen shift forming the an ions.

(a). Formation mechanism of b“ ions; (b). formation
mechanism of c.“ ions; (c). formation mechanism of (In ions,
when the side chain is aspartic acid.

Formation mechanism of fragment ions from 4-NMPA
derivatives by loss of charged moiety.

Mechanism of TMPP+-Ac charge derivatization.

Mechanisms of some C-terminal charge derivatizations.

Structure of the intermediate in the active ester reaction.

xi

29

38

40

42

63

CI

CID

CIF

CRF
C-terminus
DCC
DMAA
EDC

ESI

FAB

HPLC

GB-L-M
63'-L-M

m/z
MALDI-MS
MES
[M+H]+
MS/MS

NHS

LIST OF ABBREVIATIONS

Chemical ionization

Collision-induced dissociation

Charge-induced fragmentation

Charge-remote fragmentation
Carboxyl-terminus of peptides
dicyclohexylcarbodiirnide
dimethylalkylammonimn

1 -(3-dimethylaminopropyl)-3-ethy1carbodiimide
Electrospray ionization

Fast-atom bombardment

High-performance liquid chromatography
Kinetic energy

Linker (-CH2CO-)

Trimethylamine acetyl charged-derivative of the molecule

Dimethylamine acetyl charged-derivative of the molecule

Mass-to-charge ratio value

Matrix-assisted laser desorption/ ionization mass spectrometry
2-(N-morpholino)ethane sulfonic acid

Protonated molecule

Tandem mass spectrometry

N-hydroxysuccinimide

xii

Sulfo—NHS
NMPA
N-terminus
PE

PSD

PU

PVDF
SDS-PAGE
TMAA
TMPB
TMPP+-Ac

TOF

N-hydroxysulfosuccinimide

N-methylpyridinium acetyl

Amino-terminus of peptides

Polyethylene

Post-source decay

Polyurethane

Polyvinylidene difluoride

Sodium dodecyl sulphate polyacrylamide gel electrophoresis
Trimethylamino acetyl

trimethylpyridiniumbutyl
Tris-(2,4,6-trimethoxyphenyl)phosphonium acetyl

Time—of-flight

xiii

CHAPTER ONE. INTRODUCTION

In the early days of mass spectrometry analysis, mass spectrometry was principally used
to determine the exact masses and relative abundances of the elements and their isotopes
[1-2]. In the 19405 and 19503, mass spectrometry analysis was directed to organic
compound identification. The available ionization methods mainly electron ionization
(EI), chemical ionization (CI) [3] and field ionization (F1) [4], require thermal
evaporation of the analyte, thereby limiting mass spectrometry to the analysis of low
molecular weight and thermally stable compounds. In the last two decades, the
characterization of peptides, proteins and other biomolecules encountered in
biotechnology launched the application of mass spectrometry to large and often thermally
liable bioorganic molecules. The extension of mass spectrometry for the identification
and structural determination of large biomolecules became possible with the development
of various desorption ionization methods [5-8], which are capable of producing ions

directly from the condensed phase.

Pulses of laser light have been employed in an ion source to produce intact gas phase
peptide ions from solid samples [7]. However, the mass spectral quality depended
critically on the specific physical properties of the biological compound (e.g.
photoabsorption, volatility) under study [9-10]. This situation is changed dramatically
with the development of matrix-assisted laser desorption/ionization (MALDI) in 1988 by

Karas and Hillenkamp [11].

MALDI-MS provides the means to volatilize peptides and proteins readily and to make
the condition for volatilization largely independent of the physical properties of the
sample molecules. In this technique, the samples are prepared by cocrystallization of the
analyte with a small molecular weight, UV absorbing organic matrix, on a sample probe
target. Ionization is triggered by a UV laser. Mass analysis is done by a time-of-flight

(TOF) instrument.

MALDI-MS TOF is versatile and effective for the analysis of peptides and proteins
because of the properties and capabilities of the technique. Chapter two will discuss the
instrumentation of a MALDI-MS TOF experiment, introducing the reader to MALDI
operating principles, sample preparation, functionality of the reflectron, and post-source

decay techniques for peptide structural analysis.

In protein analysis, after a mixture of polypeptides is separated, analytes are frequently
digested and electroblotted, resulting in complex mixtures in specific locations on a
membrane. If the membrane can be prepared for direct MALDI analysis, all components
of the mixture are usually not detected, certainly not detected with equal response.
Charge derivatization, attaching a fixed positive charge to the N-terminus of peptides, can
increase the number of components detected and can provide useful MS/MS spectra [12-
14]. Chapter three will evaluate the charge derivatization methods that have been

explored.

Although charged-derivatives offer attractive advantages in the mass spectrometry
analysis of proteins, additional work is still required in this field. The goal of our
approach is to develop a method, which improves the sensitivity of peptide analysis by
MALDI, and provides structure information of peptides by MALDI-PSD, which is easy
and fast enough to be used in routine analytical work and which does not cause mass
spectral interference by non-volatile solvents or reagents. In chapters four and five, we
will summarize the results of research regarding charge derivatization methods and their

usefulness in MALDI-MS analysis.

References

8.

9.

. K.L. Busch, G.L. Glish, S.A. Mcluckey, Mass Spectrometry/Mass Spectrometry

Techniques and Applications of Tandem Mass Spectrometry, VCH: New York, 1988

DA. Skoog, Principles of Instrumental Analysis, 3rd ed., Saunders: Philadelphia,
1985

M.S.B. Munson, F.H. Field, .1. Am. Chem. Soc, 88, 2621 (1966)
E.W. Muller, K. Bahadur, Phys. Rev., 102, 624 (1956)

L. Pokai, Field Desorption Mass Spectrometry, Marcel Dekker: New York, 1990
Vol.2

D.F. Torgerson, R.P. Showronski, R.D. MacFarlane, Biochem Biophys. Res.
Commun.,60,6l6(l974)

M.A. Posthumus, P.G. Kistemaker, H.L.C. Meuselaar, M.C. Ten Noever de Brauw,
Anal. Chem, 50, 985 (1978)

A. Benninghoven, WK. Sichterrnann, Anal. Chem, 50, 1180 (1978)

M. Karas, D. Bachmann, F. Hillenkamp, Anal. Chem, 57, 2935 (1985)

10. B. Spengler, M. Karas, U. Bahr, F. Hillenkamp, J. Phys. Chem, 91, 6502 (1987)

l 1. F. Hillenkamp, M. Karas, R.C. Beavis, B.F. Chait, Anal. Chem, 63, 1193A (1991)

12. G.C. Didonato, K.L. Busch, Biomed. Mass Spectrom, 12, 364 (1985)

13. M. Bartlet-Jones, W.A. Jeffery, H.F. Hansen, D.J.C. Pappin, Rapid Commun. Mass

Spectrom, 8, 737 (1994)

14. Z.-H. Huang, J. Wu, K.D.W. Roth, Y. Yang, D.A. Gage, J.T. Watson, Anal. Chem,

69, 137 (1997)

CHAPTER TWO. INTRODUCTION TO MATRIX-ASSISTED LASER
DESORPTION/IONIZATION TIME-OF-FLIGHT (MALDI-TOF) MASS

SPECTROMETRY

Introduction

Before the introduction of matrix-assisted laser desorption/ionization (MALDI) [1-2] and
electrospray ionization (ESI) [3-5] techniques, mass spectrometric analysis of organic
compounds was limited to small, volatile sample molecules. With the extension of mass
spectrometry to large nonvolatile and ofien thermally fragile biomolecules, research
during the 19708 and 19805 has been directed to the development of ionization methods
capable of ionizing such molecules directly from the solid or liquid state. This work led
to the development of MALDI and E81 techniques [6]. With these two new techniques,
today, routine mass spectrometric analyses of proteins in the interesting mass range
between 10 and lOOkDa with only picomolar amounts of sample have become possible.
Due to the advantages and limitations of these two techniques, they both contribute to

establish mass spectrometry as a tool in fields like molecular biology and biomedicine.

There are several features that make MALDI an attractive analytical technique for
biomolecules. First of all, MALDI can be used to identify molecules in a broad mass
range. The accessible mass range for MALDI has been extended to more than 300,000
Daltons [7]. MALDI has been successfully applied to peptides, proteins, olignucleotides

[8] and oligosaccharides [9]. Secondly, MALDI spectra have high resolution. A mass

accuracy of 0.01% for proteins up to 30,000 Daltons has been reported [7]. MALDI has
an outstanding sensitivity. Typically no more than 1 picomole of analyte is the total
amount of sample required and used for analysis. The absolute sample consumption for
recording the mass spectra can be estimated to be in the low attomole (10'18 mole) range
[10], thus virtually the whole sample can be recovered afler analysis. The ability to
examine complex mixtures and tolerance to contaminations are other reasons of more and
more interest in MALDI. Although singly charged molecular ions dominate the mass
spectra, MALDI has the ability to elucidate structural information of the molecules under

investigation, by analyzing fragment ions [1 1-12].

Operation principles

Many types of lasers have been used to generate ions for mass spectrometric analysis [13-
16]. Two main types of lasers employed are CO; lasers emitting light at 10.6um in the
infrared (IR), and N2 ultraviolet (UV) lasers at 337nm. In order to reduce thermal
decomposition of labile biomolecules, the use of short pulse lasers (typically 1-100
nanoseconds) is essential. Photon beams generated by pulsed lasers have good spatial
resolution, focused down to spots of submicrometer diameters by suitable microscopic
objectives [17]. Direct laser- desorption of intact biomolecules without a matrix seems to
be limited to molecular masses of about 1000 Daltons [14,18-20]. Too high of an
irradiance results in extensive fragmentation, with increasing and finally prominent
formation of unspecific low mass ions. These limitations promoted a search for highly

UV- absorbing organic compounds to be used as a matrix.

The matrix can serve several functions to enhance desorption of large molecules. The
matrix is present in large excess relative to the analyte (typically 1000:1). In the matrix,
which is like a solvent for analyte molecules, analyte molecules are solvated, and
separated from each other. Since the analyte molecules are separated from each other, the
aggregation, which prevents molecular ion formation, is limited. As matrix desorption is
usually performed from solid material, the ability of a matrix to form a “solid solution” of
analyte molecules is essential. The matrix absorbs laser energy and transfers it into
excitation energy of the solid system. Top molecular layers of the sample then will
experience a rapid phase change. Beavis and Chait proposed that, as a result of the
volume excitation, a dense gas is formed, which expands supersonically into the vacuum
[21]. The matrix also plays an active role in the ionization process of analyte molecules
[22,23]. The analyte ionization is assumed to take place via matrix/analyte collisions and

charge transfer within the gas phase plume [24-26].

Ions formed by pulsed laser desorption will be analyzed by a time-of-flight (TOF) mass

analyzer. An illustration of a linear MALDI-TOF is shown in Figure 2.1.

After ionization, there are two major steps in a time-of-flight analysis: acceleration and
the flight time measurement itself. The TOF mass analyzer is conceptually the simplest of
the mass analyzers in common use. The measurement is based on the fact that if ions of
different masses leave the ion source simultaneously, they will take different times to
reach the detector [27]. The flight time of an ion with mass, m, formed in ion source with

kinetic energy, KE, a distance from the source to the detector, (1, is calculated as follows:

- __ __ Linear detector

 

 

 

 

 

 

 

- ......... a’
A Ion path
............................. ’
Laser path
Laser — Flight tube

' Video camera

/Aperture (grounded)

"~
\.
,
“x.

Ion source (grid) — ——-**~-- -- -

 

 

 

 

 

 

 

|
Sample
Ion source / / loading
(sample plate) I chamber

Main source chamber

Figure. 2.1. MALDI-TOF mass spectrometer (linear mode).

KE = 1/2rnv2
v = (21(13/rrr)"2
t = d/v = d(m/(2KE))“2 (1)
A TOF mass analyzer is therefore used with sources like those that use pulsed-lasers, and
requires the use of a collector with very good time resolution. Compared with other types
of mass spectrometers, TOF has several advantages besides its simplicity and robustness
[28]. The whole mass spectrum is obtained in every single cycle of the measurement
without the need to scan any voltages or currents. Furthermore, because of the few ion
optical elements through which the ions must pass, the transmission from the point of the
ionization to the detector is usually higher than in other mass spectrometers. However,
there is a severe problem associated with this technique - low resolution. This problem is
created by the initial velocity of the ions in the ion source and the duration of ionization.
As shown in Figure 2.2(a), and Figure 2.2(b), initial ion velocities [29] and the duration
of ionization are sources of broadening of the flight time distribution. In addition, when
desorption occurs in a strong electric field, energy is presumably lost by collisions within

the neutral plume, resulting in further mass-dependent energy dispersion [30].

The introduction of reflectrons and delayed-extraction led to the impressive renaissance
of TOF mass spectrometry. In 1973, Mamyrin suggested a compensation of the flight
time differences of ions with an electric deceleration field and reversing their flight
direction before detecting them [31]. Figure 2.3 presents a diagram of a reflectron TOF
mass spectrometer. The reflectron is also called an ion mirror. Ions with high energies

penetrate deeper into the reflectron field, and spend more time there than low energy

a) Initial ion velocity v

1:-

v-dv

6*

 

 

 

E1 ' E2 Detector

b) Ion formation time t

H

I.--

H
+
Q-
o-o

O—-—* O——>

 

O

 

El . 132 Detector

Figure. 2.2(a). An illustration of the initial velocity distribution of the ions in the ion
source causing the distribution on detector. 2.2(b). An illustration of ions with the same
m/z but formed at different times in the ion source, arriving at the detector at different
times.

Linear detector

 

 

.......... .}
Ion path in linear
or reflectron mode

Laser path

  

 

 

 

Reflectron detector

Reflector p.‘ A

(electrostatic i ‘-._ _;' '.

mirror) -' '1‘ ;'
Timed ion I “-J
selector

Laser — Flight tube

' Video camera

—\ /—' / Aperture (grounded)

 

 

 

 

/

\u

._____._1.._— .I-

Ion source (grid) — \ Sanllple

Ion source / / loading

(sample plate) chamber

 

 

 

 

Main source chamber

Figure. 2.3. Diagram of a MALDI-TOF mass spectrometer with a reflector.

ions. The retarding field of the reflector is adjusted to compensate for the different flight
times of the ions in the drift regions. Ions with sarrre mass will hit the detector at the same
time. The resolution of MALDI-MS spectra has been improved from hundreds to more

than 10,000 for biomolecule analytes [32-34].

More recently, the development of delayed-extraction dramatically improves the
resolution achievable in MALDI-TOF mass spectrometry. In the conventional MALDI-
TOF instrument, ions are generated by the laser beam near the surface of the sample plate
and are continuously extracted by a dc potential. In delayed-extraction, a short time delay
(less than 300 nanoseconds) is applied between laser desorption/ionization and ion
acceleration events [35]. During the delay time, the ions which have the greater initial
kinetic energy will travel further into the ion source than those with lower kinetic energy.
When the extraction field is turned on, the ions having the lower kinetic energy will
traverse a longer distance in the electric field, thereby acquiring more kinetic energy in a
compensatory effort to give all ions of the same m/z the same flight time. In principle,
delayed-extraction allows ions of the same mass to arrive at the detector at the same time,
giving rise to a sharp peak or greater resolution. Some improvement in MALDI spectral
quality with delayed-extraction may also be explained by dissipation of the very high

local pressure (tens of atmospheres) in the laser plume following the laser pulse [29, 30].

12

Sample preparation

Proper sample preparation is critical for successful analysis by MALDI-MS [36, 37]. A
saturated matrix solution, usually in water and acetonitrile, at a concentration of 5-100
mg/uL depending on the solubility of the matrix, is used. The analyte is prepared in a
solvent, which is miscible with the matrix solution, at the concentration of about 10-

lpmol/uL. The matrix and analyte solutions are mixed to achieve a final matrixzanalyte
molar ratio in the range of 100:1 to 10000:]. An aliquot of 1 to 2 uL of the mixture will

then be deposited on a metal sample plate. During the drying process, the matrix co-

crystallizes with the analyte from the solution.

The nature of matrix-analyte interactions in the mixture is not well understood. Studies
show that small molecules such as amino acids and peptides will be embedded in the
matrix crystal lattice while larger molecules are distributed preferentially on the crystal
surface [3 8]. Matrix crystal formation and size varies depending on the choice of matrix
and solvent. In general, it has been observed that a decrease in crystal morphology quality
is attributed to a lower quality MALDI- MS spectrum [39]. High salt concentrations in
the mixture may induce the decreasing quality of the crystal formation. HPLC or simply
washing with a drop of water may need to be carried out to desalt the sample [40].
Alternatively, solution dilution, which can reduce the interference from a salt, may also

lead to improved spectra.

13

The sample can be introduced into a MALDI spectrometer on various support surfaces.
The mainly used surface is metal. Other than that, inert polymeric supports such as
membranes have also been employed for picomole sample handling in protein mass
spectrometric analysis [41-43]. There is considerable interest now in the direct analysis
by MALDI-MS of biological samples that have been adsorbed onto a transfer membrane
because of implications in coupling gel electrophoresis with MALDI-MS. After proteins
are separated in a gel, they are electroblotted quantitatively onto a membrane; their
relative positions are not changed. Protein digestions often are applied to proteins in a gel
before electroblotting. Membranes with separated proteins or peptide spots will then be
washed with distilled water to remove contarninations before adding matrix and
introduction to the MALDI mass spectrometer. Various membranes, such as
polyvinylidene difuoride (PVDF) membrane [44, 45], polyurethane (PU) membrane
[46], polyethylene (PE) membrane [42] and nitrocellulose [43,47] have been reported to
be suitable for analyzing picomole levels of a protein that had been immobilized on the

membrane.

Characteristics of MALDI-MS spectra

Peaks representing the intact protonated molecules in the positive ion mode dominate
MALDI-MS spectra. Na+, KL and matrix adduct peaks are also observed. Multiply
charged ions are observed but at relatively lower abundance. Peaks corresponding to
fragmentation are rarely observed. The lack of fragmentation makes this technique ideal
for mixture analysis [48-50]. Under proper laser irradiance most of the signals in the

lower mass range can be assigned to matrix ions. The mass resolution of the ion peaks

l4

can be estimated by m/Am (Am is the width at the half-height of the peak). Besides the

spread of initial ion energies in the ionization process, unresolved adduct ion peaks on the
high mass side and peaks due to cleavage of small groups on the low mass side, are also
believed to contribute to low resolution for some peaks. Depending on the attained
signal-to-noise ratio, the precision of mass deterrrrination for proteins is typically in the
range of 10-50 mass units at 10,000 Daltons and 100-500 mass units for a protein of

100,000 Daltons in mass.

Post-source decay

The MALDI technique is powerful in determining molecular masses of large
biomolecules. However, in order to identify and characterize the biomolecules, structural
information is also desired. In linear MALDI-MS spectra, relatively little or no
fragmentation was observed. Studies using reflectron MALDI-TOF mass spectrometers
revealed that MALDI ions undergo post-source fragmentation [51-52], but the fragments
are not separated from the protonated molecules by a linear TOF mass analyzer, because
they have approximately the same velocity. All fragment ions from post-source decay are
detected at the end of the field free region, as an increase in the peak width of the intact
protonated molecule. A fragment ion has a kinetic energy KEf as in (2):
KEr=iMr/MP)KEP (2)
In equation (2), Mr and M, are masses of fragment and precursor ions respectively, and

KEp is the kinetic energy of the precursor ion.

15

Since these fragment ions and precursor ions have different kinetic energies, a reflector
can separate them. As mentioned earlier in this chapter, in a reflectron-TOF instrument,
at the end of the flight tube, an ion mirror is placed, whose electric field forces the ion in
the reverse direction of their initial movement. As illustrated in Figure 2.4, by scanning
the reflector voltage, starting at the optimum value for precursor ions down to zero, a
particular ion of a certain mass with a certain KE will be focused onto the detector. The
entire mass range of PSD fragment ions for a particular precursor ion can be detected by
lowering the reflector potential in several intervals. Typically the ratio of the mirror
voltage to the acceleration voltage is lowered in steps, in a way that each ratio is about
75% of the preceding ratio. The observable fragment ion mass is proportional to the
mirror ratio times the precursor ion mass, as shown in equation (3):

Mf = (KEp’KEp) Mp = mirror ratio x Mp (3)
Basically, the reflectron is tuned to sequentially focus the lower energy fragments and a
series of related spectra are obtained. A program then combines those spectra into a

single spectrum showing both the fragment ions and their precursor ion.

16

Reflector Signal

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

a)
b) EN
1 1 U2 M1
M1 l l l l
c) E E 5 5

 

0» A A I.

Figure. 2.4. An illustration of a PSD spectrum formed by relectron MALDI-TOF. At high
reflector voltage (111), the precursor ions are focused on the detector; at lower reflector
voltage (uz, u3), fragment ions are focused on the detector.

 

 

 

 

l7

References

10.

ll.

12.

13.

14.

15.

l6.

l7.

l8.

. M.Karas, F. Hillenkamp, Anal. Chem, 60, 2299 (1988)

A. Overberg, M. Karas, V. Bahr, R. Kaufrnann, F. Hillenkamp, Rapid Commun. Mass
Spectrom, 4, 293 (1990)

OK. Meng, M. Mann, J.B. Fenn, Z. Phys. D, 10, 361 (1988)

J .B. Fenn, M. Mann, C.K. Meng, S.F. Wang, C. M. Whitehouse, Science, 246, 64
(1989)

R D. Smith, J .A. Loo, C.G. Edmond, C.J. Barinaga, J .R. Udseth, Anal. Chem, 62,
882 (1990)

G. Siuzdak, Proc. Natl. Acad. Sci. US. A., 91, 11290 (1994)

F. Hillenkamp, M. Karas, Proceeding of the 37th ASMS Conference on Mass
Spectrometry and Allied Topics, Miami Beach, May 21-26, 1989, 1168

B. Spengler, Y. Pan, R.J. Cotter, L. —S. Kan, Rapid Commun Mass Spectrom, 4, 99
(1990)

B. Stahl, M. Steup, M. Karas, F. Hillenkamp, Anal. Chem, 63, 1463 (1991)

M. Karas, A. Ingendoh, U. Bahr, F. Hillenkamp, Biomed. Environm. Mass Spectrom,
18, 841 (1989)

KB. Tomer, Mass Spectrom. Rev., 8, 445 (1989)
RE. Tecklenburg, D.H. Russell, Mass Spectrom Rev., 9, 405 (1990)
F.J. Vastola, R.O. Mumma, A. Pirone, J. Org. Mass. Spectrom, 3, 101 (1970)

MA. Posthumus, P.G. Kistemaker, H.L.C. Meuzelaar, M.C. deBrauw, Anal. Chem,
50, 985 (1978)

R. Stoll, F.W. Rollgen, Org. Mass. Spectrom, 14, 642 (1979)
R.J. Cotter, Anal. Chem, 53, 1306 (1981)

A.H. Verbueken, F.J. Bryunseels, R.E. Van Grieken, Biomed. Mass. Spectrom, 12,
438 (1985)

M.Karas, D. Bachmann, F. Hillenkamp, Anal. Chem, 57, 2935 (1985)

19. B. Spengler, M. Karas, U. Bahr. F. Hillenkamp, J. Phys. Chem, 91, 6502, (1987)
20. B. Spengler, M. Karas, U. Bahr, F. Hillenkamp, Anal. Instrum, 17, 173 (1988)

21. RC. Beavis, B.T. Chait, “ Method and Mechanisms for Producing Large Ions fiom
Large Molecules”, NATO ASI Science Series, Plenum, New York

22. M. Karas, F. Hillenkamp, presented at Desorption 94 Sunriver, OR, 1994

23. B. Spengler, R. Kaufrnann, U. Bokelmann, M. Huubert, D. Kirsh, presented at
Desorption 94, Sunriver, OR, 1994

24. V. Bokelmann, B. Spengler, R. Kaufinann, Eur. Mass Spectrom, 1, 81 (1995)

25. K. Dreisewerd, M. Schurenberg, M. Karas, F. Hillenkamp, Int. J. Mass Spectrom. Ion
Processes, 141, 127 (1995)

26. B.H. Wang, K. Dreisewerd, U. Bahr, M. Karas, F. Hillenkamp, J. Am. Soc. Mass
Spectrom, 4, 393 (1993)

27. R.W. Kiser, “Introduction to Mass Spectrometry and Its Applications”, Prentice-Hall,
New York, 1965

28. IF. Holland, C.G. Enke, J. Allison, J.T. Stults, J.D. Pinkston, B. Newcome, J.T.
Watson, Anal. Chem, 55, 997A (1983)

29. RS. Brown, J .J . Lennon, Anal. Chem, 67, 1998 (1995)
30. ML. Vestal, P. Juhasz, S.A. Martin, Rapid Commun. Mass Spectrom, 9, 1044 (1995)

31. BA. Mamyrin, D.V. Karataev, D.V. Shmikk, V.A. Zagulin, Sov. Phys. JETP 37, 45
(1973)

32. T. Bergmann, T.P. Martin, H.Schaber, Rev. Sci. Instrum, 60, 347 and 792 (1989)

33. U. Boesl, R. Weinkauf, E.W. Schlag, Int. J. Mass Spectrom Ion Processes, 112, 121
(1992)

34. R. Frey, G. Weiss, H.Kaminski, E.W. Schlag, Z. Naturforsch. A, 40, 1349 (1985)
35. W.C. Wiley, I.H. McCaren, Rev. Sci. Instrum, 26, 1150 (1955)
36. S.L. cohen, B.T.Chait, Anal. Chem, 68, 31 (1996)

37. O.Vorm, P. Roepstorff, M. Mann, Anal. Chem, 66, 3281 (1994)

19

38. F.M.L. Amado, P. Pomingnes, M. Gracasantana-marques, A.J. Ferrercorreia, K.B.
Tomer, Rapid, Commun. Mass Spectrom., 11, 1347 (1997)

39. SJ. Doktycz, P.J. Savickas, D.A. Krueger, Rapid Commun. Mass Spectrom., 5, 145
(1991)

40. H.Zhang, P.E. Andren, R.M. Caproli, J. Mass Spectrom., 30, 1768 (1995)

41. E]. Zaluzec, D.A. Gage, J. Allison, J .T. Watson, J. Am. Soc. Mass Spectrom., 5, 230
(1994)

42. J .A. Blackledge, A.J. Alexander, Anal. Chem, 67, 843 (1995)

43. K.K. Mock, C.W. Sutton, J.S. Cottrell, Rapid Commun. Mass Spectrom., 6, 233
(1992)

44. M.M. Vestling, C. F enselau, Anal. Chem, 66, 471 (1994)

45. M.M. Vestling, C.Fenselau, Biol. Soc. T rans., 22, 547 (1994)

46. ME. Mccomb, R.D. Oleschuk, D.M. Manley, L. Donald, A. Chow, J.D.J. O’Neil,
W.Ens, K.G. Standing, H. Perreault, Rapid Commun. Mass Spectrom., 11, 1716
(1997)

47. K. Klarskov, P. Roepstorff, Biol. Mass Spectrom., 22, 433 (1993)

48. M. Karas, U. Bahr, U. Giebmann, Mass Spectrom. Rev., 10, 335 (1991)

49. M. Karas, U. Bahr, A. Ingendoh, E. Nordhoff, B. Stahl, K. Strupat, F. Hillenkamp,
Analytica Chimica Acta, 241, 175 (1990)

50. RC. Beavis, B.T. Chait, Proc. Natl. Acad. Sci. U.S.A., 87, 6873 (1990)
51. R. Kauffman, D. Kirsch, J. Phys. Chem, 96, 9678 (1992)

52. R. Kauffman, D. Kirsch, B. Spengler, Int. J. Mass Spectrom. Ion Processes, 131, 355
(1994)

20

CHAPTER THREE. CHARGE DERIVATIZATION OF PEPTIDES

Introduction

Chemical derivatization has played an important role in mass spectrometry from the
earliest practice of the technique [1]. In the early years of mass spectrometric analysis,
the primary ionization techniques were electron ionization (EI) and chemical ionization
(CI). These mass analysis techniques have a common disadvantage. The sample analyzed
must have enough volatility in order to be analyzed. The major impetus for chemical
derivatization’s early use was to confer the necessary volatility to sample compounds. In
the 19705 and 19805, the development of desorption/ionization techniques, e.g. field
desorption [2], plasma desorption [3], and fast atom bombardment (FAB) [4], enabled the
analysis of underivatized peptides by mass spectrometry. The advent of FAB ionization
has led to the use of chemical derivatization for detectability enhancement by the
intentional introduction of a charged group into the molecule. In 1985, a study reported
that by applying Girard’s reagent T in the derivatization of the ltetosteroid androsterone,
which is undetectable under normal FAB, the quaternary ammonium derivative as shown
in (1), gave an intense peak in the mass spectrum indicating the present of the compound
[5].
o o NNHCOCH2N+(CH3)

II || ||
R-C-R' + (CH3)3N+CH2CNHNH2 -> R-C-R'

Girard’s reagent T quaternary ammonium derivative (1)

21

Later on, several types of derivatives were developed to lower the detection limits of
peptide analysis. It is assumed that derivatives with a fixed charge group increase
ionization efficiencies of the analytes during analysis by desorption/ionization

techniques.

Other than the molecular weight, which is important information in identifying
molecules, structural information from fragments is also desired. The utility of FAB with
tandem mass spectrometry, MS/MS, with fragmentation methods such as collision-
induced dissociation (CID), produces a variety of structurally significant fragment ions
[6-8]. However, the many types of fragment ions that arise during desorption/ionization
and MS/MS [8] can complicate peptide mass spectra. Peptides undergo cleavage of a
single peptide bond with charge retention at either the amino-terminal (N -terminal) or the
carboxyl-terminal (C-terminal) fragment in the CID experiment. N-terminal fragment and
C-terminal fragment ions are an, b", cn and X“, y.,, as shown in Figure 3.1. Fragmentation
at a peptide bond, followed by further cleavage of the side chain of the amino acid at that
peptide bond forms ions of types (in, vn and wn (Figure 3.1). Internal fragment ions

formed by fragmentation at two distant peptide bonds are also observed.

The structure of peptide fragment ions and the mechanisms that produce them are
represented by the nomenclature recommended by Johnson et al. [9]. The mechanism
shows that the fragmentation is charge-induced [7]. Alternative an [10], c“ [1 l] and yn [7]

structures are also supposed for peptides formed by the charge-remote fragmentation

22

N-terminal ions C—terminal ions

 

 

 

 

 

 

., i i" in i
H -( NH- CH - c0);- +NH= CH x" +oec— NH- CH- co (- NH- CH- co 7;, on
i in ‘1‘" i
b“ HTNH‘ CH- (307:1 NH- CH- (350+ y" +NH3- CH- COT NH- CH- CO t1 OH
R
r. R, in .
I I zn +CH- co (- NH- CH- co 7;,orr
°n H-( NH— CH- c0); NH— CH- co—+ NH, w
w 1'1
R CH- R' vn NH= CH- co T NH- CH- co -),,—,.ori
I II
dn H-(NH-CH-CO‘mNH-CH H,
Rf—CH If
11
W“ CH- co TNH- CH- comorr

Figure. 3.1. Structure of N-terminal and C-terminal peptide fragment ions.

23

mechanism, as shown in Figure 3.2. An alternative bn structure [12, 13], a more stable

structure than that originally proposed for bn ions, is presented in Figure 3.2.

Charge-remote fragmentation (CRF) [14, 15] can be distinguished from charge-induced
fragmentation (CIF). The principle of these two reaction types is shown in Figure 3.3. In
CIF, a mobile proton is located close to a certain bond, lowering activation energies for
fragmentation. As a result the charge is moved to one or the other part of the fragmented
peptide forming a detectable fragment ion [13,16,17]. If a mobile proton is lacking,
fragmentation reactions by the CIF mechanism are suppressed since the activation energy
is too high. In the case of CRF, no mobile proton is needed for fragmentation, but the
internal energy of the peptide is high enough to overcome the activation energy [11, 13,
16, 17]. It is assumed that in low-energy CID, the CIF mechanism is the predominant

mechanism [18, 19], while CRF was observed frequently in high-energy CID [15, 20].

The charge localization influences the type of fragmentation [21]. Peptides devoid of
basic centers (acylated N-terminus, no basic amino acids) result in simple spectra
dominated by bn and yn ions. These ions are formed via the CIF mechanism [7]. One of
the arrride nitrogens is initially protonated. This is rationalized by the assumption that the
protonating hydrogen is attached randomly to any of the equally basic amide nitrogen
atoms of the peptide backbone. This triggers cleavage to both the bn ion and release of a
neutral C-terminal peptide fragment or fragmentation to form a yn ion by transfer of a

hydrogen atom from the N-terminal moiety [22].

24

N-terminal ions _C-termina1 ions

H+ R'\ /R" H+
R C v.2 [Rn R

| u I I
1H NH_ CH- co m NH- CH NHZ- CH- COTNH- CH- com on

i re ; H R"
bi: H-(-NH-CH-CO-);2 NH-CH-C< j
0 \\

 

 

 

 

H“ O

 

i in

"3 H-(-NH- CH- com NH- CH- co- NH,

 

 

Figure. 3.2. Alternative structure of an, bn, cn and yn ions.

Charge-induced fragmentation Charge-remote fragr_n entation
WW
/\/\/\/I/\/\/\/\Jr

/\/\+\/\/\/\/\/\ /\/\\/\/\/\/\/\+

Figure. 3.3. Charge-induced fragmentation (CIF) and charge-remote fragmentation (CRF)
mechanisms.

25

In the presence of a basic amino acid such as arginine (Arg), or a less basic amino acid
such as histidine (His) and lysine (Lys), the appearance of the CID spectra changed
dramatically. This effect is attributed to the tendency of the proton to be preferentially
attached to the basic amino acid, in which case charge-remote fragmentation occurs [23].
This process competes with the protonation of one of the less basic amide nitrogen atoms,
as discussed above, and produces fragments by a mixture of charge—induced and charge-
remote fragmentation mechanisms [7]. If the basic amino acid is present at the C-
terrrrinus, mainly yn, vn, wIn and b“ ions are observed. Conversely, when a basic amino

acid is at or near the N-terminus of the peptide, an, bn and (1.1 ions are formed.

Derivatization can be used in conjunction with tandem mass spectrometry to obtain
simplified mass spectra of peptides. A charged group stabilized on either the N-terminus
or the C-terminus of the peptide can facilitate interpretation of the mass spectra. It is
assumed from the observed fragmentation pattern of the peptides in MALDI-PSD
spectra, that charged peptide derivatives form fragments through a charge-remote
fragmentation mechanism [34]. As a result, only certain types of ions are observed.
Proposed structures of the fragment ions of charge-derivatized peptides are shown in
Figure 3.4 (taken from reference [50]). Because the protonated peptides and charged
peptide derivatives differ in mass depending on the type of derivatization, an asterisk is

added to denote the presence of the charged moiety.

Charge derivatization limits the types of fragmentation of peptides. Derivatization at the

N-terminus causes N-terminal charge remote fragment ions to be observed; derivatization

26

® 15 fixed charge moiety

N-terminal ions C-terminal ions
* —— R. R.. —-——-
an R \C / R,1 R
l u x" * | |
@TNH CH_ co 7.7.1 NH, CH 0 = c = NH- CH- co (—NH- CH- com 6)
R R. l‘n ’3
* -2
GT NH_ 'CH_ co m N z CIH Yn NH= c- cor NH- CH- co m G)
R R. y . If" i
b. * GT NH- ICH- co )5 NH- i: = c = o n N”? CH‘ COT NH' CH“ C02?! Q
R“ R
When Asp is residue n I l
o 2,, * +1 oCH- co (- NH- CH- co RIG

R
l
@-(- NH- CH- co in: Nil-<1 1'1

/
R 0}; vn * NH= CH- co 19 NH- CH- co m Q
Ca " @1- NH- '01- co )— NH- |cu.- CO-NH
n-l 2 R._ CH 1}
R , II
I 04- R wn * CH- co (- NH- CH- co m G)

11
dn " @-(— NH- CH- co ),-,-, NH- CH

Figure. 3.4. Structures of the fragment ions of charge-derivatized peptides.

27

at the C-terminus causes C-terrninal fragments to be observed. A study using high energy
CID showed that CID mass spectra of N-terrninal charged-derivatives produce primarily
an", and (1,," and a few bf“ and cn“ ions[24], while C-terminal charged-derivatives give

primarily yn*-2, yn", vn" and w.“ ions [25-26].

The diversity of the fragmentation of an N-terminal charged-derivative results in various
sources of the hydrogen shifi. When amide hydrogen atoms shift, a." ions are formed as

illustrated in Scheme 3.1.

The shifi of a hydrogen atom from a side chain produces ions with a cyclic termination.
For example, when the side chain is aspartic acid, the terminal hydrogen on aspartic acid
side chain is labile. This hydrogen atom is at a position that facilitates formation of three
different fragmentation intermediates. Through these intermediates, low energy paths for
dissociation are possible. The mechanism of the side chain hydrogen shifting produces

bn*, cn-I* and (1,,“ ions as shown in Scheme 3.2.

Recently, MALDI-TOF mass analysis has received more and more attention in its
potential for biomolecule analysis. Charge derivatizations have been applied to the
analysis of peptides by MALDI [27-35]. In linear MALDI-TOF, only a peak representing
the C+ cation for such charged-derivatives was observed. However, the ions do fragment
following their generation and acceleration. Mass spectra representing these fragment
ions can be acquired in reflectron TOF experiments, and are referred to as post-source

decay (PSD) spectra [36-37]. As in high-energy collision-induced dissociation, the data

28

R2

11“ 1) CH 0
\ /
WCH—C— QCW
H

i" ll 1?
WCH- C—N=CHR2 + H—C/\/\/\/\

an

Scheme 3.1. Mechanism of amide hydrogen shift forming the an ions.

29

a).

b).

//O
R... cu,—c

W I / \O
G NH—CHwfi—NH—C< \4
H
o ,4."
CHRmI
l
o:c/\/\/\/\
O
R //
n-l CH _C
W I / ’ \
G NH— CH —c -——NH—CH
ll \ /O
O //C
O
bi.
0 O

O
0 | W
O=C- NH— CHRM

 

1"...
G /\/\/\/\NH— CH —-c __NH2

II
o

cn-l

A)

Rn-I CHI—p C

/ \
(DA/WAN“... ('3H_C __NH__CH 0
11 \m
o ,c
o ’ |
NH
C'HR...
l
o :C W
’ng
G) WNH— CH —|C| ——NH—CH = CH,
0
a,

Scheme 3.2. (a). Formation mechanism of bn ions; (b). formation mechanism of cm ions;
(c). formation mechanism of (In ions, when the side chain is aspartic acid.

30

obtained by MALDI-TOF illustrated that the charge-remote fragmentation mechanism is
the primary fragmentation mechanism in MALDI-TOF PSD [34]. The PSD spectra of
peptides are complicated with mixed types of fragment ions. It is charge derivatization
that makes PSD an effective technique for obtaining structural information, through

interpreting simplified PSD spectra.

In summary, the major advantages of charged-derivatives are that they simplify the
fragmentation of peptides. This is important for peptides with a completely unknown
sequence. Also, they improve the detectability of the peptide analyte in the mass

spectrometer.

There exists a diversity of charged-derivatives in that both positive and negative charge
derivatives have been studied [38, 39]. A charged group can be attached to the N-
terrninus or the C-terminus, or to the side chain of the peptides [7, 40, 41]. Also
attachment of a charged group to both the N-terminus and side chain of the peptides, or

both the C-terrninus and side chain of the peptides, has been observed [7, 40, 41].

N-terminal charged-derivatives

Several N-terminal charge derivatization approaches have been developed since 1984.

The structures of those charged groups in the derivatives are shown in Figure 3.5.

31

Orr-(erg), -co—

WmmybyrIqumbutyl
(TM PB)

       

N-mathyH-pyrldinkrmacem
(4-NMPA)

O Griz-co-

+N
/

Nmem-pyddlrfimcetyl
(ts-NMPA)

Qua...

pyridWmceM
(PM

(crgrmicnfco-

trimethylammonMMtyl
(TMAA)

(0.11") (or-1,), trier-1,00-
dinnthyloctytanmoniumtyt
(DMOMJ

,1 l?
—rli-CH2CH25 421-120-

thiochoiinoacoty!

0
1+. (crux/K

€50

(©):teu,eu,.

apps)

regs-c»:-

tyl
(TMPPA)

Figure. 3.5. Structures of charged groups in N-terminal derivatization.

32

The earliest N-terminal charge derivatization on peptides for sequence mass
spectrometric analysis was performed in 1984 by Kidwell et a1. [42, 43]. The peptides

were derivatized by reaction with methyliodide, as presented in (2):

CHgI .
NHz-Pep-COOH -> (CH3)3N+-Pep-COOH (2)

This derivatization approach had low yield, so another approach [42, 43] involving
successive derivatization of the peptide N-terminus with chloroacetyl chloride followed

by reaction with trimethylamine to give a trimethylamino acetyl derivative was proposed.

Other amines have also been studied, as shown in (3).

i, ii
NHz-Pep-COOH -> (CH3)3N+-CH2CO-NH-Pep-COOH (3)
R3
|
Reagents: i= Hal-CHZCO-RI ii=R2 —N-R4
a) Hal = or, R1 = Cl, R2=R3=R4=C2H5
b) Hal = CI, R1 = CI, R2=R3=R4=CH3
C) Hal = I, R] = OC(O)CH21, R2=R3=CH3, R4: CH3, C6H|3 01' Can

These derivatives were analyzed by FAB. The derivatives had detection limits in the low

picomole range.

In 1989, iodoacetic anhydride followed by tlriocholine iodide was selected to attach a
quaternary ammonium group to a peptide N-terminus [44]. By controlling pH at each

step, selective derivatization of the N-terminus can be achieved. The pH control allows

33

selective derivatization of the N-terrninus because of the differences between the pKa of
the N-terminus and those of the basic amino acid side chains. However, protection of
cysteines was required to prevent derivatization of the side chain [45]. The reaction was
completed after several hours and needed HPLC to purify the product before they could
be analyzed in the mass spectrometer. Analysis by FAB followed by CID-MS/MS gave
typical N-terminal charge-remote fragment ions, but the spectrum was complicated by the

loss of trimethylamine from the derivatives.

In 1990, Vath and Biemann proposed a gas phase N-tenninal charge derivatization by
devising a technique for peptides deposited on a capillary tube as solid phase derivatized
with first gas phase chloroacetyl chloride, then with gas phase trimethylamine [46]. The

reaction vessel is shown in Figure 3.6.

This two step reaction attached a trimethylamino acetyl moiety to the N-terminus of the
peptide. It took three hours to complete, and could be carried out at a subnanomole level.
The CID spectra of these derivatives are simplified compared with spectra of
underivatized peptides. The products did not need to be purified since excess reagents
had been evacuated. However, it is not possible to eliminate derivatization on a side chain

by controlling pH in the gas phase reaction.

Three years later, a similar reaction scheme for attaching a dimethylalkylammonium

(DMAA) group was developed [47]. Appropriate nucleophiles can be trimethylamine,

dimethyloctylarnine or pyridine. The derivatization involved reaction of at least 100

34

Sample Tube\

   
 
  

 

Vacuum Stopcoek

V
Vacuum

Figure 3.6. Picture of a gas phase N—terminal charge derivatization reaction vessel. (taken
from reference [46])

picomoles of peptides with iodoacetic anhydride followed by reaction with
dirnethylalkylamine. Control of pH can be used to achieve 70-90% selectivity for the N-
terminus over the side chain of the peptides. The procedure involved a total of two hours
of reaction time, and gave a yield of 60-80%. Excess reagents and nonvolatile buffers
must be removed by HPLC. Increased signal intensity was observed in mass spectra of
derivatives. This is presumably the result of increased surface activity. However,
trimethyl ammonium derivatives gave a slight decrease in signal intensity compared to
that for the underivatized peptides. In ClD-MS/MS spectra, a small loss of the alkyl

group from the dimethylalkylammonium functionality was observed.

In 1994, Bartlet-J ones et a1. developed another derivatization approach, which is called

 

 

“CSQ” [27], as shown in (4).
Reagent
NHz-Pep-COOH -) (CH3)3N+-(CH2)5-CO-NH-Pep-COOH (4)
0
ll
/_
Reagent = (CH3)3+N-(CH2)5(O)CO-N
L
H
O

The reaction can be performed sequentially in little more than four hours with quantities
of 5-10 pmol for the complete procedure. HPLC was required to remove excess reagents.
When the peptide was charge derivatized, and basic side chains were modified, only N-
tenninal fragment ions were produced in MALDI-PSD spectra [32,35] primarily an" and

bn‘“ ions. The bn“ ions are formed by a hydrogen shift from the beta position. The bn"

36

ions of charge derivatized peptides are usually present in spectra when very high laser
power was used for desorption [32]. In another study of the “CSQ” charged-derivative,
by using MALDI-PSD, several an* and bn“ fragment ion peaks were observed, but the
most abundant fragments were resulting from a combination of backbone cleavage and

the loss of trimethylamine from the derivative, e. g. an*-59 and bn*-59 ions.

N-methyl-4-pyridinium acetyl (4-NMPA), N-methyl-3-pyridinium acetyl (3-NMPA) and
trimethylpyridiniumbutyl (TMPB) derivatives have also been studied. However, the
reaction took place in a relatively long period of time and the yield of the reaction was
low [24]. In CID-MS/MS spectra, the 4-NMPA derivatives produce ions due to
fi'agmentation of derivative-peptide bonds. It is believed that these losses are produced
when arginine on the side chain extracts a proton, which eliminates the derivative moiety

as shown in Scheme 3.3.

The spectra of 3-NMPA derivatives show that elimination of the derivative is less
favored for the 3-NMPA derivatives than for 4-NMPA. An explanation proposed is that
the pyridinium nitrogen is para to the acetyl linkage in the 4-NMPA and meta in the 3-
NMPA. The former is able to transfer charge via resonance to the peptide, which leads to
the leaving of the derivative moiety as a neutral molecule. In TMPB derivative MS/MS

spectra, loss of trimethylpyridine has been observed which complicates the spectra.

N-terminal charge derivatization using quaternary phosphonium groups have also been

explored. Vinyl triphenylphosphonium bromide [40] and 2-bromoethyl

37

(H‘ ‘:X

—N‘:\ Q1}; co- (N- CH ( RI)- (NH- CH- C0)-
I_IIX+

—©CH2+O= C= N-CH(RI) co-(NH-CH-Cor OH

H Ajf

+ l
—N‘:\ CH- co- NH- CH ( RI)- (NH- CH- co *);OH
H/+
I

—©=i:‘65-H N- CH(R,)-(NH CH C0);0H

Hx+
l
—©C= C= o + H2N- CH ( R,)- co- ( NH- CH- co )gon

Scheme 3.3. Formation mechanism of fragment ions from 4-NMPA derivatives by loss of
charged moiety.

38

triphenylphosphonium bromide [25] have been used to attach a triphenylphosphonium
positive charged moiety on peptides. The derivatization reactions have greater than 75%
yield but with some difficulties in reproducibility. pH control was used to prevent
reaction with basic side chains, but it required removal of the buffer salts prior to mass
spectrometric analysis. In CID-MS/MS spectra, 2-bromoethyl triphenylphosphonium
bromide derivatives produced abundant charge-remote fragments arising from the
derivatized terminus. But in the study of vinyl triphenylphosphonium bromide
derivatives, the spectra are dominated by ions that are solely produced from the
derivative moiety and thus contain no sequence information. The peptide fragment ions

were of rather low relative abundance in the spectra [24].

Huang et a1. introduced a derivatization procedure that attached a tris(2,4,6-
trimethoxyphenyl) phosphonium acetyl (TMPP+-Ac) group to the peptide N-terrninus

[29-30]. The reaction is shown in Scheme 3.4.

A nearly quantitative modification can be achieved with 5-10 fold molar ratio of the
reagents and base over the analytes. The reaction took 15 minutes, and lO-picomole of
peptide can be derivatized. Selective N-terminal derivatization was achieved by pH
control. No HPLC was necessary before mass spectrometric analysis of the products. But
TMPP+-Ac derivatization can be adversely influenced by the presence of certain amino
acids located at the terminus. For example, when the N-terrninus bears an arginine in the
chain, the derivatization reaction efficiency is reduced by enhanced rate of hydrolysis of

the reagent [29].

39

Reagent Br“ 9
NHz-Pep-COOH > TMPP+-CH2-C-NH-Pep-C00H

 

Br'
Reagent = TMPP+-CH2C-S Q F

F
OCH3

.M. (err-‘57...—

OCH3

Scheme 3.4. Mechanism of TMPP+-Ac charge derivatization.

40

The TMPP+-Ac derivative has been used to direct the fragmentation of peptides during
analysis by MALDI-PSD [29,30,32]. In the spectra, a..* ions are dominating, a few bn“,
cn“ and (1,," ions are also observed. When an aspartic acid residue is present at the nth
residue, b..*, (1.." and cn-I* ions are more abundant than a..* ions. This is due to the
hydrogen shift from the aspartic acid side chain forming three different intermediates,
through which low energy paths of fragmentation take place. This reagent has been used
to derivatize peptides located in an electrophoretic gel or membrane [35]. The modified
peptides were then analyzed by MALDI and MALDI-PSD directly from the gel or the

membrane surface.
C-terminal charged-derivatives

C-terminal charge derivatization has been explored. Several approaches are shown in

Scheme 3.5.

Coupling of the peptide C-terminus to a charged pyridinium salt with 1-(3-
dimethylaminopropyl)-3-ethylcarbodiimide (EDC) was used to generate charged-
derivatives [43]. Alternatively, after EDC reacted with the peptide C-terminus, by adding
methyliodide, the terminal nitrogen of EDC was quaternized, to form a charged-
derivative. However, these reagents also derivatize the unprotected acidic side chains of
the peptide. The reaction shown in Scheme 3.5(a) and 3.5(c) are selective C-terminal
derivatization procedures [48, 49]. These two procedures are very time consuming and

required large samples. A C-terminal triphenyl phosphonium derivatization approach was

41

NHZ-Pep-CO—OCHr—lfii» NHZ-Pep-CO-NH-N+-

Reagents: i=carboxypeptidase, NHz-N(CH3)2, ii=CH3I

. o
l D II I
NHz-Pep-COOH—’ NHZ-Pep-CO-N-C-NHCHZCHZCHz-N
O ‘

ll 1
_"_+ NH2—Pep-Co-N-C-NHCH2CH,CH,-N+-

Reagents: i= :N-CHZCHZCHZN=C=NCH2CH3, ii = CH3I

NHz-Pep-COOH ‘_ii_, CH3CO-NH-Pep-CO-OCHZCHz-N(CH3)3
CH3‘ , CH,co-NH-Pep-Co-OCH2CH,-+N(CH,)3

Reagents: i = AczO, ii = HOCH2CH2N(CH3)2

NH2CH2CH2-P+-( @ )3
NH2~Pep-COOH , NH2-Pep-Co-NHCH2CH2-P +-( @ )3
DCC

 

NH2,pep,COOH I‘mzcflchz'+N O NHZ-Pep-Co-NHCHZCHz-w (O)
EDC

 

Scheme 3.5. Mechanisms of some C-terminal charge derivatizations.

42

reported in 1991 [25]. The peptide was mixed with 2-aminoethyltriphenyl phosphonium
bromide and dicyclohexylcarbodiimide (DCC) at pH 5 (Scheme 3.5.d). The
derivatization products can be analyzed without purification, and increased signal was
observed in MALDI spectra of the derivative [30]. However, for all C-terminal charge

derivatization reactions, there is a potential to derivatize acidic side chains.

C-terminal charge derivatization and N-terminal charge derivatization have both been
explored. However, under certain pH control, specific N-terminal derivatization can be
accomplished without modifying the side chains of the peptide [24]. In contrast, it is
difficult to prepare C-terminal derivatives without simultaneous alkylation of the
carboxyl group of the side chains. Also, N-terminal charged-derivatives produce fewer
types of fragment ions than C-terminal charged-derivatives making the spectra of N-
terminal charged-derivatives simpler than those of C—terminal charged-derivatives. In
response to all these advantages of N-terminal charged-derivatives, more efforts were put

into N-terminal charge derivatization.

Research goal

A useful N-terminal charge derivatization must have several attributes. First of all, the
derivatives produce a series of charge-remote fragments upon analysis by MALDI. The
fragment ions must be dominated by an“ and (1,,” ions. The derivatized peptides should
not lose the charged moiety as a neutral fragment to an appreciable extent because

fragment ions that lack the derivative moiety would complicate the interpretation of the

43

spectrum. Secondly, derivatization should not adversely affect ionization efficiency of the
peptide. The derivative should have a lower detection limit in the mass spectrometer than
the original peptide. Third, the reaction should be fast and with high yield of the
derivative desired, be specific to the N-terminus without unwanted derivatization of
peptide side chains. The ideal charge derivatization also should not require purification

prior to mass spectrometric analysis.

Although charged-derivatives offer attractive advantages in the mass spectrometry
analysis of proteins, additional work is still required in this field. The goal of our
approach is to develop a method, which improves the sensitivity of peptide analysis by
MALDI, and provides structure information of peptides by MALDI-PSD, which is easy
and fast enough to be used in routine analytical work and which does not cause mass

spectral interference by non-volatile solvents or reagents.

We put our effort on peptide N-tenninal quaternary amino group derivatization. It is
assumed that since a quaternary amino group is smaller than a quaternary phosphonium
group, the attachment of a quaternary amino group will be less hindered. Charged
quaternary amino groups may direct peptide fragmentation more efficiently through

secondary structure effects.

44

References

1.

10.

ll.

12.

13.

14.

15.

l6.

17.

18.

K. Biemann, “Mass Spectrometry, Organic Chemical Applications”, McGraw-Hill,
New York, 1962

H.D. Beckey, “Principles of Field Ionization and Field Desorption Mass
Spectrometry”, pergamon, Oxford, 1977

RD. MacFarlane, Anal. Chem, 55, 1247A (1983)

M. Barber, R.S. Bordoli, R.D. Sedgwick, A.N. Tyler, J. Chem. Soc. Commun., 325
(1981)

G.C. Didonato, K.L. Busch, Biomed. Mass Spectrom., 12, 364 (1985)
K. Biemann, H.A. Scoble, Science, 237, 992 (1987)

RS. Johnson, S.A. Martin, K. Biemann, Int. J. Mass Spectrom. Ion Processes, 86,
137(1988)

I.A. Papayannopoulos, Mass Spectrom. Rev., 14, 49 (1995)

RS. Johnson, S.A. Martin, K. Biemann, J .T. Stults, J .T. Watson, Anal. Chem, 59,
2621 (1987)

BS. Wagner, Ph.D. Dissertation, Michigan State University, East Lansing, MI (1992)
K.M. Downard, K. Biemann, J. Am. Soc. Mass Spectrom., 4, 874 (1993)

T. Yalcin, I.G. Csizmadia, M.R. Peterson, A.G. Harrison, J. Am. Soc. Mass
Spectrom., 7, 233 (1996)

T. Yalcin, C. khouw, I.G. Csizmadia, M.R. Peterson, A.G. Harrison, J. Am. Soc. Mass
Spectrom., 6, 1165 (1995)

ML. Gross, Int. J. Mass Spectrom. Ion Processes, 118/119, 137 (1992)

NJ. Jensen, K.B. Tomer, M.L. Gross, J. Am. Chem. Soc., 107, 1863 (1985)
M.M. Cordero, J .J . Houser, C. Wesdemiotis, Anal. Chem, 65, 1594 (1993)
N.N. Dookeran, T. Yalcin, A.G. Harrison, J. Mass Spectrom., 31, 500 (1996)

K.A. Cox, S. Gaskell, M. Morris, A. Whiting, J. Am. Soc. Mass Spectrom., 7, 522
(1996)

45

19. NE. Scherman, N.A. Yates, J. Shabanocoitz, D.F. Hunt, W. Jeffery, M. Bartlet-Jones,
D.J.C. Pappin, Proceedings of the 43rd ASMS Conference on Mass Spectrometry and
Allied Topics, Atlanta, May 21-26, 1995 P. 626

20. V.H. Wysocki, M.E. Bier, R.G. Cooks, Org. Mass Spectrom., 23, 627 (1988)

21. D. Renner, G. Spiteller, Biomed. Environ. Mass Spectrom., 13, 405 (1986)

22. DR. Mueller, M. Eckersley, W. J. Richter, Org. Mass Spectrom., 23, 217 (1988)

23. KB. Tomer, F.W. Crow, M.L. Gross, J. Am. Chem. Soc., 105, 5487 (1983)

24. J .Zaia, K. Biemann, J. Am. Soc. Mass Spectrom., 6, 428 (1995)

25. D.S. Wagner, A. Salari, D.A. Gage, J. Leykam, J. Fetter, R. Hollingsworth, J .T.
Watson, Biol. Mass Spectrom., 20, 419 (1991)

26. J .T. Watson, D.S. Wagner, Y.S. Chang, J .R. Strahler, S.M. Hanash, D.A.Gage, Int. J.
Mass Spectrom. Ion processes, 111, 191 (1991)

27. M. Bartlet-Jones, W.A. Jeffery, HF. Hansen, D.J.C. Pappin, Rapid Commun. Mass
Spectrom., 8, 737 (1994)

28. W. Hines, J. Peltier, F. Hsieh, S.A. Martin, Proceedings of the 43rd ASMS Conference
on Mass Spectrometry and Allied Topics, Atlanta, Georgia, May 21-26, 1995

29. Z.-H. Huang, J. Wu, K.D.W. Roth, Y. Yang, D.A. Gage, J.T. Watson, Anal. Chem,
69, 137 (1997)

30. Z.-H. Huang, J. Wu, DA. Gage, J .T.Watson, Proceedings of the 45th ASMS
Conference on Mass Spectrometry and Allied Topics, Palm Springs, CA, June 1-5,
1997

31. P.-C. Liao, J. Allison, J Mass Spectrom., 30, 511 (1995)

32. P.-C Liao, Z.-H. Huang, J. Allison, J. Am. Soc. Mass Spectrom., 8, 501 (1997)

33. T]. P. Naven, W.A. Jeffery, M. Bartlet-Jones, D. Rahman, D.J.C. Pappin,
Proceedings of the 45th ASMS Conference on Mass Spectrometry and Allied Topics,
Palm Spring, CA, June 1-5, 1997

34. B. Spengler, F. Luetzenkirchen, S. Metzger, P. Chaurand, R. Kaufrnann, W. Jeffery,

M. Bartlet-Jones, D.J.C. Pappin, Int. J. Mass Spectrom. Ion processes, 169, 127
(1997)

46

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

J .R. Strahler, Y. Smelyanskiy, G. Lavine, J. Allison, Int. J. Mass Spectrom. Ion
Processes, 169, 111 (1997)

B. Spengler, D. Kirsch, R. Kaufinann, J. Phys. Chem, 96, 9678 (1992)

R. Kaufmann, D. Kirsch, B. Spengler, Int. J. Mass Spectrom. Ion Processes, 131, 355
(1994)

J .P. Kiplinger, L. Contillo, W.L. Hendrick, A. Grodski, Rapid Commun. Mass
Spectrom, 6, 747 (1992)

I. Lindh, W.J. Griffiths, T. Bergman, J. Sjovall, Rapid, Commun. Mass Spectrom, 8,
797 (1994)

D.M. Bunk, R.D. MacFarlane, Int. J. Mass Spectrom. Ion Processes, 111, 55 (1991)
D.M. Bunk, R.D. MacFarlane, Int. J. Mass Spectrom. Ion Processes, 126, 123 (1993)
DA. Kidwell, M.M. Ross, R.J. Colton, J. Am. Chem. Soc., 106, 2219 (1984)

DA. Kidwell, M.M. Ross, R.J. Colton, Springer Ser. Chem. Phys, 36, 412 (1984)

J .T. Stults, R. Halualani, R. Wetzel, Proceedings of the 37th ASMS Conference on
Mass Spectrometry and Allied Topics, Miami Beach, Florida, May 21-26, 1989

R. Wetzel, R. Halualani, J .T. Stults, C. Quan, Bioconjugate Chem, 1, 114 (1990)
J .E. vath, K. Biemann, Int. J. Mass Spectrom. Ion Processes, 100, 287 (1990)
J .T. Stults, J. Lai, S. McCune, R. Wetzel, Anal. Chem, 65, 1703 (1993)

ED. Bennett, R.A. Day, Proceedings of the 35th ASMS Conference on Mass
Spectrometry and Allied Topics, Denver, May 24-29, 1987

K. Hirayama, S. Akashi, R. Yuji, U. Niitsu, Y. Fujimoto, Org. Mass Spectrom., 28,
1516 (1993)

K.D.W. Roth, Z.-H. Huang, N. Sadagopan, J .T. Watson, Charge Derivatization of
Peptides for Analysis by Mass Spectrometry, Mass Spectrom. Rev., submitted 1998

47

CHAPTER FOUR. LIQUID PHASE CHARGE DERIVATIZATION

Derivatization of peptides with betaine hydrochloride and dicyclohexyl

carbodiimide

Introduction

Betaine hydrochloride with a quaternary amino group has been chosen as the derivatizing
reagent. Formation of an amide bond between betaine hydrochloride and the N-terminus
of peptides is an endothermic reaction [1]. Carboxylic acids do react with amines at
elevated temperatures and amides can be produced this way. The temperatures, however,
at which such transformations occur at reasonable rate, far exceed the limits considered
safe for complex peptides. Therefore, in order to form an amide bond, one of the groups,
the carboxylic group of betaine hydrochloride or the N-terminus of the peptides, must be

activated.

Activation of the N-terminus of peptides is a challenging problem. Electron-releasing
substituents should enhance the nucleophilicity of the nitrogen atom, but the substitution
will also decrease the rate of acylation because of the bulkiness of the substituent. The
activation of the carboxyl group remains the underlying principle of the derivatization

methods in use [2], as shown in (1).

48

O O 0

II II NHz-Pep ll
GB—C—OH -) EB—C—X -) GB—C—NH—Pep + HX (1)
X = Halide or carboxylate group
The nucelophillic species undergoes addition at the carboxyl group, followed by
elimination of the halide or carboxylate group. Acyl halides and anhydrides are reactive
acylating reagents because of a combination of the inductive effect of the halogen or

oxygen substituent on the reactivity of the carboxyl group and the ease with which the

tetrahedral intermediate can expel such relatively good leaving groups [1, 2].

In (1), “X” is an electron-withdrawing group, which renders the carbon atom of the
carboxyl group sufficiently electrophillic to facilitate the nucelophillic attack by the
amino group. The tetrahedral intermediate thus formed is stabilized by the elimination of

X' as shown in (2):

O O’ H 0
II | I ll
EB—C—X + NH2——Pep —) [$—C—N+—Pep] -> [ea—C—NH—Pep] + HX
l
1 |
X H
tetrahedral intermediate (2)

Therefore, an appropriate activation method should be found. An attractive approach to
the formation of the charged-derivative is the use of “coupling reagents”. These are
compounds used in peptide synthesis. In this case, we can use them as coupling reagents

that connect the peptides’ N-terminus (NHz—Pep) with betaine hydrochloride through an

49

amide bond. One of the most successful coupling reagents used is 1,3-

dicyclohexylcarbodiimide (DCC) [3].

DCC will first react with betaine hydrochloride to form an intermediate as shown in (3):

O 0

II II
H—OH + (C6H11)N=C=N(C5H11) ‘9 “—0

I
(C6H11)N=C_NH(C6HI I)

Betaine hydrochloride DCC intermediate (3)

Then into the reaction mixture is added the peptides whose N-terminus will attach to the

carboxyl group of betaine hydrochloride, releasing N, N’-dicyclohexy1urea as shown in

(4):

NHz—Pep
O O O

H H H
6 C—0 —) H—NH—Pep + (C6H11)NH—C—NH(C6HII)

l
(C6H1 1)N=C—NH(C6H l 1)
intermediate charged-derivative dicyclohexylurea (4)

The micro-scale derivatization using DCC and betaine hydrochloride with peptides was

demonstrated.

50

Experimental

1. Materials

Betaine hydrochloride ((CH3)3N(C1)CH2C02H, FW 153.61) [4] , 1,3-dicyclohexy1
carbodiimlde ((C6H11)N=C=N(C6H1|), DCC, FW 206.33, 1.0 M 8011111011 in

dichloromethane) [5] were obtained from Aldrich (Milwaukee, WI).

Lys-Arg-Thr-Leu-Arg-Arg (FW 829.0), methionine enkephalin-Arg-Phe (FW 877.0),
Tyr-adrenocorticotropic hormone fragment 4-9 (FW 1068.2), allatostatin 11 (FW 1067.2),
and bradykinin (F W 1067.2) were purchased from Sigma Chemical Co. (St. Louis, MO)

and used without further purification.

a—cyano-4-hydroxycinnamic acid [6] obtained from Aldrich Chemical Co. has been

recrystallized and used as the MALDI matrix.

2. Instrument

MALDI-MS experiments were carried out on the PerSeptive Biosystems Voyager Elite

time-of-flight mass spectrometer equipped with a model VSL-337ND nitrogen laser

(Laser Science, Newton, MA) (337nm, 3-nsec pulse length) and a dual microchannel

plate detector (Galileo, Sturbridge, MA). The instrument has delayed extraction

51

capabilities, and can be operated in either linear or reflectron mode. The acceleration
voltage in the ion source used was 20kV. Data were acquired with the data system

provided and based on a transient recorder with 2 ns resolution.

Time-to-mass conversion was achieved by either external or internal calibration using

matrix (m/z 172.2) and a standard of bradykinin (m/z 1067.2).

All experiments were performed using Ot-cyano-4-hydroxycinnamic acid as matrix.

3. Sample preparation

Peptides were prepared by dissolving the peptides in 1:1 (v/v) solution of

acetonitrile/water at a concentration of 1nmol/uL.

Betaine was first prepared as 4.5nmol/ttL aqueous solution. Then luL of the betaine
solution was air dried in a plastic vial. Dichloromethane was added in, while keeping the

concentration of betaine constant at 4.5nmol/uL.

DCC was diluted to 4.5nmol/uL using dichloromethane.

A saturated matrix solution was prepared in a 1:1 (v/v) solution of acetonitrile/water and
mixed in equal volumes with peptides or derivatization products. The mixture was

allowed to air dry before being introduced to the mass spectrometer.

52

4. Charge derivatization

Pipette the peptide solution (4uL of lnmol/ttL in 1:1 (v/v) acetonitrile/water) into a

plastic vial and let it air dry.

To a solution of the betaine hydrochloride (4ttL of 4.5nmol/ttL in dichloromethane) was
added DCC (4uL of 4.5nmol/uL in dichloromethane). Then luL of the betaine and DCC

solution was pipetted into a plastic vial. After standing at room temperature for 30
minutes, the mixture was added into the vial with solid peptide. The mixture was
vortexed for 1 minute. The vial was sealed and set in room temperature. The reaction was
stopped after 2 hours, by adding 1:1 (v/v) acetonitrile/water into the reaction mixture,

diluting the mixture to a solution at the concentration of 5pmol/uL. Then luL of this

diluted reaction solution was mixed with equal volume of matrix.

Results and Discussion

In the MALDI-MS spectra shown below, we use ‘L’ as linker between the charged

moiety and the peptide N-terminus, ‘9’ as the charged moiety and a two-letter symbolic

name for each peptide.

53

The yield of the reaction is relatively low. For a particular peptide, allatostatin II (m/z
1067.2), the charged-derivative peak (m/z 1167.4) height in the spectra (Figure 1) is 1/3

of the original peptide protonated peak (m/z 1068.4).

Not only the desired derivatives (GB—CO—NH—Pep) are formed for the peptides, but

also the coupling products of molecules of peptides themselves are formed (5):

NHz-Pepl-COOH + NHz-Pepz-COOH -) NHz-Pepl-CO-NH-Pepz-COOH + H20 (5)

As shown in Figure 4.1, the peaks at m/z 2116.8, m/z 2216.8 for allatostatin 11 (Al, FW
1067.2) are two peptide molecules coupled and the charged—derivative of the coupled
molecules respectively. Compared with the MALDI-MS spectrum of allatostatin H before
the derivatization, the intensity of the protonated peptide peak decreased. The intensity of
the peptide protonated peak in the spectrum taken afier derivatization is only 1/10 of the
intensity of the peptide protonated peak in the spectrum taken before the reaction (Figure
41,42). This suggested that 90% of the peptide molecules have been transformed to a

derivative, coupled molecules or coupled-derivative molecules.

In a mixture of peptides, the coupling can take place between molecules of one peptide
themselves, and also can take place between molecules of one species with the molecules
of the other species. These phenomena are shown in Figure 4.3 methionine enkephalin

(Me, FW 877.0) coupled with Tyr-adrenocorticotropic hormone fragment 4-9 (Ty, FW

54

1068.2). In the MALDI-MS spectrum, the peak of the protonated product is at m/z 1928.5

(877.0 + 1068.2 —18= 1927.2). The cross-coupled product was derivatized by betaine

 

a
'3

C

s

s.

3 +
i [A1+H]
a? 1068.4

éB-L-Al
l 1167 .4

[Al+Al-H20+H]*

2116 . 8
$-L-(Al+Al-H20)
2216 . 8

 

 

1500

1.4100 2500
n/z

Figure. 4.1.MALDI—MS spectrum of trimethylamino acetyl charged-derivative €19-L-Al
(m/z 1167.4) of allatostatin 11 (A1, FW 1068.4), peptide coupling reaction product
[Al+Al-H20+H]+ (m/z 2116.8) and its trimethylamino acetyl charged-derivative $-L—
(Al+A1-H20) (m/z 2216.8).

Relative Intensity

+

1068 .0 Wm]

 

1. ll 1L4

 

 

 

600 1200 1600
m/z

Figure. 4.2. MALDI-MS spectrum of allatostatin II protonated peak [A1+H]+ at m/z

1068.0.

55

hydrochloride and DCC, formed the trimethylamino acetyl derivative represented by a
peak 100 m/z units higher (m/z 2027.5). The coupling between the molecules of the same
species Tyr-adrenocorticotropic hormone fragment 4-9 (F W 1068.2) is also present. The
coupling product and its trimethylamino acetyl derivative are at m/z 2119.8 and m/z

2218.8.

lMfi+Hl+ [TWHT

 

 

 

€‘878-5 1069.3 +
g $-L-[Me+Ty-H20] Ty+Ty-H20+H]
E 2119.8
:3 Q-L-Ty - + - _

.5 $—L-Me 1168 . 4 [Me+Ty H20+H] Q-L [Ty+Ty H20]
0: 1928 . 5 2213 s
“978 - 1, 2027.5 ’
1000 1500 2000

m/ 2

Figure. 4.3. MALDI-MS spectrum of a two-peptide mixture: methionine enkephalin (Me,
F W 877.0) and Tyr-adrenocorticotropic hormone fragment 4-9 (Ty, FW 1068.2),
protonated peaks [Me+H]+ (m/z 878.5) and [Ty+H]+ (m/z 1069.3), their derivative peaks
at m/z 978.5 and 1168.4 respectively. In the m/z 2000 range, there are peaks related to
the coupling between peptide molecules.

The MALDI-MS spectrum of the mixture of betaine hydrochloride and DCC is shown in
Figure 4.4. We can locate the peak for protonated DCC at m/z 207.9. At 18 m/z units
higher, is a peak representing the product of DCC hydrolysis. The yield of the
intermediate of this two-step derivatization reaction, produced by betaine hydrochloride

and DCC reaction, is approximately 40% of the total of the original reagent. The peak

represent the intermediate is at m/z 325.0.

56

E [DCC+H]*
E o-L-Dcc
fig [DCC+H20+H]’ 325 . 0
“ 226.0
07 . 9
l'r .L..L - I ll ..h[L 1.]. Li.“ J J_

 

 

 

 

 

 

 

 

200 900 400
m/z

Figure. 4.4. MALDI-MS spectrum of the mixture of betaine hydrochloride and DCC.

There are several shortcomings of the DCC method. First of all, DCC has very low
solubility in water. The reaction must taken place in an organic solvent. However,
peptides and betaine hydrochloride are hydrophilic compounds. They usually are
prepared as aqueous solution. The differences in the hydrophilic and hydrophobic
properties of the reagents tend to complicate the experimental procedure, and may cause

the low product yield.
Second, the N, N'-dicyclohexylurea by-product, while indeed insoluble in most organic

solvents, is not entirely insoluble, therefore it frequently contaminates the product of

coupling. Contamination is one of the causes of the suppression of signals.

57

A more disturbing side reaction is the intramolecular rearrangement of the intermediate.
The attack on the activated carboxyl group by the nearby nucelophillic N results in an
O-) N shift, yielding an N-acylurea derivative as by-product. This by-product is
undesirable not only because they represent a loss of the carboxyl component from
betaine, but also because they introduced another contamination to the product of peptide

derivatization.

Obviously, some amount of the “coupling reagent”, DCC, has not reacted with betaine
hydrochloride, but reacted as a coupling reagent between two peptide molecules. The
formation of coupled peptides and cross-coupled peptides tends to complicate the
MALDI-MS spectrum of the final derivatization product. This may cause difficulties

when we try to identify peptides in peptide mixtures by the MALDI-MS spectrum.

58

Derivatization of peptides with betaine hydrochloride, 1—(3-dimethylaminopropyl)—

3-ethy1 carbodiimide and N-hydroxysuccinimide

Introduction

The DCC coupling reaction has several shortcomings, such as low solubility of DCC in
water, contamination induced by the by-products of the reaction, and the side reaction
forming N-acylurea. A remedy for this imperfection could be in the use of water-soluble
carbodiimides [7, 8]. 1-(3-dimethylaminopropyl)-3-ethyl carbodiimide (EDC, FW
191.71), an ionic carbodiimide ( CH3CH2N=C=N(CH2)3N+H(CH3)2 Cl'), is one of the
water soluble carbodiimides. The experimental procedure could be simplified and the
yield of the reaction should be improved [9] by using reagents with similar
hydrophobic/hydrophilic properties. The by-product of the EDC coupling reaction can be
removed by a drop of water applied on the dried sample spot, on a sample plate, before it

is loaded into a MALDI mass spectrometer.

However, the amine-reactive intermediate formed by EDC and betaine hydrochloride, an
O-acyl isourea, is unstable in aqueous solutions. The similar side reaction as in the case
of DCC, renders rearrangement of the carboxyl and releasing of an N-substituted urea as

shown in reaction (6).

59

O 0

II II
GB—C—O -> H o

I I ll
CH3CH2N=C—NH(CH2)3N+H(CH3)2 CH3CH2N—C—NH(CH2)3WH(CH3)2 (6)
We need to find a way to activate the carboxyl group of betaine hydrochloride. The
intermediate of the reaction should be able to react with amines readily, and it must be

stable and not rearrange or decompose into other species.

Active esters [10, 11] have the properties that we want. They have activated carboxyl
groups and they are stable. One of the active esters is N-hydroxysuccinimide (NHS) [12-
14] ester. So, N-hydroxysuccinimide and its water-soluble analog sulfo-NHS are chosen
to modify carboxyl group on betaine to form an amine active ester. This is accomplished
by mixing the NHS with betaine hydrochloride molecules and a dehydrating agent such
as EDC. The reactivity of O-acyl hydroxyarnines is usually explained with their
anhydride character, and with anchimeric assistance [15, 16] provided by the nitrogen
atom adjacent to the ester oxygen as shown in (7):

O 0

II II
$—C—O—NR2 + NHz—Pep 9 H \
NR2

I
a
I

H—N—H '

Pep (7)

6O

The chemical reaction in the derivatization process is shown as follows:

€B—CH2COOH + CH3CH2N=C=N(CH2)3N+H(CH3)2 or —>

0
II
EB—C—O + HO—NC4H3S05 ->

I
CH3CH2N=C—NH(CH2)3N+H(CH3)2 0‘ (NHS)

0 0

|| ||
EB—C—O—NC4H3S05 + NHz—Pep -) GB—C—NH—Pep (8)
Experimental

1. Materials

Betaine hydrochloride ((CH3)3N(C1)CH2C02H, FW 153.61) and B-mercaptoethanol
(HSCHzCHzOH, FW 78.13) were obtained from Aldrich Chemical Co. (Milwaukee, WI).
(l—(3-dimethylaminopropyl)-3-ethyl carbodiimide (EDC, FW 191.71), 1-hydroxyl-2,5-
dioxo-3-pyrrolidine sulfonic acid monosodium (Sulfo-NHS, FW 217.13) and 2-[N-

morpholino] ethane sulfonic acid (MES) were purchased from Pierce (Rockford, IL).

Allatostatin II (F W 1067.2) was purchased from Sigma Chemical Co. (St. Louis, MO)

and used without further purification.

61

or-cyano-4-hydroxycinnamic acid, obtained from Aldrich Chemical Co., has been

recrystallized and used as the MALDI matrix.

2. Sample preparation

Peptides were prepared by dissolving the peptides in a 1:1 (v/v) solution of

acetonitrile/water at a concentration of 1nmol/ttL. 4.5uL of the peptide solution was air

dried on the bottom of a plastic vial.

Betaine, EDC, and Sulfo-NHS were dissolved separately in MES buffer at pH 6, to

lnmol/uL, 2.1nmol/ttL, and 5.1nmol/uL respectively.

A saturated matrix solution was prepared in a 1:1 (v/v) solution of acetonitrile/water and
mixed in equal volumes with solution of peptides or derivatization products. The mixture
was allowed to air dry on a sample plate before being introduced to the mass

spectrometer.

3. Charge derivatization

First, 4.5uL EDC solution was added into 4uL of betaine solution. Into this mixture was
added 4pL of the sulfo-NHS solution. The reaction took place in room temperature for 15
minutes. Then 0.5}.tL B-mercaptoethanol was added to quench the EDC. Then the mixture

was added into the vial with air-dried 4.5nmol peptide on the bottom of the vial. The

62

mixture was vortexed immediately, and then was set in an ultrasonic water bath for 2
hours. The reaction was stopped after 2 hours, by adding 1:1 (v/v) acetonitrile/water into

the reaction mixture, diluting the mixture to a solution at the concentration of 5pmol/uL.

Then 1p.L of this diluted reaction solution was mixed with equal volume of matrix.

Result and Discussion

The yield of this EDC and sulfo-NHS derivatization method is low, as shown in Figure
4.5. The peak at m/z 1184.9 corresponds to the hydrolyzed derivative of Allatostatin II

(m/z 1067.2).

There exists a certain ambiguity in the reaction of N-hydroxysuccinimide esters. The
strained five membered ring is fairly sensitive to nucelophiles which can open it [17].
There are three carboxyl groups in the intermediate active ester as present in Scheme 4.]:

O SO3'Na+
0 H

H
o—f—o—
I.
0

Scheme 4.1. Structure of the intermediate in the active ester reaction.

However, the open ring product is not observed in the MALDI-MS spectrum. (Figure

4.5)

63

[AHHT

 

gr 1 0 6 8 . '7
E
[$~L-A1+H20]
1 1 8 4. 9

 

 

 

1000 1100 1200
m/z

Figure. 4.5. MALDI-MS spectrum of hydrolyzed derivative [EB-L-AHHZO] (m/z 1184.9)
of allatostatin 11.

Besides other drawback of EDC, there is the potential that the positive charge on EDC

may hinder the formation of an O-urea intermediate with the positively charged betaine

molecule.

64

a

Derivatization of peptides with trimethylamine and bromoacetic acid N-

hydroxysuccinimide ester

Introduction

Bromoacetic acid N-hydroxysuccinimide ester is commercially available. This is a

heterobifunctional active ester with two potential leaving groups: HO-NC4H402 and Br'.

It allows bromoactylation of the N-terminus of peptides, followed by coupling to other

 

compounds, such as NR3 (10).
O
0 II
II ‘
Br—CHz—C—O— + NHz—Pep -)
||
0
O 0

II II
Br—CHT—C—NH—Pep + NR3 —> R3N+—CH2—C—NH—Pep (10)

In this way, we introduce a positively-charged amine group onto the N-terminus of the

peptides.

65

Experimental

1. Materials

Bromoacetic acid N-hydroxysuccinimide ester (FW 236.0) was purchased from Sigma
Chemical Co. (St. Louis, MO). Trimethylamine ((CH3)3N, FW 59.11) in 25% (wt.) water
solution was obtained from Aldrich Chemical Co. (Milwaukee, WI). Poly(vinylidene
difluoride) (PVDF) membrane, pore size 0.45 run, was bought from Millipore Co.
(Bedford, MA). 2-[N-morpholino] ethane sulforic acid (MES) was purchased from Pierce

(Rockford, IL).

Tyr-Ala—Gly-Phe-Leu-Arg (FW 725.8), des-Asp-angiotensin (F W 1181.4), Lys-Arg-Thr-
Leu-Arg-Arg (FW 829.0), methionine enkephalin-Arg-Phe (FW 877.0), Tyr-
adrenocorticotropic hormone fragment 4-9 (FW 1068.2), and allatostatin II (F W 1067.2),
were purchased from Sigma Chemical Co. (St. Louis, MO) and used without further

purification.

or-cyano-4-hydroxycinnamic acid, obtained from Aldrich Chemical Co., has been

recrystallized and used as the MALDI matrix.

66

2. Sample preparation

Peptides were prepared by dissolving the peptides in 1:1 (v/v) solution of

acetonitrile/water at a concentration of 0.1nmol/ttL. Then a IuL aliquot of peptide was

added into 12p.L 0.3M MES at pH 6.

Trimethylamine was diluted with water to 5 pmol/11L.

Bromoacetic acid N-hydroxysuccinimide ester was prepared as 0.01M solution in dry

THF.

A saturated matrix solution was prepared in a 1:1 (v/v) solution of acetonitrile/water and
mixed in equal volumes with peptides or derivatization products. The mixture was

allowed to air dry on a sample plate before being introduced into the mass spectrometer.

3. Charge derivatization

(1). The peptide buffer solution was cooled to 0°C, then added in SuL solution of
bromoacetic acid N-hydroxysuccinimide ester (0.01M in THF) solution. The mixture was
immediately vortexed for 1 minute, then cooled to 0°C for 5 minutes. Afier 5 minutes,
the temperature of the solution was elevated to ambient temperature and at that
temperature for 2 to 3 minutes. Then 5 [1L trimethylamine (Sumol/uL in water) was

added to the mixture. The final mixture was vortexed and heated up to 37°C for 20

67

minutes. The mixture was vortexed every 10 minutes to ensure mixing. After 20 minutes,
the mixture was cooled down to ambient temperature and the solution was diluted to

5pmoI/uL with 1:1 (v/v) acetonitrile/water.

(2). This reaction was also conducted on a PVDF membrane. First, 2p.L, 20pmol/p.L
peptide solution in 1:1 (v/v) acetonitrile/water, was dropped on a piece of membrane. Into
this droplet was added luL of lnmol/uL bromoacetic acid N-hydroxysuccinimide ester
in THF. The mixture droplet was air dried on the membrane. Then 2uL of Sumol/uL

trimethylamine aqueous was dropped on the same place of the dried peptide bromoacetic

acid N-hydroxysuccinimide ester mixture. The droplet was dried in air after 30 minutes.

Results and Discussion

Relatively high yield was obtained for this derivatization method. As shown in the
MALDl-MS spectrum for allatostatin II (m/z 1067.2), the derivative peak is the most
intense one in the spectrum. The intensity of the protonated peptide peak is only 1/6 of

the intensity of the derivative’s peak (Figure 4.6).
However, at shorter reaction times, such as 10 minutes, the reaction is not complete. In

Figure 4.7, we located the bromoactyl peptide peak of allatostatin II (m/z 1067.2),

[Br79—CHz—CO—NH—Pep+H]+ at 1189.5

68

1167 . 5 G-L-Al

Relative Intensity

 

 

 

 

 

[AHHY
1068.5

_ .LELLLuJJLLLLU,A g,
1000 1200 1400

m/z

Figure..4.6. MALDI-MS spectrum of trimethylamine actyl derivative of allatostatin II 69-
L-Al (m/z 1167.5), and the peak of protonated allatostatin II [A1+H]+(m/z 1068.5).

[AHHT
1068.6

[Br°'-L-A1+H]*
ums

Relative Intensity

[Br79-L-A1+H]+ 1189.5
\

 

 

M

 

 

1050 1100 1150 1200
m/z

Figure. 4.7. MALDI-MS spectrum of protonated bromoactyl allatostatin II [Br79-L-
A1+H]* (m/zl 189.5) and its isotopic peak [Br8‘-L-A1+H]* at m/z 1191.5.

69

Mixture of peptides can be derivatized as presented in Figure 4.8, methionine enkephalin-
Arg-Phe (Me, F W 877.0), Tyr-adrenocorticotropic hormone fragment 4-9 (Ty, FW

1068.2), and their derivative peaks at 977.7 and 1168.8 respectively, were observed in the

 

 

 

 

 

spectrum.
[Me+H]+
g. 8 78 . 6
g l
E
$-L-Me [Ty+H]+ .
977-5 1069.?
$—L-Ty
1 1 68 . 8
8 0 0 90 0 1 0 O 0 1 1 0 0
m/ 2

Figure. 4.8. MALDI-MS spectrum of the derivatives of methionine enkephalin-Arg-Phe
(Me, FW 877.0), Tyr-adrenocorticotropic hormone fragment 4-9 ( Ty, FW 1068.2), at
m/z 977.6 (Q-L-Me) and m/z 1168.8 (GB-L-Ty) respectively.

The derivatization of mixture peptides on a membrane is also successful (Figure 4.9).
Tyr-Ala-Gly-Phe-Leu-Arg (TA, FW 725.8), methionine enkephalin-Arg-Phe (Me, FW
877.0), Tyr-adrenocorticotropic hormone fragment 4-9 (Ty, F W 1068.2) and des-Asp-

angiotensin (AA, FW 1181.4) were components in the mixture on the membrane. After

the derivatization process as described above, the corresponding derivatives for each

70

peptide can be detected on the membrane by MALDI-MS. However, the derivatization
efficiencies for these peptides are different. In the best case, for Tyr-adrenocorticotropic
hormone fragment 4-9 (Ty, FW 1068.2), the derivative’s yield is 50%, estimated from the
peak height ratio between the protonated peptide peak and the derivative peak. For Tyr-
Ala-Gly-Phe-Leu—Arg (TA, FW 725.8) and des-Asp-angiotensin (AA, FW 1181.4), 30%

yield was estimated.

[mm]+ e-L-Ty

1000.3

M+H+ ,1168.9 +
726.4 I e ] lTy+Hl .1182.8[AA+H]
878.4 1069.3

Relative lntenslty

0-L-TA 63-L-AA
825 . 8 , 1281 . 3

 

f

 

1 l
. . 0-L-Me . y

W l

800 1000 “1,2 1200 1400 1600

 

 

 

 

 

 

 

 

 

 

 

Figure. 4.9. MALDI-MS spectrum of charged-derivatives of peptide mixture on
membrane.

71

References

10.

11.

12.

13.

14.

15.

16.

17.

. P.P.N. Satchell, Q. Rev. Chem. Soc., 17, 160 (1963)

FA. Carey, R.J. Sundberg, “Advanced Organic Chemistry”, Plenum, New York,
1990

. J.C. Sheehan, G.P. Hess, J. Am. Chem. Soc., 71, 1067 (1955)

The Merck Index 9‘’1 edition, 1208, Merck & Co., Inc., Rahway, New Jersey, 1976
The Merck Index 9th edition, 3075, Merck & Co., Inc., Rahway, New Jersey, 1976
RC. Beavis, T. Chaudhary, B.T. Chait, Org. Mass Spectrom, 27, 156 (1992)

J.C. Sheehan, J .J. Hlavka, J. Org. Chem, 21, 439 (1956)

J .C. Sheehan, P.A. Cruickchank, G.L. Boshart, J. Org. Chem, 26, 2525 (1961)
F.M.F. Chen, K. Kukoda, N.L. Benoiton, Synthesis, 928 (1978)

R. Schwyzer, M. Fenrer, B. Iselin, Helv. Chim. Acta, 38, 69 (1955)

R. Schwyzer, M. F enrer, B. Iselin, H. Kiagi, Helv. Chim. Acta, 38, 80 (1955)

G.W. Anderson, J .E. Zimmermann, F.M. Callaharr, J. Amer. Chem. Soc., 85, 3039
(1963)

Z. Grabarek, J. Gergely, Anal. Biochem, 185, 131 (1990)

J .V. Staros, R.W. Wright, D.M. Swingle, Anal. Biochem, 156, 220 (1986)

SM. Beaumont, BO. Hanford, J .H. Jones, G.T. Young. Chem. Commun., 53 (1965)
B0. Hanford, J .H. Jones, G.T. Young, T.N.F. Johonson, J. Chem. Soc., 6814 (1965)

J. Savrda, J. Org. Chem, 42, 3199 (1977)

72

CHAPTER FIVE. GAS PHASE CHARGE DERIVATIZATION

Derivatization of peptides with chloroacetyl chloride and trimethylamine on a

membrane

Introduction

The spectral simplicity and apparent lack of fragmentation of ionized proteins
characterize MALDI-MS spectra. This suggests that MALDI could be applied to the
analysis of mixtures [1]. Peptide and protein mixtures are important analytical targets.
However, when a mixture of proteins is encountered, the resulting MALDI spectra
frequently show fewer peaks than the number of components in the mixture. In response
to this suppression effect, a separation technique should be used. Protein separation and
purification by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-
PAGE) [2] is a fast and above all an extremely sensitive and almost quantitative method.
By use of high-resolution 2-dimensional polyacrylamide gel electrophoresis (2D—PAGE),
complex protein mixtures can be separated. The 1 or 2 dimensional separations, based on
characteristics of analyte, such as molecular weight, isoelectric point, mobility, yield a

spatial array of separated polypeptides.

However, the interfacing of the two analytical methods is not straightforward. A gel is

not like a metal surface, which is a conducting surface for a time-of-flight measurement

[2]. While success in time-of-flight measurements directly from a gel surface has been

73

reported [3], there still exist many problems of on-gel detection. The obvious problem is
the fact that gels become fragile when dehydrated. They are moisture sensitive, and can
crack and disintegrate in the vacuum conditions of the mass spectrometer [2]. They form
uneven surfaces, which cause a spread of flight times in MALDI-TOF. Also, detergents,
critical to separations, notably SDS, can severely interfere with crystal growth and

MALDI analysis [4-6].

Another way to interface gel electrophoresis to MALDI is via membranes. It is
increasingly common for separated proteins to be transferred onto more robust polymer
membranes by the application of an orthogonal electric field [2]. When electroblotting is
carried out correctly, the proteins are transferred quantitatively and the relative position
of each is retained. Proteins are transferred free of buffers, SDS, and other
contaminations. MALDI-MS analysis of proteins, electroblotted onto polymer
membranes afier SDS-PAGE separation has been reported [7-9]. The proteins or peptides
are desorbed directly from the blot membranes afier matrix application. Several
commercially available membranes, such as polyvinylidene difuoride (PVDF) membrane
[10-11], polyurethane (PU) membrane [12], and polyethylene (PE) membrane [13] have
been adopted as sample supports in MALDI-MS. On these chemically inert membranes,
proteins are present in a highly purified state, in comparison to classical techniques,

essentially free of buffers, salts, detergents or contaminations by other proteins.

Several laboratories have had success in desorbing peptides and proteins directly from

PVDF [11, 14, 15]. It is the membrane that has been used in this research. Our goal in

74

this project is to combine charge derivatization chemistry, which has been successfully
demonstrated for enhanced MALDI analysis of peptides, with PAGE and membranes.
After a mixture of polypeptides is separated, analytes are digested and electroblotted,
resulting in complex mixtures in specific locations on a membrane. If the membrane can
be prepared for direct MALDI analysis, all components of the mixture are usually not
detected, certainly not detected with equal responses. Charge derivatization, attaching a
fixed positive charge to the N-terminus of the peptides, can increase the number of the
components detected and can provide useful MS/MS spectra. We are investigating
formation of charged-derivatives of peptides on membranes. Our approach is to use gas

phase reagents to react with peptides on membranes to form charged-derivatives.

Because of the high sensitivity of the mass spectrometer, only a small amount of the
derivatives needs to be introduced into the instrument, usually by using a small aliquot if
the derivatization is carried out by conventional solution chemisu'y. However, when
solvents are added to a membrane, the separated spots of proteins or peptides can spread.
Damage of the separation is undesired. Other advantage of gas phase reactions over
liquid phase reactions is that excess reagents can be removed by evacuation of the
reaction vessel. We have therefore developed this two step derivatization of picomole
quantities without transfer of the sample or exposing it to bulk solvent or reagents. After

the derivatization, the entire product on a membrane was transferred onto a sample probe.

75

two protein mixture

 

 

fi 4— P, Dag... $911301

" . Pl -->{Di}1
P2 —’ P2 --> {D32 {Di}2 6% >

gel gel

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

@4— P1 <—{Di}1
P2 ’ $ 1 {0,12 9%
membrane ‘ membrane
MS Analysis

Figure. 5.1. Illustration of protein separation on gel followed by electroblotting onto a
membrane; separated proteins digested on gel and then electroblotted onto a membrane.

76

The only limitations are the requirements that the reagents have an appreciable vapor
pressure at the working temperature. The reagents we used are chloroacetyl chloride and
trimethylamine. They both have relatively high vapor pressures at room temperature.
Chloroacetyl chloride is a bifunctional compound, used as linker between the peptide’s

N-terminus and a trimethylamine moiety (1).

o 0
II II
C1—CH2—C—Cl + NHZ—Pep 9 C1—CH2—C—NH—Pep

o 0
II II

C1—CH2—C—NH—Pep +N(CH3)3 9 (CH3)3N+——CH2——C—NH—Pep (1)
Experimental

1. Materials

Chloroacetyl chloride (ClCHzCOCl, FW 112.94) and trimethylamine (FW 59.11, 25wt.%
solution in water) were obtained from Aldrich (Milwaukee, WI). Poly(vinylidene
difluoride) (PVDF) membrane, pore size 0.45 run, was bought from Millipore Co.

(Bedford, MA).

Tyr-Ala-Gly-Phe-Leu-Arg (FW 725.8), des-Asp-angiotensin (F W 1181.4), Lys-Arg-Thr-
Leu-Arg-Arg (FW 829.0), methionine enkephalin-Arg-Phe (FW 877.0), Tyr-
adrenocorticotropic hormone fragment 4-9 (FW 1068.2), neoromedin U-8 (FW 1111.3)

and allatostatin 11 (FW 1067.2), were purchased from Sigma Chemical Co. (St. Louis,

MO) and used without further purification.

Ot-cyano-4-hydroxycinnamic acid, obtained from Aldrich Chemical Co., has been

recrystallized and used as the MALDI matrix.

2. Instrument

MALDI-MS experiments were carried out on the PerSeptive Biosystems Voyager Elite
time-of-flight mass spectrometer equipped with a model VSL-337ND nitrogen laser
(Laser Science, Newton, MA) (337nm, 3-nsec pulse length) and a dual microchannel
plate detector (Galileo, Sturbridge, MA). The instrument has delayed extraction
capabilities, and can be operated in either linear or reflectron mode. The acceleration
voltage in the ion source used was 20kV. Data were acquired with the data system

provided and based on a transient recorder with 2 ns resolution.

Time-to-mass conversion was achieved by either external or internal calibration using

matrix (m/z 172.2) and a standard of bradykinin (m/z 1067.2).

All experiments were performed using or-cyano-4-hydroxycinnamic acid as the MALDI

matrix.

78

3. Sample preparation

Peptides were prepared by dissolving the peptides in 1:1 (v/v) solution of

acetonitrile/water at a certain concentration. Then lpL of the peptide solution was

dropped onto the membrane surface and air-dried.

The PVDF membrane was washed with 1:1 (v/v) acetonitrile/water then methanol and

dried in air before use.

A saturated a-cyano-4-hydroxycinnamic acid matrix solution, luL, prepared in a 1:1

(v/v) solution of acetonitrile/water was dropped on the membrane surface with peptides
or derivatization products. The mixture was allowed to air dry before being introduced

into the mass spectrometer.

4. Charge derivatization

Membrane with peptides dropped and dried on the surface was put in a reaction vessel,

which has 10 mL of the proper reagent in a vertical sidearm (Figure 5.2).

79

Picture of the glass reaction vessel

 

 

 

 

 

 

 

 

 

 

 

: Vacuum stopcock 1
Membrane — .1.
J\Reagent
Peptide dot l 1 Vacuum stopcock 2
Vacuum

Figure. 5.2. A glass reaction vessel with two side arms which containing membrane and
reagents, is connected to a vacuum pump.

80

At the beginning, the reagent in the right side arm was chloroacetyl chloride, and the
vacuum stopcock 1 was closed. The vessel was evacuated to remove the air, then sealed.
Vacuum stopcock 1 was then opened to let the reagent vapor fill the reaction vessel at the
room temperature. After the reaction between the peptides and chloroacetyl chloride on
the membrane has taken place for a period of time, the vacuum stopcock 1 was closed.
Another reagent, trimethylamine was loaded and before opening vacuum stopcock 1, the
vessel was evacuated to remove the excess chloroacetyl chloride reagent vapor from the
vessel. When the vacuum stopcock 1 opened again, the peptides were exposed to

trimethylamine vapor, which was heated up to 45°C for a period of time, to finish the
derivatization. Onto the derivatized peptide dots on membrane, 1 [1L of or-cyano-4-

hydroxycinnamic acid matrix solution was added. After drying in the air, the membrane

was introduced to the MALDI-MS on a sample plate.

Another reaction vessel we employed is a plastic mold, into which a reservoir with
reagent can be attached. The membrane with air-dried peptides' dots can be fitted in the
mold and it is the only place through which the reagent vapor can go, as shown in Figure

5.3.

In these experiments, with the glass vessel or with the mold, only the second reagent,

trimethylamine, was warmed to 45°C. All other parts of the reaction vessels and the

reaction with chloroacetyl chloride were kept at room temperature.

81

Another reaction vessel employed

Reagent vapor

  
  
 

 

Two
Membrane 4._._.__ > pieces
between the 0f the
, mold
two preces of the
mold [ \ reservoir
Reagent

 

 

 

Figure. 5.3. Another vessel for the reaction, the membrane was fixed on a mold and
connected with a reagent container. The reagent molecules collided with the peptide
molecules stabilized on the membrane when the reagent vapor diffused out.

82

Results and Discussion

Peptides in mixtures can be detected directly from a membrane. The resulting spectrum

has good resolution and signal-to-noise ratio, as shown in Figure 5.4.

[AA+H]+
1182 .3

—‘E
g
E M +H *
a I e ] [AH'HT

878 .O

1068 .2
[TA+H]+
726 .3

 

 

 

. .. -1001 - 111111100001111...

7’00 800 900 mIgooo 1f00 1200 13'oo

 

 

 

Figure. 5.4. MALDI-MS spectrum of a four peptide mixture of Tyr-Ala-Gly-Phe-Leu-
Arg (TA, FW 725.8), methionine enkephalin-Arg-Phe (Me, FW 877.0), allatostatin II
(Al, FW 1067.2), and des-Asp-angiotensin (AA, F W 1181.4), deposited on PVDF. Peaks
at m/z 726.3, 878.0, 1068.2, and 1182.3 represent the four protonated peptide species

respectively.

After the peptides were exposed to the first reagent vapor, chloroacetyl chloride vapor,
for 10 minutes, the membrane was taken out from the reaction vessel and the product of
the first step of the derivatization was detected on the membrane. The spectrum of three

component peptide mixture afier the first step derivatization is presented in Figure 5.5.

83

 

1 1 a 7 . 7 [Cl-L-Ne+H]+

 

"é“ [Ne+H1*
1%
'3 1112.3
5
2
A1+H *
[ ] [Cl-L-A1+H]+
1068 . 7
//‘1‘“:5 [Cl-L-AA+H]+
1257 . 9
[AA+H]*
- 1182. 1

 

 

 

 

W Wm:-

900 1000 11'00 1200 17900 12100
n 2

 

 

Figure. 5.5. MALDI-MS spectrum of chloroacetyl derivatives of allatostatin 11 (Al, FW
1067.2), neoromedin U-8 (Ne, FW 1111.3), and des-Asp-angiotensin (AA, FW 1181.4).
The m/z values shift up by 76 for each peptide.

However, the resolution and signal-to-noise ratio of those peaks decreased dramatically

compared with those peaks detected on membrane without going through the reaction.

As presented in Figure 5.6, the two step derivatization was complete. The peptides were
first reacted with the chloroacetyl chloride vapor for 2 hours. The chloroacetyl
derivatives of this step were then reacted with trimethylamine in the gas phase, for 2
hours, to yield trimethylamino acetyl derivatives. The peaks of protonated peptides as
shown in Figure 5.4, are virtually disappeared from the spectrum, only the charged-

derivatives of these peptides appear and gave good responses in the MALDI-MS

spectrum.

84

 

GB-L-AA

 

 

 

 

 

e

g 1280.9

i e-L—Ar

:3 1166.9

.2

875 . 6
998 . 7
800 1000 1200 1400
m 2

Figure. 5.6. MALDI-MS spectrum of allatostatin 11 (Al, FW 1067.2) and des-Asp-
angiotensin (AA, FW 1181.4) derivatization products. The derivatization reaction was
completed afier 2 hours of reaction with chloroacetyl chloride and 2 hours with
nimethylamine in the gas phase.

In fact, the reagents we used are very active. Here is a spectrum taken after 10 minutes of
reaction with chloroacetyl chloride and then 10 minutes with trimethylamine in gas phase
(Figure 5.7). The whole reaction consumed only 20 minutes in total. Not all peptide
molecules reacted and were derivatized; we can still observe peaks of protonated peptides
that remained unreacted. But we already get more than 50% derivatization yield,

estimated from the relative height of derivative peaks and original peptide peaks, for all

of the peptides in the mixture.

85

5 0-L-Ne
E ea-L-Al 1212 . 0
f 1167 . 0
>
'5:
£2 $-L-TA
825 . 6 GB L X
- - e-L-AA
992 . 5
1281 . 1
”Amy 1X+Hr [Ne+H]*
727,5 893.5 1112.4

 

 

 

 

 

 

 

 

 

     

 

 

7'00 800 900 1000 11172.00 1200 1900

Figure. 5.7. MALDI-MS spectrum of peptide mixture derivatization products on a
membrane. After a 20-minute reaction, with first chloroacetyl chloride, then
trimethylamine, the 50% yield of the reaction was obtained.

With low quantities, some peptides will not give good response, occasionally no response
at all in a MALDI-MS spectrum. For example, 500 fmol des-Asp-angiotensin (AA, F W
1181.4) was not detectable directly from a membrane in the spectrum as Figure 8a. This
may be caused by the hydrophobic interactions between peptides and membrane surface,

so not enough peptide molecules have been desorbed from the surface to be ionized and

give a signal in the spectrum.

After we added onto the peptide N-terminus, a polar chloroacetyl group, the chloroacetyl

86

derivative of the peptide gave higher response than the original peptide (Figure 5.8.b). At
the end, peptide molecules have been derivatized as the trimethylamino acetyl derivative,
and we can identify it represented by a peak at 100 m/z units higher than the m/z value of

the original peptide (Figure 5.8.0).

In another case, the analyte is a small peptide, VGVAPG (FW 498.9). Usually, small and
hydrophobic peptides do not form as peaks at the expected m/z values in MALDI-MS
spectra 5.9(a). There exist difficulties in desorption and ionization of these molecules
from the membrane in the MALDI-MS technique. After the charged-derivatization

process, the derivative shows good response at m/z 599.3 in Figure 5.9(b).

Other amine compounds that have been employed instead of trimethylamine in the
derivatization are dimethylamine and cyclohexylamine. The resulting spectrum of
derivatization of neoromedin U-8 (m/z 1111.3) with dimethylamine is shown in Figure
5.10. The dimethylamine group substituted the chloride on the other end of the
chloroacetyl peptide and the total m/z value shift of the derivative from the original

peptide m/z is 86.

87

 

 

1281.0 Q'L'AA

+100—>

Relative Intensity

 

 

 

1258.1 [Cl-L-AA+H]+

 

+76

 

+H *
1181.] [AA ]

A. A --.- La.— A A.

'v w

 

 

1000 1600 zooo

Figure. 5.8(a). In MALDI, 500 frnol of a peptide on a membrane is hardly detectable.
5.8(b). MALDI-MS spectrum of the first step derivatization product, which is 76 m/z
units higher than the original peptide, indicating the presence of the low amount of
peptide on the membrane. 5.8(c). MALDI-MS spectrum of the two step derivatization
product, which is 100 m/z unit higher in the spectrum from the m/z value of peptide.

88

 

 

 

 

 

 

 

551 .. 6
£- a.
6
E
VGVAPG
l 626 - 4
b. 551 .. 6
GB-L-VGVAPG
599 - 4
VGVAPG
- 26 .. 4
+100 >
500 550 600 650
m/z

Figure. 5.9(a). 50 pmol VGVAPG (FW 498.9) on a membrane is not detectable in
MALDI-MS. 5.9(b). MALDI-MS spectrum of charged-derivative of VGVAPG. Charge
derivatization enhances the response of VGVAPG (F W 498.9), which has no response as
an underivatized peptide in MALDI-MS.

89

£- ®'-L-NC
3 H974
2
§—" +86 —F
me+
nnA

 

 

 

W M

10570 1100 1130
m

 

 

1200 1250 1500
[2

Figure. 5.10. MALDI-MS spectrum of dimethylamineacetyl derivative of neoromedin U-
8. Derivatization with chloroacetyl chloride and dimethylamine reagent vapors of
neoromedin U-8 (Ne, FW 1111.3) produces the charged-derivative ($'-L-Ne), which in
the MALDI-MS is at m/z 1197.3.

However, the derivatization is not successful in the case with cyclohexylamine. In the

MALDI-MS spectrum of the final mixture for the same peptide, only the chloroacetyl

peptide was observed at 76 m/z units higher from the original m/z value of the peptide, as

shown in Figure 5.11.

90

 

 

 

 

 

 

 

’E‘.‘
s lNe+H]‘
f 1112.2
.5
in
1000 1100 1200 1300

m]:

Figure. 5.11. MALDI-MS spectrum of the result of an unsuccessful neoromedin U-8 (Ne,
FW 1111.3) derivatization with chloroacetyl chloride and cyclohexylamine reagent
vapors, only chloroacetyl derivative [Cl-L-Ne+H]+ is observed at m/z 1187.1.

A possible reason for unsuccessful derivatization with cyclohexylamine is the lower

vapor pressure of the reagent [16].

Post-source decay (PSD) spectra have been taken for the trimethylamino acetyl derivative
and dimethylamine acetyl derivative of neoromedin U-8. The base peaks in the spectra
are the charged-derivatives in both cases. Extensive fragmentation within the charged
moiety on the N-terminus of peptide was observed, as shown in Figures 5.12 and 5.13.

Little structural information can be obtained from these PSD spectra.

91

Q'L'Ne 1211.77

Relative Intensity

(ea-L-Newirg,

1152.3
1188..

   

 

 

1.1.0

 

 

 

 

40° 603,, . 800 1000

Figure. 5.12. Post-source decay spectrum of trimethylamino acetyl neoromedin U-8. The
base peak is trimethylamino acetyl neoromedin U-8, at m/z 1211. The peak at m/z 1152
is from the fragmentation and loss of trimethylamine group from 1211.

92

Relatlvc Intenalty

 

613’~L-Ne
1197. 3

1180.2
($’-L-Ne)-NH(CH3)2

1152.3

 

 

 

200 400 600

800 1000 1200

Figure. 5.13. Post-source decay spectrum of dimethylamine acetyl neoromedin U-8. The
base peak is dimethylamine acetyl neoromedin U-8, at m/z 1197. The peak at m/z 1152 is
from the fragmentation and loss of dimethylamine group from 1197.

93

Although the derivatization was successful for most of the peptides used in the
experiments, there were some peptides that have no peptide signals in MALDI-MS
spectra, also gave no signals afier derivatization, such as Thr-Ser-Lys (m/z 334.4), Pro-
Phe-Gly-Lys (m/z 507.6), and or-neurokinin (m/z 1133.3). Some times the derivatization
product detected on membrane shows both the chloroacetyl derivative and the
trimethylamino acetyl derivative. Long hours of trimethylamine reaction have been

adopted, but the yield of the reaction still needs to be improved.

Matrix crystal formation is critical to MALDI-MS spectra. Ot-cyano-4-hydroxycinnamic
acid matrix is usually applied by adding into the peptide (either dried or still in droplet)
an equal volume of saturated matrix solution. The or-cyano-4-hydroxycinnamic acid
matrix solution is acidic and has a high percentage of organic solvent (1:1
acetonitrile/water). The application of matrix solution can cause the protein to spread out

on the PVDF. A spot can become a ring if a drop of matrix is placed on a protein spot and

the concentration of the protein per mm3 of membrane is reduced. Improvements in the
composition and application of matrix solution to reduce protein and peptide spreading
will assist in the development of the direct MALDI-TOF analysis of these biopolymers

from PVDF.

Another contribution to the low resolution and low intensity of the spectra taken from the
membrane is from PVDF itself. The laser intensity required to reach the ionization
threshold on PVDF was higher than that required for a metal target. This was attributed to

the porosity of PVDF, which permits a distribution of the analyte and matrix within the

94

membrane [13]. Larger variance, lower resolution in flight times was observed with
PVDF due to the spatial distribution of sample within the pores. In Figure 5.14 (adopted
from reference 13), the electron micrograph of PVDF membrane shows us that the matrix

crystals are embedded in the pores on the membrane [13].

The main advantage of charge derivatization is that the PSD spectra of the derivatives
should be much simpler and identification characterization should be done by interpreting
the fragment peaks in PSD spectra [17, 18]. Unfortunately in the PSD spectra obtained
for the trimethylamino acetyl derivative, extensive fragmentation was in the charged
group. The charged moiety is not as stabilized on the peptide as a TMPP group in the
TMPP+-Ac derivatives [5]. Further study needs to be conducted to find a proper base or a
linker (between the charged group and N-terminus of peptides, in this study, chloroacetyl
chloride) of a different structure as derivatization reagents in gas phase, in order to

enhance the PSD spectral quality.

95

 

Figure. 5.14(a). Electron micrograph of PVDF membrane obtained at a x4800
magnification. (taken from reference [13])

{3 ‘ . .7 .“r’fga ‘ ”2..
o ., _. . 9' '3‘ _ ‘ ”T” r Matrix crystal
p- 7.

    

“.t_ ..
0 , :‘A
. .64;
Mfif'.$§fs

Figure. 5.14(b). Electron micrograph of PVDF after deposition of sample and matrix
solution. The matrix crystals are embedded into the membrane surface. (taken from
reference [13])

96

References

1. ML. Vestal, P. Juhasz, S.A. Martin, Rapid Commun. Mass Spectrom, 9, 1044 (1995)

2. RR Ogorzalek Loo, T.I. Stevenson, C. Mitchell, J .A. Loo, D.C. Andrews, Anal.
Chem, 68, 1910 (1996)

3. K. Strupat, M. Karas, F. Hillenkamp, C. Eckerskom, F. Lottspeich, Anal. Chem, 66,
464 (1994)

4. F.M.L. Amado, M.G. Santana-marques, A.J. Ferrer-Correia, K.B. Tomer, Anal.
Chem, 69, 1102 (1997)

5. N. Xu, Z.-H. Huang, J .T. Watson, D.A. Gage, J. Am. Soc. Mass. Spectrom., 8, 116
(1997)

6. J .A. Loo, Bioconjugate Chem, 6, 644 (1995)
7. C. Eckerskom, F.Lottspeich, Electrophoresis, 11, 554 (1990)

8. R. Aebershold, “ Advances in Electrophoresis”, VCH Weinheim, New York, 1990,
81-163

9. C. Eckerskom, K. Strupat, D. Schleuder, D. Hochstrasser, J .-C. Sanchez, F.
Lottspeich, F. Hillenkamp, Anal. Chem, 69, 2888 (1997)

10. M.M. Vestling, C. Fenselau, Anal. Chem, 66, 471 (1994)
11. M.M. Vestling, C. Fenselau, Biol. Soc. Trans, 22, 547 (1994)

12. ME. Mccomb, R.D. Oleschuk, D.M. Manley, L. Donald, A. Chow, J .D.J. O’Neil, W.
Ens, K.G. Standing, H. Perreault, Rapid Commun. Mass Spectrom, 11, 1716 (1997)

13. J .A. Blackledge, A.J. Alexander, Anal. Chem, 67, 843 (1995)

14. M. Karas, U. Bahr, A. Ingendoh, E. Nordhoff, B. Stahl, K. Strupat, F. Hillenkamp,
Anal. Chim. Acta, 241, 175 (1990)

15. M.M. Vestling, C.Fenselau, W.-M. Ching, at the 9th Symposium of the Protein
Society, Boston, July 8-12, 1995

16. The Merck Index 9th edition, 2739, Merck & Co., Inc., Rahway, New Jersey, 1976

17. B. Spengler, D. Kirsch, R. Kaufrnann, J. Phys. Chem, 96,9678 (1992)

97

18. R. Kaufmann, D. Kirsch, B. Spengler, Int. J. Mass Spectrom. Ion Processes, 131, 355
(1994)

98

anHroAN STATE UNIV. LIBRARIES
lllllllllllHIlllllll"mull“Wllllllllllllllllllll"WI
31293016885299