fl

 
  

W}!

’lh

MU

   

IHJWHWWUW

l
1

 
 

{IWHHIIHHWIN

 
     

—'l.p._x
1A0.)
0100

.5

‘ ' ”l

\\\\\‘l\\\\\\

\
312

llllllllllll

\\\\

This is to certify that the
thesis entitled
A Study of Cyanylation Reagents for the Assignment

of Cysteine Residues and Disulfide Bond Pairings in Proteins
by MALDI-TOF Mass Spectrometry

 

presented by
Qing Li

has been accepted towards fulfillment
of the requirements for

Master degree in Chemistry

 

Date

 

Lori/ML; (5??? %

0.7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

 

LIBRARY

Michigan State

 

 

University

PLACE IN RETURN BOX

to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DA

bus

..0

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1m WM“

 

 

 

A STUDY OF CYANYLATION REAGENTS FOR THE ASSIGNMENT OF
CYSTEINE RESIDUES AND DISULFIDE BOND PAIRINGS IN PROTEINS BY
MALDI-TOF MASS SPECTROMETRY
By

Qing Li

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
A STUDY OF CYANYLATION REAGENTS FOR THE ASSIGNMENT OF
CYSTEINE RESIDUES AND DISULFIDE BOND PAIRINGS IN PROTEINS BY
MALDI-TOF MASS SPECTROMETRY
By

Qing Li

A novel methodology to locate cysteines and disulfide bond pairings in peptides
and proteins has been developed by the Watson group. This new approach employs a
specific reaction between free sulfliydryls and a cyanylation reagent such as 2-nitro-5-
thiocyanobenzoic acid or l-cyano-4-dimethylamino-pyridium tetrafluoroborate to
cyanylate free sulfliydryl groups. The N-terminal peptide bond of the cyanylated
cysteine residue can be cleaved under alkaline conditions to form an amino-terminal
peptide and a series of 2-iminothiazolidine-4-carboxyl (IT C) peptides which can be
mapped to the sequence by Matrix-Assisted Laser Desorption/Ionization Mass
Spectrometry.

This thesis focuses on two new cyanylation reagents: 2-thiocyanopyridine
(PYRSCN) and 4-(dimethylamino) phenyl thiocyanate. These two reagents were found
to be similar in reactivity to NTCB when reacting with free sulfhydryl groups; a
comparison is also made with CDAP. The two new reagents tend to form both the
cyanylated product and a mixed disulfide product. Structural features of the cyanylation
reagent and the pH of the reaction medium were studied using model peptides. Based on
a plausible reaction pathway for the cyanylation process, a new series of reagents is

proposed which may avoid the formation of the mixed disulfide during the reaction.

TABLE OF CONTENTS

List of Figures ........................................................................................ vi
List of Tables ......................................................................................... ix
List of Schemes ....................................................................................... x

Chapter I. Introduction

I.

II.

III.

IV.

Introduction ................................................................................... 1
Characterization of Cysteine and Disulfide Bond Status in Proteins ................. 3
A. Localization of Sulfhydryl Groups ............................................... 3
1. Classical Approach .......................................................... 3
2. Cleavage at Cysteine Residues ............................................. 9
B. Localization of Disulfide Bonds ................................................ 11

Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MS)

................................................................................................ 14
A The Matrix ......................................................................... 15

B Sample Preparation ............................................................... 16
C. Instrumentation .................................................................... 18
D Novel Techniques in MALDI-TOF MS ....................................... 19

E Application of MALDI-TOF MS ................................................ 20
References ................................................................................... 23

iii

Chapter II.

Novel Methodologies for The Assignment of Both Cysteine Residues and

Disulfide Bond Pairings in Proteins

I. A Strategy to Locate Cysteine Residues in Proteins by Specific Chemical

Cleavage Followed by Mass Mapping by MALDI-TOF MS ........................ 29
A. Introduction .......................................................................... 29
B. Cyanylation of Cysteine Residues ................................................ 30
C. Chemical Cleavage at Cysteine Residues ....................................... 31
II. A Strategy for Assignment of Disulfide Bond Pairings in Proteins ................. 35
A. Introduction .......................................................................... 35
B. A Novel Strategy for Disulfide Assignment .................................... 36
111. References .................................................................................... 41
Chapter III. Study on Cyanylation Reagents
I. Introduction ................................................................................. 43
A. Organic Mercurial Reagents ...................................................... 43
B. Sulfhydryl/Disulfide Exchange ................................................... 45
C. NTCB (2-nitro-5-thiocyanobenzoic acid) As a Reagent ....................... 48
D. CDAP (1-cyanO-4-dimethylamino-pyridium salt) AS a Reagent ............. 49
11. Materials and Methods ..................................................................... 51
A. Chemicals ............................................................................ 51
B Synthesis of 2-thiocyanopyridine (PYRSCN) .................................. 52
C. Detection of PYRSCN by GC/MS ............................................... 53
D Mass Spectrometry ................................................................. 53

III.

IV.

VI.

E. Cyanylation of Nascent Sulfliydryls ............................................. 54

F. HPLC Separation of Partially Reduced and Cyanylated Protein Isomers

........................................................................................ 54

G. Cleavage of Singly Reduced and Cyanylated Protein Isomers ............... 55

H. Complete Reduction of Remaining Disulfide Bonds ........................... 55

Results and Discussion .................................................................... 55
A. Locate the Free Sulfhydryl Groups in Rabbit Muscle Creatine

Phosphokinase using PYRSCN as the cyanylation reagent .................. 55

B. pH Effect on Cyanylation Reaction ............................................. 58

1. Results of Study on Peptide 2428 ........................................ 6O

2. Results of Study on Peptide 1657 ........................................ 63

3. Results of Study on Peptide 919 .......................................... 66

C. Introduction to the Reaction Pathway .......................................... 70

D. Effect of Reagent Structure on the Cyanylation Reaction ................... 73

Summary .................................................................................... 76

Future Work ................................................................................. 78

References ................................................................................... 80

Figure 1.1

Figure 1.2

Figure 1.3

Figure 1.4

Figure 1.5

Figure 1.6

Figure 1.7

Figure 1.8

Figure 1.9

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

LIST OF FIGURES

The modification of cysteine with iodoacetic acid to form S-

carboxymethylcysteine ............................................................ 5
Derivatization of the cysteine group by SBD-F or ABD-F ................... 6
p-mercuribenzoate .................................................................. 7
Ellman’s Reagent: 5 ,5’-dithio-bis-(2-nitrobenzoic acid) ...................... 7

5-[2-((iodoacetyl)amino)-ethyl] napthalene-l-sulfonic acid (IAEDAN S)
......................................................................................... 9

Structures of 2-nitro-5-thiocyanobenzoic acid (NTCB) and l-cyano-4-

dimethylaminO-pyridium salt (CDAP) ......................................... 10
Illustration of sulfliydryl/disulfide exchange in proteins ..................... 12
Classical strategy for disulfide bond assignment of a protein ............... 13

Scheme of matrix-assisted laser desorption/ionization time-of-flight mass
spectrometer (MALDI-TOF MS) ............................................... 21
Reaction between cysteine residue and 2-nitro-5-fl1iocyanobenzoic acid
(NTCB), (A) Cyanylation and (B) Cleavage Reaction. B-elimination
competes with cleavage reaction under alkaline conditions ................. 33
A mechanism for the base-catalyzed cleavage reaction ..................... 34
Descriptive overview of the cyanylation/cleavage methodology for
assignment of disulfide bond pairings in proteins ............................ 38

Chemical overview of the cyanylation/cleavage methodology. Partial

reduction means that the protein of interest is reduced under controlled

vi

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9

Figure 3.10

Figure 3.11

Figure 3.12

conditions into a mixture of isoforms, each of which corresponds to
reduction of only one of its disulfide bonds; for a protein containing n
cysteines, n isoforrns of the singly-reduced protein will result. ITC stands
for iminothiazolidine carboxyl residue at the amino terminus .............. 39
Chemical reaction between p-mercuribenzoate and sulfhydryl group. . ...44

The reaction of Ellman’s reagent, 5,5’-dithio-bis-(2-nitrobenzoic acid),

with cysteinyl residues in proteins .............................................. 45
Structure of DTNB and 3-carboxypropyldisulfide ........................... 46
Cysteine Modification Reagents ................................................ 48
Reaction Of NTCB with SH ...................................................... 49
Reaction of CDAP with a generic free sulfhydryl group(RSH) ............ 50
HPLC chromatogram of cyanylated creatine phosphokinase ............... 57

MALDI spectrum of cleavage product of creatine phosphokinase HPLC
peak 3 in Figure 3.7 ............................................................... 58

HPLC chromatogram of cyanylated peptide 2428, at pH 7.0 & 8.0, room

temperature, 50-fold PYRSCN for 30 minutes ................................ 60
(a) MALDI spectrum of peptide 2428 HPLC peak 1 ........................ 61
(b) MALDI spectrum of peptide 2428 HPLC peak 2 ........................ 61
(c) MALDI spectrum of peptide 2428 HPLC peak 3 ........................ 62

HPLC chromatogram of cyanylated peptide 1657, at pH 7.0 & 8.0, room

' temperature, 50-fold PYRSCN for 30 minutes ................................ 63

(a) MALDI spectrum of the above peptide 1657 HPLC peak 1 ............ 64

(b) MALDI spectrum of the above peptide 1657 HPLC peak 2 ............ 64

vii

Figure 3.13

Figure 3.14

Figure 3.15

Figure 3.16

(c) MALDI spectrum of the above peptide 1657 HPLC peak 3 ............ 65

HPLC chromatograms of cyanylated peptide 919, at pH 7.0 & 8.0, room

temperature, SO-fold PYRSCN for 30 minutes ................................ 66
(a) MALDI spectrum of the above peptide 919 HPLC peak 1 .............. 67
(b) MALDI spectrum Of the above peptide 919 HPLC peak 2 .............. 67
(c) MALDI spectrum of the above peptide 919 HPLC peak 3 .............. 68

HPLC chromatograms of the cyanylation product of peptide 2428 reacted
with 5-fold molar excess of 4-(dimethylamino) phenyl thiocyanate at pH
7, room temperature for 30 minutes ............................................. 73

Structure of N—cyanoimide salt .................................................. 79

viii

Table 3.1

Table 3.2

Table 3.3

Table 3.4

Table 3.5

Table 3.6

LIST OF TABLES

Calculated and observed masses for cleavage products of creatine
phosphokinase ..................................................................... 57
Calculated and observed masses for cyanylation product of peptide 919,
1657 and 2427.9 ................................................................... 59
Product ratios of peptide 2428 after reacting with PYRSCN under both pH
7 and pH 8 .......................................................................... 60
Product ratios of peptide 1657 after reacting with PYRSCN under both pH
7 and pH 8 .......................................................................... 63
Product ratios of peptide 919 afier reacting with PYRSCN under both pH
7 and pH 8 .......................................................................... 66
Product ratios of peptide 2428 after reacting with PYRSCN and 4—

(dimethylarnino) phenyl thiocyanate at pH 7 .................................. 73

ix

Scheme 3.1

Scheme 3.2

Scheme 3.3

LIST OF SCHEMES

Formation of cyanylated and mixed product using PYRSCN reagent
....................................................................................... 69
Reaction proposed to occur in GSH and NTCB reaction by Degani and

Patchornik .......................................................................... 70
Two proposed pathways for the reaction of NTCB with GSH by Price

....................................................................................... 71

Chapter I

Introduction

I. Introduction

Since its discovery in 1912, mass spectrometry (MS) has provided key insight for
broad and diverse disciplines. However, many compounds of biological interest are
difficult to volatilize directly and are usually derivatized in order to increase volatility for
ion production by electron ionization in the vapor phase. This frequently leads to thermal
degradation at the same time to produce a complicated mass spectrum which is difficult
to interpret (1). Advances in chemical technology have been the engine powering the
biotechnology industry. Analytical chemists have added fresh impetus to bioresearch
with two new mass spectrometry ionization tools: matrix-assisted laser
desorption/ionization (MALDI) (2) and electrospray (E81) (3). Commercial availability
of these instruments has made routine the analysis of high-mass compounds including
proteins, peptides, carbohydrates, oligonucleotides, natural products, and drug
metabolites with picomole to femtomole sensitivity (4). Today, MS has been
revolutionary in spurring research in protein biochemistry, glycobiology, and
biotechnology, and most recently in DNA sequencing (5).

Protein analytical techniques are essential for determining a protein’s primary
structure as well as detecting and defining posttranslational modifications of proteins so
that the complete structure of a protein can be Obtained (6, 7). Nucleotide sequences of
genes embody the information required to deduce the primary structure (amino acid

sequence) of proteins, but they do not reveal whether the side chains of amino acids are

chemically modified in viva after translation of genes or the extent of any modifications.
Defining posttranslational modifications is essential in the case Of expressed proteins that
are defined for use as pharmaceuticals. This is required to ensure that they have the same
chemical structure as their natural counterparts or an acceptable alternate structure.
Unacceptable structures could be detrimental to recipients due to unwanted
pharmaceutical or immunological activities.

Defining the locations and structures of posttranslational modifications in proteins
is one of the most important contributions that mass spectrometry has made and
continues to make in the field of protein biochemistry (8). The unique potential of MS is
most evident when it is used in conjunction with conventional methodologies like
automated protein sequence and amino acid analysis. In contrast to the latter two
approaches in which identification is based entirely on chromatographic retention relative
to one of the 20 commonly occurring amino acids or a derivative thereof, MS relies on
measurement of molecular mass, an intrinsic physical property. Mass spectrometry is
clearly the method of choice for characterization of posttranslationally modified proteins,
since in most cases well-developed, accurate, and sensitive chemical or biochemical
approaches are not available.

Disulfide bond formation after gene expression is one of the most important
posttranslational modifications frequently encountered in protein characterization (9, 10).
Mass spectrometry, combined with other chemical/biochemical techniques, has been
extensively applied to the recognition of both free sulflrydryls and disulfide bonds in

proteins.

In this thesis, novel methodologies for recognizing fi'ee sulfhydryl groups and
disulfide bonds will be discussed in Chapter 2, and a study of reagents for the cyanylation
reaction will be described in detail in Chapter 3. The Objective of this chapter is to
introduce basic aspects of MALDI and current methodologies for characterizing cysteine

sulfhydryl status in proteins.

11. Characterization of Cysteine and Disulfide Bond Status in Proteins

Cysteine contributes to protein biological functions by using its free sulfhydryl
group (-SH) as the active site for enzyme catalysis (11). Sulfhydryl groups can also play
an important role in a variety of physiological and biochemical processes (1 1). Disulfide
bonds, on the other hand, comprise the major covalent cross-linkage in proteins and
maybe intra- or inter—molecular in nature. Intra-chain disulfide bonds serve to confer
conformational stability on the folded polypeptide chain. While inter-chain disulfide
bonds are functional in maintaining the quaternary structure of multi-chain proteins,
serving as the only linkage between subunits or providing covalent stability to structures
otherwise maintained by non-covalent forces. Both free cysteines and disulfides,

therefore, play a unique role in the activity of proteins (9).

A Localization of Sulfhydryl Groups

1. Classical Approach
The classical approach for locating free sulflrydryl groups in proteins involves

several steps. The first step is to modify fiee sulfhydryl groups, usually by an irreversible

reaction such as alkylation, under conditions that can prevent or minimize
sulfhydryl/disulfide exchange. Second, the protein is cleaved by enzymes or chemical
reagents between cysteine residues. Third, the digestion mixture is separated by
chromatography. Finally, the derivatized sulfhydryl-containing peptides are identified,
mapped to sequence by the Edman technique and/or mass spectrometry, and related to
specific segments of the protein.

A key to the success of this approach is to choose an appropriate derivatization
reaction and reagent for sulfhydryl groups. The reagent should be soluble in the reaction
medium and label sulfhydryl groups selectively, rapidly, and irreversibly under mild
conditions (ideally, under weak acidic conditions to avoid sulflrydryl/disulfide bond
exchange which is minimized at pH 2-6.5). Furthermore, the reagent should possess
strong UV or fluorescent absorption which does not overlap with the maximum
absorption of proteins, or it should easily attach a radioactive element to facilitate the
identification of derivatized peptides after HPLC separation. Finally, the derivative of
cysteine should be distinguishable fiom other amino acids by the Edman degradation
technique in which the identification of amino acids exclusively relies on the retention
time of PTH-derivatives.

There are many reagents that had been used in the modification of cysteine groups.
Alkylation of the free sulflrydryl group is one of the most traditional ways. This is usually
achieved by reaction with compounds containing an alkyl halide group such as
iodoacetate, iodoacetamide, bromoacetate, etc., among which iodoacetamide shows the
highest reactivity and specificity to SH, alkylation can be accomplished by converting the

SH to S-carboxymethylcysteine (Figure 1.1) (11-13). The iodide liberated can be

determined by an electrochemical method (14) or by absorbance at 226nm (15). TO

avoid the formation of iodine by photooxidation of iodide ions, the carboxymethylation is

performed in the dark.
SH COOH
I
CH CH2
I2 O + ICHZCOOH —> g + HI
in—'2 I
—NH—CH—C—

Figure 1.1 The modification of cysteine with iodoacetic acid to form S-carboxymethylcysteine

This commonly used cysteine alkylation method does not yield material which is
easily amenable to Edman sequence analysis. This is either because of solubility
problems with the alkylated proteins and peptides or because the 3-phenyl-2-
thiohydantoin (PTH) derivatives (formed by Edman degradation ) of the alkylated
cysteine residues cannot be readily identified by stande procedures such as gas-liquid
chrrnatography. Because of these reasons, some other alkylation reagents, such as 4-
vinyl pyridine (4-VP), which yields the pyridylethylcysteine (PECys) derivative were
developed. Alkylation with 4-VP provides several advantages. For instance, the PECys
residues are stable to hydrolysis in 6N HCl, and they increase the solubility of peptides at
neutral and lower pH, etc. (16). Although this provides another choice of alkylation
reagent, iodoacetate is still the most popular derivatization reagent for modifying

sulflrydryl groups because of its simplicity and specificity (17-18).

Two derivatives of 7-fluoro-2,l,3-benzoxadazole (19), 7-fluoro-2,1,2-benzoxadiazole
4-sulfonate (SBD-F) and 4-(aminosulfonyl)-7-fluoro-2,1,3-benzoxadiazole (ABD-F),
have proven useful for the assay of sulflrydryls by reason of their formation of fluorescent

products ( Figure 1.2).

F F SR SR
N N / N N
/ / \ \ ¢ \
’ \0 O o RSH 0 O o
O / - / \ /
\\ N / \\N Sodium Borate §N N
so; NH: so2 NH2 80, Na * so 2 NH2
SBD-F ABD-F SBD-SR ABD—SR

Figure 1.2 Derivatization of the cysteine group by SBD-F or ABD—F

The adducts exhibit fluorescence at a long wavelength (~515nm) and show a spectral
shift corresponding to environmental hydrophobicity (20, 21). The reaction is performed
under borate buffer, pH 9.5, containing lmM EDTA to prevent metal-catalyzed oxidation
Of the sulflrydryls. For the derivatization of sulflrydryls in large proteins, the proteins
must be denatured to ensure that every SH can be attacked by the reagents (22, 23).
Organomercurial compounds are high affinity and specific reagents for SH
groups. In contrast to alkylation, the mercaptidation reaction is reversible. Among
various organomercurial reagents, monofunctional organomercurial compounds have
been widely exploited. Spectrophotometric titration with p-mercuribenzoate (Figure 1.3)
proposed by Boyer is the most preferred because of its simplicity, high sensitivity,

selectivity and precision (24).

COO-

©
Hg.

Figure 1.3 p-mercuribenzoate

Another method to modify free SH groups is based on sulflrydryl/disulfide exchange.
Among oxidants of SH groups, disulfides occupies a special position since their
reactions with sulfhydryls are absolutely specific. This is referred to as
sulflrydryl/disulfide exchange. Among the various disulfides proposed, Ellman’s reagent
5,5’-dithio-bis-(2-nitrobenzoic acid), DTNB (Figure 1.4), has received the widest use
(25). Grassetti and Murray (26, 27) demonstrated that 2,2’-dithiodipyridine or 4,4’-
dithiodipyridine were good substituents for Ellman’s reagent. These compounds showed
strong reactivity to SH groups in a wide pH range (pH 1-8) (28). The particular value of
pyridyl disulfides as reactivity probes and labelling reagents is that they are two-protonic-
state electrophiles, i.e., they markedly increased their electrophilicity when protonated on

a pyridyl nitrogen atom.

COOH COOH

WOO

Figure 1.4 Ellman's Reagent: 5,5'-dithio-bis-(Z-nitrobenzoic acid)

The other method used in the modification of —SH groups is oxidation. Sun and
Smith used perforrnic acid to modify certain residues and thereby cause a characteristic
change in the peptide molecular weight. This change in molecular weight is determined
by FABMS (fast-atom-bombardment mass spectrometry) and used to help identify
peptides. Results for a series of small peptides demonstrate that Cys, Met, and Trp are
the only residues that undergo a change in molecular weight under the conditions they
used. The molecular weight of a peptide increases by 48 Da for each cysteinyl-
containing peptide (29). When treating with perfomic acid, destruction of tryptophan and
partial destruction of tyrosine will occur, thus the practical application of this approach is
limited.

Chin and Wold (30) employed the combination of tributylphosphine (Bu3P) to
reduce disulfide bonds and of 4-(aminosulfonyl)-7-fluoro-2,1,3-benzoxadiazole (ABD-F)
to block free sulfhydryls to characterize both free sulfhydryls and disulfide bonds in a
number of proteins. Since the two reagents apparently do not react with each other, their
combination offers a convenient and quite general method for the complete
characterization of free Cys (-SH) and cross-linked Cys (S-S) in proteins. In their paper,
they demonstrated their strategy on model compounds (e.g., Cys and glutathione),
proteins such as enolase from yeast and rabbit muscle, containing only fiee Cys, and a
protein, superoxide dismutase, containing both free and crosslinked Cys. Using a similar
procedure, Kirley (31, 32) determined the location of three disulfide bonds and one free
sulfhydryl group in the B-subunit of (Na+, KU-ATPase.

Another reagent, 5-[2-((iodoacetyl)amino)-ethyl]naphthalene-l-sulfonic acid

(IAEDANS) (Figure 1.5) has been proven powerful for sulfhydryl derivatization (33-35).

Recently, Sturrock (36) et al used this reagent to label a free cysteine residue in human
testicular angiotensin-converting enzyme (tACE). After isolating the fluorescent peptide
from enzymatic digests by HPLC, the sequence of the fluorescent peptides was mapped
by MALDI-PSD ( matrix assisted laser desorption/ionization - post source decay). The

sequence data established that Cys496 is a free thiol in tACE.

ll
lH,CC-NH-CH,CH2-NH

so,H
Figure 1.5 5-[2-((iodoacetyl)amino)—ethyl] napthalene-l-sulfonic acid (IAEDAN S)

2. Cleavage at Cysteine Residues
This method utilizes a cleavage reaction at cysteine residues. The conversion of

an SH group into an SCN group was first achieved in two stages by successive treatment
of a protein with Ellman’s reagent and then with cyanide (37, 38). Catsimpoolas (37)
showed that the cyanide cleavage of the disulfide bond yields a sulfhydryl and a
thiocyano group. In pH ranges below 8, the latter cyclizes to produce an
acyliminothiazolidine moiety on the peptide chain. This product is unstable and
hydrolyzes spontaneously to effect cleavage of the peptide link on the amino group.
Stark (39) showed that 2-nitro-5-thiocyanobenzoic acid (NTCB) (Figure 1.6) specifically
cyanylates cysteine sulflrydryls, which subsequently cleave at the N-terminal side of the
cyanylated cysteinyl residues under mildly alkaline conditions to form an amino-terminal

peptide and series of 2-iminothiazolidine-4-carboxyl (ITC) peptides. If a protein contains

n cysteine residues, the cleavage reaction results in the formation of n+1 peptide
fragments, mass assignment of which indicates the number and location of cysteine
residues. While potentially a very useful method, it has seldom been used for sequence
determination because all but the N-terminal peptide is blocked by the ITC group, which
is not amenable to Edman degradation (40, 41) . Papayannopoulos and Biemann (42)

first used CID tandem mass spectrometry to sequence the NTCB cleavage reaction

0
N02 C N
NTCB CDAP

Figure 1.6 Structures of 2-nitro-5-thiocyanobenzoic acid (NTCB) and
1-cyano-4-dimethylamino-pyridium salt (CDAP)

products of a protease inhibitor isolated from Sarcophaga bullata. Mass spectrometry
provides much more accuracy than SDS page for mass assignment and can be applied to
the sequencing of ITC blocked peptides. To date, the NTCB approach has primarily been
employed to locate total cysteines in various proteins although selective identification of
free cysteine in the presence of disulfide bonds was also reported (42-47).

Another cyanylation reagent that had been used is 1-cyano-4-dimethylamino-

pyridinium salt (CDAP) ( Figure 1.6). Reported by Wakselman (48) this reagent was

10

synthesized from 4-dimethylaminopyridine. CDAP was found to be reactive toward
protein sulflrydryl groups in neutral or acidic medium, then the cyanylated product can
be cleaved under alkaline conditions (3 7). It was also reported that this reaction is highly
specific; papain when reacting with CDAP at pH 3.6 for 11 minutes, loses more than
98% of its catalytic activity. The cyanylated product is isolable and this reaction has no
oxidative side reactions.

In this thesis, more cyanylation reagents will be investigated, and the chemistry and

mechanism of the cyanylation reactions will be discussed in detail in Chapter 3.

B. Localization of Disulfide Bonds

Although the amino acid sequences of proteins are readily deduced from the
corresponding cDNA, modifications occuring to the protein after translation cannot be
predicted accurately. Although there are good methods for quantifying the number of
disulfide bonds in proteins, the unambiguous determination Of the locations of disulfide
bonds remains a difficult task.

Two most often used terms in cysteine chemistry are sulfl'iydryl/disulfide exchange
and disulfide bond scrambling (or disulfide bond interchange). Although
sulflrydryl/disulfide exchange (such as Ellman’s reagent) is the most specific reaction,
both sulfhydryl/disulfide exchange and disulfide bond scrambling in proteins occur in
mildly alkaline media when structures permit and result in mismatched nonnative
disulfide bond structures (Figure 1.7). While under acidic conditions from pH 2 to pH
6.5, this sulfhydryl/disulfide exchange and disulfide bond scrambling reactions can be

minimized (49, 50).

11

'S
S\S

 

Figure 1.7 Illustration of sulfhydryl/disulfide exchange in proteins

Classical Strategy

As illustrated in Figure 1.8, a classical approach (51, 52) in locating the disulfide
bond linkage includes three steps. First, a protein is cleaved by enzymes or chemical
reagents between half-cystinyl residues under conditions that can avoid or minimize
disulfide scrambling. Second, the digestion mixture is separated by reversed-phase
HPLC. Third, the amino acid sequence or molecular masses of these peptides are
determined using Edman degradation or mass spectrometry. Finally, the disulfide
linkages are deduced, but this method is usually time-consuming. The other traditional
method is enzymatic digestion. The method usually requires an enzyme to be Specific so
that it can produce well-defined fragments that can simplify the identification of the
disulfide bond. It is very likely that some proteins cannot be cleaved by specific cleavage
reagents; thus, they must be cleaved by non-specific reagents. However, when using
nonspecific reagents, it is difficult to relate the disulfide-bridged peptides to specific

segments of the protein simply by determining their molecular masses by MS. .

12

Recently, more methods have been developed for the identification of disulfide-
containing peptides, especially various mass spectrometric methods. A new methodology

developed in Watson’s research group will be discussed in detail in Chapter 2.

Multiple step enzymatic/chemical cleavage of protein

Fractionate peptides by HPLC

l

Reduce disulfide bonds

l

Modification of nascent SH groups

Separation of the resulting peptides

l

Sequencing by Edman degradation

Figure 1.8 Classical strategy for disulfide bond assignment of a protein

13

III Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MS)

Lasers originally found applications in mass spectrometry in the early 19703 for
desorption/ionization of small organic compounds that contained appropriate
chromophores (53). Abrupt energy absorption initiates a phase change in a rrricrovolume
of the absorbing sample fiom a solid to a gas while also inducing ionization. In the late
19808, research into the desorption/ionization of nonvolatile materials for mass
spectrometry has fostered the development of matrix-assisted laser desorption ionization,
MALDI (54-56). Although the first attempt to generate ions of organic molecules by
direct laser desorption/ionization dates back to early 19708 (57). However, the size of the
analytes which can be desorbed and ionized was limited to ~1000 daltons. When larger
molecules are analyzed, more energy is required to desorb them. However, such a high
energy flux leads to pyrolytic or photochemical decomposition of analytes. The
breakthrough came in the late 19803 when Hillenkamp and Karas successfully
experimented with the use of a matrix, nicotinic acid (58, 59). It was discovered that
desorption/ionization of large, nonvolatile molecules such as proteins could be similarly
obtained when such samples were irradiated by a laser after being codeposited with a
large molar excess of an energy-absorbing “matirx” material, even though the analyte did
not strongly absorb the wavelength of the laser radiation. Initial experiments with this
technique were conducted with frequency-quadrupled Nd-YAG lasers that emit at
266nm. The ejected ions were analyzed by a time-Of-flight (TOF) mass spectrometer.
After that MALDI has quickly gained notoriety as an analytical tool that can accurately

determined the molecular weight of proteins up to (and exceeding) 100kDa from low-

14

picomole amounts of sample (60). Despite the lack of an established ionization
mechanism, the dependence on the matrix material in the desorption/ionization process
gave rise to the terminology matrix-assisted laser desorption/ionization (MALDI). Today
MALDI has developed into an extremely powerful technique for accurate and sensitive
analysis of molecular masses extending to greater than 200,000 Da. With the
introduction of two new techniques, post-source decay (PSD) and delayed extraction
(DE), MALDI-TOP MS is also becoming an effective tool for structure elucidation of
biopolymers.

All the work that will be discussed later was performed on the MALDI instrument
(PerSeptive Biosystem Voyager Elite) available in MSU/NIH Mass Spectrometry
Facility. This instrument can be operated in the reflectron mode to acquire PSD spectra
for structural analysis, and it also has the delayed extraction function which improves its

resolution.

A. The Matrix

For the most commonly employed laser wavelengths (337 or 355nm), several
hundred different organic compounds, selected for their light absorbing properties, have
been investigated for use as MALDI matrices (61, 62). However, only a few of these are
widely applicable, e.g., sinapinic acid (SA), 2,5-dihydroxybenzoic acid (DHB) and or-
cyano-4-hydroxycinnamic acid (a-CHCA) (63, 64). This is largely because features
other than absorption of the appropriate wavelength of laser radiation are necessary for a
compound to be a useful MALDI matrix. Other important matrix characteristics include

formation of a single phase with the analyte in the solid phase, solubility in the same

15

solvents required for dissolution of the analyte, vacuum compatibility (low vapor
pressure), a chemical composition that promotes ionization (matrix substituents that can
donate protons to the analyte), nonreactivity with the analyte, and other physical
properties such as low heat of sublimation and a capacity to crystallize readily (65).
Some of these criteria may explain the variable success in analyzing different classes of
macromolecules (i.e., small peptides, proteins, glycoproteins, polysaccharides, or
oligonucleotides) by MALDI using different matrices.

The physicochemical events leading to the transfer of proteins to the gas phase
and their ionization in MALDI have not yet been fully elucidated (66). The matrix is
believed to serve to minimize sample fragmentation from the laser beam by absorbing the
incident laser energy, resulting in the sample and matrix molecules being ejected into the
gas phase. One model for the mechanism assumes that the uppermost layers of matrix
are induced to undergo a phase transition from the solid to the gas phase. The subsequent
expansion of these matrix molecules into the vacuum drags the matrix-isolated protein
molecules into the gas phase. During the transfer to the gas phase, the protein undergoes
ionization through proton transfer reactions with the matrix by a reaction process that

remains to be explained.

B. Sample Preparation
Proper sample preparation is critical for successful analysis by MALDI-MS (60).
Matrix solutions are prepared in water, water-acetonitrile, or water-alcohol mixtures at a
concentration of about 5-10ug/ul, depending on the solubility properties of the matrix.

The analyte is prepared at a concentration of about 0.1ug/ul in a solvent that is miscible

l6

with the matrix solution (for peptides or proteins, aqueous 0.1% trifluoroacetic acid
(TFA) is frequently used). The matrix and analyte solutions are mixed to give a final
matrix : analyte molar ratio of approximately 10,000:1 and a final volume of 0.5-2ul.
The mixture is applied to a stainless steel MALDI-MS probe tip, and allowed to dry by
either ambient evaporation, heating with a stream of warm air, or under vacuum. During
the drying process, the matrix codeposits from solution with the analyte. The nature of
the matrix-analyte interaction has not yet been defined, but it may involve analyte
molecules that have become embedded in the matrix crystal lattice or on the surface of
rapidly forming matrix crystals. Studies based on optical microscopy have shown that
matrix crystal formation and size vary, depending on the matrix and solvent, but seem to
be independent of the analyte. The decrease in the “quality” of matrix crystal
morphology (where the matrix-analyte deposit appears as a powder, film, or
noncrystalline glassy surface) leads to a lower quality MALDI mass spectrum. These
conditions can result from the presence of high levels of contaminating salts, buffers, or
surfactants.

Several factors affect the sensitivity of a MALDI-MS experiment. In many cases,
analysis of smaller amounts of sample yields a better MALDI signal. This may simply be
the result of approaching a more Optimum molar ratio between the matrix and the analyte,
especially with macromolecules whose masses exceeds lOkDa. Alternatively, dilution of
a contaminant which can interfere with analyte signal acquisition may also lead to
improved spectra.

In this thesis, the experiment is performed on the Voyager Elite system which

uses a plate holder that can accommodate 100 samples and permit a much higher sample

17

throughput. During the drying process, the matrix codeposits from solution with the
analytes. The sample is inserted into the vacuum chamber by a probe and a vacuum lock
in the VT2000, and by an “automatic mechanical han ” in the Voyager Elite MALDI-
TOF mass spectrometer, and irradiated with short duration(l to lOns) pulses of an UV
laser beam.

To date, the chemistry of “matrix-assistance” remains incompletely understood;
the choice of matrix and method of application is still empirical. Sample preparation by
different matrices and solvents (69), matrix additives (70), and evaporation rate (71,72)
affects the resolution and sensitivity of MALDI. Optimal results require parallel analyses

under different conditions.

C. Instrumentation

Time-of-flight (TOF) mass spectrometers (60) are instruments noted for their nearly
unlimited mass range and high sensitivity. In these mass analyzers, ions are formed by a
pulsed ionization process (pulsed laser radiation for MALDI) in a short (~50m) source
region containing an electric field as shown in Figure 1.9. The electric field accelerates
the ions into a long (~1m) field-free drift region. The time required for ions to traverse
the drift region (TOF) is dependent on their mass and is described by the relationship:
TOF = L/v = L(m/22e10”2, where L is the length of the flight tube, v is the ion velocity, m
is the mass of the ion, 2 is the charge state of the ion, and V is the acceleration potential.
Thus, low-mass ions have a shorter flight time than heavier ions. The clock used to
measure the TOF of the MALDI ions is triggered by the laser pulse. All the ions that are

formed from a single laser pulse and that are accelerated by the electric field give rise to a

18

transient TOF signal from the detector at the end of the flight tube. To improve ion
counting statistics (60), multiple transient TOF mass spectra can be summed from 10 to
several hundred laser pulses. Time-to-mass conversion is achieved by incorporating the
tof of ions of known mass (from internal or external standards) into a calibration
algorithm from which mass values can be computed for unknown ions with
experimentally determined flight times.

Additional virtues of the time-of-flight instrument are simplicity and low cost.
However, because the TOP recorded for a particular ion reflects many different initial
conditions experienced in the ion source (such as the time/location of ion formation and
initial kinetic energy distributions), the mass resolution is disappointingly low (73). The
poor resolving power (defined as the capacity to distinguish ions differing slightly in
mass) is a result of peak broadening which increases with the mass of the ionized
macromolecule. This limits the capacity to detect certain protein modifications and/or
protein sequence variations, especially at higher masses. In spite of this, it is possible to
determine the mass of MALDI-TOF ions with an accuracy up to +/- 0.01% for proteins
with molecular masses between 1 and 40 kDa, and with somewhat poorer accuracy for

proteins above 40 kDa (60).

D. Novel Techniques in MALDI-TOF MS
MALDI has been a superior method for molecular mass determination of proteins
and peptides, and other biological materials, but lacks the capacity for structural analysis,
because it is a soft ionization technique. This deficiency has been overcome since the

introduction of post-source decay (PSD) analysis of MALDI-generated ions. The PSD

l9

(74,75) technique is based on mass analysis of product ions from unimolecular or
collision-induced decay taking place in the field-free region between the ion source and
the reflectron. In contrast to a linear instrument, reflection instruments can be used as an
as an energy analyzer to differentiate ions that are otherwise detected as species with the
same velocity in the linear mode. Ions formed as a result of metastable decomposition of
fully accelerated precursors are detected at the same arrival time (same apparent mass) as
their precursor in the linear mode of TOF because they have the same velocity.
However, because the product ions have lower kinetic energy, they can often be resolved
in the reflectron mode of the TOF instrument by lowering the potential of the reflector.
Therefore, PSD, like MS/MS techniques, can provide full or partial sequence information
of medium sized peptides (<2500 Da). Since its commercialization about seven years
ago, MALDI-PSD has quickly evolved into a powerful technique for sequence

determination of peptide at the low picomole scale.

E. Applications of MALDI-TOF MS
Over the past few years, MALDI has become among the most powerful methods yet
available for macromolecular characterization of living systems. A wide range of
achievements in proteins, nucleotides, and glycobiology have been attained in a short
time.
MALDI has been extensively used for the determination of primary covalent

structures of proteins derived from both natural and recombinant sources (76). MALDI

20

Beam Splitter

::::g1%333_

 

 

 

 

 

 

 

 

 

    

/ .
/ ' l Transrent
/
/ 1:}
/// Trigger I Recorder
//
'/
Sample // ”I I I—
\
Probe ~‘ (+9) 6% gee '.
TH
. \ [\l\\ FlightTube h
' \ ' Detector
I Ion \\
ISourcel \
\\
\
Probe Tl

Figure 1.9 Scheme of matrix-assisted laser desorption/ionization time-Of-flight mass

spectrometer (MALDI-TOF MS)

21

provides routine and reliable means to analyzing tryptic digests and glycoproteins which
typically have a high degree of heterogeneity (77). The high tolerance of MALDI to
extraneous compounds has been used to study the degradation of hemoglobin by a
malaria parasite. This remarkable achievement shows direct analysis of cell contents is
now within the reach of mass spectrometry (78).

MALDI-MS is an excellent method when dealing with protein posttranslational
modifications. Covalently modified protein N- and C- termini (79), disulfide bonds (80,
81), phosphorylation (82), glycosylation (83), lipidation (84) and protein DNA
interactions (85) have all been studied by MALDI.

In addition to the use of MALDI as a sequencing tool, its ability toward the analysis
of combinatorial libraries has also been explored.

The capacity for mass spectrometric analysis of oligonucleotides has generally
lagged behind that of oligopeptides. However, this position has changed markedly in the
last few years (86, 87), especially with the development of new matrices. Investigation
of a 50-mer is now somewhat routine (88, 89) and there are individual reports of the
analysis of much larger RNAs (90).

It is believed that MALDI-MS will play more and more important roles in biological
studies. As is ofien true in a new and rapidly developing field, the most exciting

breakthroughs may, in the end, occur in areas not even anticipated at the present time.

22

111. References:

1. Thorgersen, D. F., Skowronski, R. P., and Macfarlane, R. D., Biochem. Biophys.
Res. Commun, 60, 616 (1974)

2. Karas, M., and Hillenkamp, F., Anal. Chem, 60, 2299-2301 (1988)

3. Meng, C. K., Mann, M., and Fenn, J. B., Proc. 36th ASMS Conference on Mass
Spectrometry and Allied Topics, San Francisco, CA, (1988), p771.

4. Siuzdak, G., Proc. Natl. Acd. Sci., 91, 11290 (1994)

5. Burlingame, A. L., Boyd, R. K., and Gaskell, S. J., Anal. Chem, 68, 599R-651R
(1996)

6. Carr, S. A., Roberts, G. D., and Hemling, M. E., in Mass Spectrometry of Biological
Materials, C. N. McEwen and B. S. Larsen (Editors), Marcel Dekker, New York,
(1990), Ch. 3, p87.

7. Yan, S. C. B., Grinnell, B. W., and Wold, F., Trends Biochem. Sci., 14, 264 (1989).

8. Carr, S. A., and Biemann, K., Methods Enzymol., 106, 29-58 (1984)

9. Creighton, T. E., in Mechanism of Protein Folding, Edited by Pain, R. H., IRL Press,
(1992)

10. Hirayama, K., and Akashi, S., in Biological Mass Spectrometry: Present and Future,
Edited by Matsuo, T., Caprioli, R. M., Gross, M. L., and Seyama, Y., John Wiley &
Son Ltd, (1994)

11. Torchinsky, Y. M., in Sulfitr in Protein, (Metzler, D., Translation Ed.), Pergamon,
Oxford, (1981)

12. Jocelyn, P. C., Biochemistry of the SH Group, Academic Press, London, (1974)

23

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29

30

Jocelyn, P. C., in Methods in Enzymol., Vol. 143, Academic Press, edited by Jakoby,
W. B., Griffith, O. W. Inc., (1987)

Stark, G. R., and Stein, W. H., J. Biol. Chem, 239, 3755 (1964)

Swoboda, G., and Hasselbach, W., Hoppe-Seyler ’s Z. Physiol. Chem, 354,
1611(1973)

Mak, A. S., and Jones, B. L., Anal. Biochem., 84, 432 (1978)

Makinen, A. L., and Nowak, T., J. Biol. Chem, 264, 12148 (1989)

Wang, Z. X., Preiss, B., and Tsou, C. L., Biochemistry, 27, 5095 (1988)

Toyo’oka, T., and Imai, K., Anal. Chem, 56, 2461-2464 (1984)

Toyo’oka, T., and Imai, K., Analyst, 109, 1003 (1984)

Sueyoshi, T., Miyata, T., Iwanaga, S., Toyo’oka, T., and Imai, K., J. Biochem.
(Tokyo), 97, 1811 (1985)

Toyo’oka, T., and Imai, K., Anal. Chem, 57, 1931 (1985)

Kobayashi, T., Kudo, 1., Karasawa, K., Mizushima, H., Inoue, K., and Nojirna, S., J.
Biochem. (Tokyo), 98, 1017 (1985)

Boyer, P. D., J. Am. Chem. Soc, 76, 4331 (1954)

Ellman, G. L., Arch. Biochem. Biophys., 82, 70 (1959)

Grassetti, D. R., and Murray, J. F ., Arch. Biochem. Biophys, 119, 41 (1967)
Grassetti, D. R., and Murray, J. F ., J. Chromatogr. 41, 121 (1969)

Stuchbury, T., Shipton, M., Norris, R., Malthouse, J. P. G., Brocklehurst, K., Herbert,
J. A. L., and Suschitzky, H., Biochem. J., 151, 417 ( 1975)

Sun, Y., Smith, D. L., Anal. Biochem., 172, 130-138 (1988)

Chin, C. C. Q., and Wold, F., Anal. Biochem., 214, 128-134 (1993)

24

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

Kirley, T. L., J. Biol. Chem, 264, 7185-7192 (1989)

Kirley, T. L., Anal. Biochem., 180, 231-236 (1989)

Pardo, J. P., and Slayman, C. W., J. Biol. Chem, 264, 9373 (1989)

Miyanishi, T., and Borejdo, J ., Biochemistry, 28, 1287 (1989)

Bishop, J. E., Squier, T. C., Bigelow, D. J., and Insesi, G., Biochemistry, 27, 5233
(1988)

Sturrock, E. D., Yu, X. C., Wu, Z., Biemann, K., and Riordan, J. F., Biochemistry, 35,
9560-9566 (1996)

Catsimpoolas, N., and Wood, J. L., J. Biol. Chem, 241, 1790 (1966)

Degani, Y., and Patchornik, A., Biochemistry, 13, l (1974)

Stark, G. R., Methods in Enzymology, Vol. 47, 129-132 ( 1977)

Biemann, K., in Methods in Protein Sequence Analysis, Plenum Press, New York,
(1993), p119

Schaffer, M. H., and Stark, G. R., Biochem. Biophys. Res. Commun., 71, 1040-1047
(1976)

Papayannopoulos, I. A., and Biemann, K., Protein Sci. 1, 278-288 (1992)

Mahboub, 8., Richard, C., Delacourte, A., and Han, K., Anal. Biochem., 154, 171-182
(1986)

Altamirano, M. M., Plumbridge, J. A., and Calcagno, M. L., Biochem., 31, 1153-1158
(1992)

Tang, C., Yuksel, K. U., Jacobson, T. M., and Gracy, R. W., Arch. Biochem. Phys.,

283, 12-19, (1990)

46. Nefsky, B., and Bretscher, A., Proc. Natl. Acad. Sci. USA, 86, 3549-3553 ( 1989)

25

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

Minard, P., Desmadril, M., Ballery, N., Perahia, D., and Mouaward, L., Eur. J.
Biochem., 185, 419-423 (1989)

Wakeselman, M., and Guibe-Jampel, E., J. C. S. Chem. Comm, 21, 1976

Lumper, L., and Zahn, H., Adv. Enzymol., 27, 199 (1965)

Spackman, D. H., Stein, W. H., and Moore, 8., J. Biol. Chem, 235, 648-659 (1960)
Staswick, P. E., Hermodson, M. A., and Nielson, N. C., J. Biol. Chem, 259, 13431
(1984)

Haeffner-Gormley, L., Parente, L., and Wetaufer, D. B., Int. J. Pept. Protein Res., 26,
83 (1985)

Cotter, R. J., Anal. Chem, 56, 485 (1984)

Hillenkamp, F., Karas, M., Beavis, R. C., and Chait, B. T., Anal. Chem, 63, 1193
(1991)

Hillenkamp, F., Karas, M., Methods in Enzymology, Academic Press, New York, Vol.
193,p280

Karas, M., Bahr, U., and Giessman, U., Mass Spectrom. Rev., 10, 335 (1991)
Posthurnus, M. A., Kistemaker, P. G., Meuzelaar, H. L. C., and Brauw, M. C., Anal.
Chem, 50, 985 (1978)

Karas, M., and Hillenkamp, F., Anal. Chem, 60, 2299-2301 (1988)

Karas, M., Bachmann, D., Bahr, U., and Hillenkamp, F ., Int. J. Mass Spectrom Ion
Processes, 78, 53 (1987)

Zaluzec, E., Gage, D., Watson, J. T., Protein Expression and Purification, 6, 109
(1995)

Ehring, H., Karas, M., and Hillenkamp, F., Org. Mass Spectrom, 27, 472 (1992)

26

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

Fitzgerald, M. C., Parr, G. R., and Smith, L. M., Anal. Chem, 65, 3204 (1993)
Beavis, R. C., and Chait, B. T., Rapid Commun. Mass Spectrom, 3, 432 (1989)
Beavis, R. C., Chaudhary, T., and Chait, B. T., Org. Mass Spectrom, 27, 156 (1992)
Juhasz, P., Costello, C. E., and Biemann, K., J. Am. Soc. Mass Spec., 4, 399 (1993)
Chait, B. T., and Kent, S. B. H., Science, 257, 1885-1894 (1992)

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

Busch, K. L., J. Mass Spectrom, 30, 233 (1995)

Cohen, S. L., and Chait, B. T., Anal. Chem, 68, 31 (1996)

Preston, L. M., Murray, K. K., and Russell, D. H., Biol. Mass Spectrom, 22, 544
(1993)

Vorm, 0., and Roepstorff, P., and Mann, M., Anal. Chem, 66, 3281 (1994)

Vorm, O., and Mann, M., J. Am. Soc. Mass Spectrom, 5, 955 (1994)

Cotter, R. J ., Biomed. Environ. Mass Spectrom, 18, 513 (1989)

Kaufrnann, R., Kirsch, D., and Spengler, 3., Int. J. Mass Spectrom. Ion Proc., 131,
355 (1994)

Kaufrnann, R., Spengler, B., Rapid Commun. Mass Spectrom, 7, 902 (1993)

Carr, S. A., in Mass Spectrometry in the Biological Sciences, Burlingame, A. L., Carr,
S. A., Eds.; Humana Press: Totowa, NJ, (1996)

Billeci, T. M., and Stults, J. T., Anal. Chem, 65, 1709 (1993)

Goldberg, D. E., Slater, A. F. G., Beavis, B. T., Chait, B. T., Cerami, A., and
Henderson, G. B., J. Exp. Med., 173, 961 (1991)

Specht, B., Oudenarnpsen-Kruger, E., Ingendoh, A., Hillenkamp, F., Leaius, A. G.,

and Spener, F., J. Biotechnol, 33, 259 (1994)

27

80. Robertson, J. G., Adams, G. W., Medzihradszky, K. F., Burlingame, A. L., and
Villafranca, J. J ., Biochemistry, 33, 11563 (1994)

81. Chang, J. Y., Schindler, P., and Chatrenet, B., J. Biol. Chem, 270, 11992 ( 1995)

82. Liao, P-C., Leykam, J ., Andrew, P.C., Gage, D., and Allison, J ., Anal. Biochem., 219,
9 (1994)

83. Burlingame, A. L., Curr. Opin. Biotechnol., 7, 4 (1996)

84. Wilcox, M. D., Schey, K. L. Busman, M., and Hildebrandt, J. D., Biochem. Biophys.
Res. Commun., 212, 367 (1995)

85. Zhao, Y., and Chait, B. T., Anal. Chem, 66, 3723 (1994)

86. Nordhoff, E., Kirpekar, F., and Roepstorff, P., Mass Spectrom. Rev., 15, 67 (1996)

87. Lirnbach, P. A., Crain, P. F., and McCloskey, J. A., Curr. Opin. Biotechnol., 6, 96
(1995)

88. Nordhoff, E., Kirpekar, F., Karas, M., Cramer, R., Hahner, S., Hillenkamp, F.,
Kristiansen, K., Roepstorff, P., Lezius, A., Nucleic Acids Res., 22, 2460 (1994)

89. Nordhoff, E., Karas, M., Crarner, R., Hahner, S., Hillenkamp, F., Kirpekar, F.,
Lezius, A., Muth, J ., Meier, C., and Engles, J. W., J. Mass Spectrom, 30, 99 (1995)

90. Kirpekar, F ., Nordhoff, E., Kristiansen, K., Roepstorff, P., Lezius, A., Hahner, S.,

Karas, M., and Hillenkamp, F., Nucleic Acids Res., 22, 3866 (1994)

28

Chapter II
Novel methodologies for the assignment of both cysteine

residues and disulfide bond pairings in proteins

1. A strategy to locate cysteine residues in proteins by specific chemical

cleavages followed by mass mapping by MALDI-TOF MS

A. Introduction

Cysteine contributes to protein biological functions by using its free sulflrydryl
(SH) group in the active site for enzyme catalysis, as the chelating site for metal ions, and
as the active site of disulfide reshuffling enzymes. Pinpointing the number and location
of cysteine residues thus becomes an important strategy for determining protein structure.
While it is true that the large majority of new protein sequences are now deduced from
nucleotide sequences of clones cDNA (or, in the case Of prokaryotic proteins, from
cloned genomic DNA fragments), differentiating between free cysteine residues
(sulflrydryl groups) and cystines (disulfide bonds) is still cumbersome and tedious;
furthermore, the cDNA technique, like other conventional techniques, cannot identify
posttranslational modifications.

Classical approaches for localizing protein free sulflrydryl groups usually involve
modification of free sulflrydryls before and after reduction of a protein, degradation of the
derivatized protein chemically or enzymatically into smaller peptide fragments, followed

by HPLC fractionation of the peptides and Edman degradation of the derivatized peptides

29

(1,2). Recently, peptide mapping by mass spectrometry of proteolytic digests before and
after HPLC fractionation also has been reported (3-5). However, this procedure requires
the sulflrydryl derivatization under a carefully controlled conditions so that sulfhydryl-
disulfide bond exchange is not an issue. Secondly, this approach requires that a protein
be cleaved at both sides of cysteine residues to give peptides that contain only one
cysteine. Multiple—step enzymatic or chemical digestions usually are performed in this
case which is tedious and requires a sample size on the order of nanomoles. Thirdly,
derivatizing reagents must be chosen to facilitate HPLC separation with UV detection
and to avoid the retention time overlap of derivatized cysteine and other amino acids on
an Edman sequencer. Although other approaches such as methods based on affinity
chromatography of sulflrydryl groups have also been proposed, their feasibility remains

to be tested.

B. Cyanylation of Cysteine Residues

One of the most widely used techniques for examination of structure-fimction
relationships in proteins is that of chemical modification Of amino acid side chains. The
role of thiol groups in proteins has been particularly well studied, mainly because of the
availability of a number of specific thiol-modifying reagents such as organo-mercurials
(Boyer, 1954) and 5,5’-dithio-bis-(2-nitrobenzoic acid), DTNB (Ellman, 1959). Jacobson
et al. (6) showed that 2-nitro-5-thiocyanobenzoic acid (NTCB) selectively cyanylates
cysteine's free sulflrydryl group under mild conditions. Here in this methodology, we
chose NTCB as the cyanylation reagent. Since the reactivity of NTCB toward a thiol

group is related to the easy displacement by the sulfur nucleophile of the good leaving

30

group p-nitrothiophenolate. The cyanylation is selective for free sulfhydryl groups,
disulfide bonds are unreactive to NTCB. The selective cyanylation and subsequent
cleavage of free sulflrydryls can thus be achieved in the presence Of disulfide bonds by
carrying out the reaction without prior reduction of the protein (6). Other reagents for
cyanylation of SH groups have been prepared by Wakselman (7) and by Brocklehurst et
a1 (8, 9). These reagents may offer some advantages over NTCB, but their utility needs
to be firrther examined. Further research on cyanylation reagents was done in our

laboratory as discussed in Chapter 3.

C. Chemical Cleavages at Cysteine Residues

NTCB cleavage of peptide chains at the N-peptide bonds of cysteine residues was
frrst observed by Catsimpoolas and Wood (10). However, because of the several side
reactions, such as the reversibility of the cyanolysis and the elimination of thiocyanylate,
the yields were low and the cleavage reaction was not acceptable to protein chemists.
Further research on NTCB by Jacobson (6) showed that after cyanylation reaction by
NTCB, subsequent cleavage occurs on the N-terrninal side of the cyanylated cysteinyl
residue under mildly alkaline conditions to form an amino-terminal peptide and a series
of 2-iminothiazolidine-4-carboxylyl (ITC) peptides (Figure 2.1). If a protein contains n
cysteine residues, the cleavage reaction results in the formation of n+1 peptide fragments,
and mass analysis of the fragments indicates the number and location of cysteine
residues.

Although the original paper claimed that the cleavage reaction can come to

completion with little side reactions for most of the peptides and proteins tested, Degani

31

and Patchornik (11) found that a B-elimination reaction also occurred depending on the
structural properties of the respective peptides.

Jacobson et a1 (6) proposed a mechanism for the cleavage reaction. This
mechanism implicates that hydroxide ion catalyzes the cleavage, as shown by the
influence of rising pH on the rate of cleavage reaction. A possible explanation of the role
of OH' is shown in Figure 2.2. The amide nitrogen is apparently too weak a base to carry
out the nucleophilic attack (the pKa of a protonated amide is 4.0 or less), and the reaction
probably proceeds only after attack of OH' on the carbonyl carbon of the amide. This
generates a much more basic nitrogen, which then can participate in the concerted
cyclization and cleavage, without formation of an acyliminotlriazolidine intermediate.

Since both cleavage and B-elimination occur under mildly alkaline conditions, the
two reactions are competitive, sometimes resulting in only a moderate degree of cleavage
of a peptide bond. Another inherent problem of the cleavage reaction is that it takes too
long to accomplish. Various side reactions may rise after long hours of exposure to
alkaline solution, which brings about some uncertainty into the analytical results. From a
systematic study done by Watson’s group (17) on the cleavage reaction as a function of
pH, solvents and amino acid structures for polypeptides containing different amino acids
adjacent to N-terminus of cyanylated cysteines, we know that higher pH can greatly
promote both cleavage and B—elimination reactions, but the extent of B—elirnination does
not increase significantly even at a pH as high as 12. Ammonia, a stronger nucleophile
than hydroxyl anions, greatly accelerates the cleavage of peptide chains and minimizes
side reactions related to the prolonged incubation. The Optimized conditions were

applied to the recognition of free sulflrydryl groups in a variety of peptides.

32

SIH
+ II I II //
H,N —-—C—NH—CH—C—NH---C

\ -

(A) Cyanyhtion NTCB, HOOC
pH 8-9; :1, 15' o,N SCN

N-C-S\
+ II I II //
H,N —-—C—NH-CH-C-NH--—C
\O-

...J L -

OH
B Clea H
“year \‘rm‘ Wm” '“
+ [/0 HNYS (II) 9 ('3le (u)
H3N ---C-OH + HN C-NH— —-—C—NH-C — C—--

Figure 2.1 Reaction between cysteine residue and 2-nitro-5-thiocyanobenzoic acid
(NTCB), (A) Cyanylation and (B) Cleavage Reaction. B-elimination competes with
cleavage reaction under alkaline conditions.

33

N-C—s
I? 9H2 ‘3 .0
H,N*—-—C—NH-CH—C—NH—-—C
\O-

l
Hr»

. s
- C/ 0 ,O
19 'mr—lh—NH—«de

OH

OH
O HN S O
. /, o
HN—--C + Y x,
3 ‘orr rmJ—&‘NH"‘C.O.

Figure 2.2 A mechanism for the base-catalyzed cleavage reaction

34

II. A Strategy for Assignment of Disulfide Bond Pairings in Proteins
A. Introduction

Many biological and therapeutically important proteins (peptides) contain
disulfide bonds. Because the disulfide bond is an important element of protein structure,
it is necessary to know the locations of disulfide bonds in order to understand more fully
any unique contributions they may make to protein (peptide) structure and function.
Although there are good methods for quantifying the number of disulfide bonds in
proteins, the unambiguous determination of the location or pairing of disulfide bonds
continues to challenge protein chemists.

Current methodology (3, 12) for assignment Of disulfide bonds in proteins
involves cleavage of protein chains between half-cystinyl residues with specific cleavage
reagents, such as cyanogen bromide or proteases, such as trypsin, to Obtain peptides that
contain only one disulfide bond. The resulting mixture of peptides is separated, and the
amino acid compositions, sequences or molecular masses of the peptides are determined
by Edman degradation or mass spectrometry or both. Assignment of these peptides to
specific segments of the protein leads to the recognition of disulfide crosslinkages.

Although this approach is well established and has been used with much success,
it is limited in many aspects. First, disulfide-containing peptides can be identified by
their amino acid composition or sequence only if they are purified to homogeneity. This
requirement may not be achieved in the case of large proteins, or when the quantity of
protein is very small. Second, the above methodology requires the protein chain to be
cleaved between every half-cystinyl residue so that the resulting peptides contain no more

than one disulfide pair. However, even multi-step enzymatic or chemical digestions

35

rarely yield a full set of diagnostic fragments because proteins are often resistant to some
cleavage reagents. It will be even more difficult or impossible to find an enzyme which
is likely to cleave between two closely spaced or adjacent cysteines (3). It is likewise
important to note that the conditions that are fiequently used to cleave a protein chain are
also the conditions that frequently lead to disulfide scrambling (13). As a result, the
location of disulfide bonds determined by this approach may not be the same as the
location of disulfide bonds in the native protein. The problem of disulfide exchange
plagues all attempts to locate disulfide bonds in proteins, irrespective of the general
method currently used. Although supplementary methodologies such as non-specific
fragmentation by partial acid hydrolysis (14) have been proposed to avoid disulfide bond
scrambling, it is difficult to deal with the data obtained by these non-specific techniques.
Gray (15, 16) has described an approach for analyzing disulfide linkage patterns
in highly bridged small peptides with close or adjacent cysteine residues. In his
experiments, peptides were partially reduced under controlled conditions, the isomers of
the partially reduced protein separated by HPLC, the nascent free thiols alkylated, and the
positions of alkylated cysteines recognized from the results of sequence analysis and
related to the disulfide bond pair that had been reduced and cyanylated. However, it is
Obviously tedious, if not impractical, to sequence an alkylated high-mass peptide or

protein using this approach.

B. A Novel Strategy for Disulfide Assignment

Watson’s group has developed a simple methodology for recognizing the location

of free cysteine groups in peptides and proteins (17). This approach employs a specific

36

chemical reaction between sulflrydryls and 2-nitro-5-thiocyanobenzoic acid (NT CB) to
selectively cyanylate cysteine thiols (18). The N-terminal peptide bond Of the modified
cysteinyl residue can then be cleaved under alkaline conditions to form an amino-
terrninal peptide and a series of 2-iminothiazolidine-4—carboxyl (ITC) peptides which can
be mass mapped to the sequence of the original molecule by MALDI-MS. Disulfide
bonds do not react with NTCB and therefore do not interfere with the determination of
sulfhydryl groups. The Optimized cleavage conditions greatly facilitated the cleavage and
make this methodology much more attractive in terms of speed of the analysis and
interpretation of the data.

Another novel strategy (outlined in Figure 2.3) developed by Watson’s group for
the assignment of disulfide bond pairings in proteins using the above specific chemical
cleavage of partially reduced and cyanylated protein isomers with mass mapping of the
resulting peptides by MALDI-TOF MS (19). In this methodology, as shown in Figure
2.4 for a simple protein containing only two disulfide bonds, the denatured protein is
partially reduced by tris(2-carboxyethyl)phosphine (TCEP) in a buffer solution at pH 3.0
to produce a mixture of residual intact protein and isomers of partially reduced species.
Conditions can be Optimized so that the predominant products are isomers in which only
a single disulfide bond has been reduced . Nascent sulflrydryls are immediately
cyanylated by 1-cyano-4-dimethylaminO-pyridinium tetrafluoroborate (CDAP) (20)
under the same buffer conditions as used for reduction. The partially reduced and
cyanylated protein isomers are then separated by reversed-phase HPLC, followed by
analysis of HPLC fractions by MALDI-MS to determine which isomers are singly

reduced/cyanylated. Those shified by +52 Da correspond to a singly reduced/cyanylated

37

Denatured Protein

partial reduction lTCEP, pH 3

Partially Reduced Protein Isomers

cyanylation l CDAP, pH 3

Cyanylated Protein Isomers

HPLC separation 1

Fraction of Protein lsomers—pldentified by MALDI- MS

cleavage l 1M NH4OH

reduction I TCEP
A Mixture of Peptide Fragments

l

MALDI-MS

Figure 2.3 Descriptive overview of the cyanylation/cleavage methodology for
assignment of disulfide bond pairings in proteins.

38

1 | 20 1 4o

 

10 I 30 | 50
S——S

l partial reduction; TCEP @ pH3

 

 

 

 

 

 

 

 

Intact SH 20 SH 40 81—20 SI 40 Completely Reduced
Protein 1 10 I - 30 | 5° 4' 1 10 1 30 5° + Specres
———- SH SH
l cyanylation; CDAP @ pH3
Intact $CN 20 ?CN 40 IS T SI 40 Completely Reduced /
Protein 1 10 I 30 1 5° + 1 10 1 30 l 50 + Cyanylated Specres
S — S SCN SCN
HPLC fractionation
identification by MALDI-MS
cleavage; 1N NH 4OH
1 9 1 10 19
20 S— S
ITC1n é S 29 ITC 2°_—'L_30 39
ITC ao——15— 50 ITC 40—— so
i reduction; TCEP
1 9 SH
1 -"—10 19
ITC o 20 SH
1 éH 79 ITC 20_—30—- 39
SH
ITC 3°'_l—"4o 50 ITC 40—— so
i MS analysis
( CystO - Cys 30) ( Cyszo - Cys40 )

Figure 2.4 Chemical overview of the cyanylation/cleavage methodology. Partial
reduction means that the protein of interest is reduced under controlled conditions into a
mixture of isoforms, each of which corresponds to reduction of only one of its disulfide
bonds; for a protein containing n cystines, n isoforms of the singly-reduced protein will
result. ITC stands for iminothiazolidinyl carboxyl residue at the amino terminus.

39

species; a shift of +104 Da corresponds to a doubly reduced/cyanylated species, etc.
These isomers with a 52-Da mass shifi from the mass of the intact protein are dried and
subjected to specific chemical cleavage in aqueous ammonia. The cleaved peptides,
which may be linked by residual disulfide bonds, are then completely reduced to give a
mixture of peptides that can be mass-mapped by MALDI-MS. The masses of the
resulting peptide fragments are related to the location of the paired cysteines that had
undergone reduction, cyanylation, and cleavage.

A primary advantage of this approach is its underlying simplicity. The data
obtained from this approach are straightforward and easily interpreted because only a few
protein isomers are produced, and each is relevant. Typically, only n-l intermediates are
needed to define an n-bridge system. Redundant information is frequently obtained to
confirm an assignment. Secondly, we describe a practical way to circumvent problems
related to disulfide bond scrambling because both reduction and cyanylation are
performed in an acidic media. Finally, our approach can be used for the assignment of
adjacent or closely spaced cysteines for which conventional methodology fails.
Therefore, two problems associated with current methodologies are solved using this new
methodology.

In Chapter 3, the cyanylation reagent, CDAP, is compared with other reagents;

also, the results of studies of cyanylation reactions are summarized.

40

III.

9.

References

. Dabre, A., Practical Protein Chemistry: A Handbook, John Wiley & Sons, (1985)

p.149.

Chin, C. C. Q., and Wold, F., Anal Biochem., 214, 128 (1993)

Hirayama, K., and Akashi, S., in Biological Mass Spectrometry: Present and Future
(Matsuo, T., Caprioli, R. M., Gross, M. L., and Seyama, Y., Eds.), John Wiley &
Sons Ltd., (1994). P. 299-312

prertson, J. G., Adams, G. W., Medzihradszky, K. F., Burlinggame, A. L., and
Villafranca, J. J ., Biochemistry, 33, 11563 (1994)

Glocker, M. 0., Arbogast, B., Milley, R., Cowgill, C., and Deinzer, M. L., Proc. Natl.
Acad. Sci. USA., 91, 5868 (1994)

Jacobson, G. R., Schaffer, M. H., Stark, G. R., and Vanaman, T. C., J. Biol. Chem,
248, 6583 (1973)

Wakeselman, M., Guibe-Jampel, E., Raoult, A., and Busse, W. D., J. Chem. Soc.
Chem. Commun., 1, 21 (1976)

Brocklehurst, K., Malthouse, J. P. G., Baines, B. S., Blenkinsop, R. D., Churcher, J.
A., Mushiri, M. S., and Ormerod, F., Biochem. Soc. Transactions, 6, 261 (1978)

Brocklehurst, K., Malthouse, J. P. G., Biochem. J., 175, 761 (1978)

10. Catsimpoolas, N., and Wood, J. L., J. Biol. Chem, 241, 1790 (1966)

11. Degani, Y., and Patchornik, A., Biochemistry, 13, 1 (1974)

12. Smith, D. L., and Zhou, Z., Methods Enzymol, 193, 374 (1990)

13. Ryle, A. P., and Sanger, F., Biochem. J., 60, 535 (1955)

14. Zhou, Z., and Smith, D. L., J. Protein Chem, 9, 523 (1990)

41

15. Gray, W. R., Protein Sci., 2, 1732 (1993)

16. Gray, W. R., Protein Sci., 2, 1749 (1993)

17. Wu, J., Gage, D. A., and Watson, J. T., Anal. Biochem., 235, 161 (1996)

18. Jacobson, G. R., Schaffer, M. H., Stark, G. R., and Vanaman, T. C., J. Biol. Chem,
248, 6583 (1973)

19. Wu, J. and Watson, J. T., Protein Sci., 6, 391 (1997)

42

Chapter 111 Study on Cyanylation Reagents

I. Introduction

Among the 20 amino acids that compose proteins, cysteine has unique properties.
For a new protein with unknown sequence, the quantitative determination of cysteine
content is very beneficial to further characterization of the protein.

The cysteine sulflrydryl group is distinguished fiom other side chains of amino
acids by its high reactivity and by the exceptionally diverse chemical reactions in which it
participates, such as alkylation, acylation, oxidation, sulflrydryl-disulfide bond exchange,
charge-transfer complexes, reactions with organic heavy metal compounds (1). SO there
are many ways for the modification of free sulflrydryl group. In the strategy developed
by Watson’s group as discussed in Chapter 2, modification of cysteine’s free sulflrydryl
group is the first step. Research on sulflrydryl modification can be dated back to the

1950s’ (2) and several reagents were involved in the modification of the cysteine group.

A. Organic Mercurial Reagents:
Organic mercurial compounds are high affinity, specific reagents for the protein
SH group. They have been widely used in studies of the biochemical function of
sulflrydryl groups since the pioneering research of Hellerman (3) with p-mercuribenzoate
(p-MB). Spectrophotometric measurement of the extent and rate of reaction of p-
mercuribenzoate with sulflrydryl groups proposed by Boyer (2) stands in first place

because of its sensitivity, simplicity, selectivity, and precision. This classical method is

43

selective because although combination of p-mercuribenzoate with other groups or
compounds in the protein preparations may occur, only the combination with the -SH
groups results in a large increase in the molecular absorptivity at 250nm. Spectral
measurements may thus serve to distinguish between combinations Ofp-mercuribenzoate
with -SH groups and other groups (2) (Figure 3.1). Interestingly, the rate and the extent
of the reaction of SH groups of many proteins with p-MB is higher at pH 4.6 than at pH
7-8 because of the high affinity of p-MB for hydroxide ions (4). As a result, the
reactivity of the reagent increases as the pH of the solution is decreased. Other reagents,
such as 1-(4-chloromercuriphenylzaO)-napthOl-2 (CMPN; mercury-orange), and p-
(hydroxymercuri)-benzoic acid (5), were also proposed to improve the

Spectrophotometric detection of SH groups in lysine monoxygenase (6) and in

carboxyhemoglobin (7).
COO -
COO SH
l ”9 +
+ C 2 0| —-—> S + H
I
—NH —CH — c — I
Hg+ CH2

I O
—NH—CH—Ii—

Figure 3.1 Chemical reaction between p-mercuribenzoate and sulflrydryl group

B. Sulfhydryl / Disulfide Exchange:

Among oxidants of SH groups, disulfides occupy a special position since their
reactions with sulfhydryls are absolutely specific. This reaction is referred to as
sulflrydryl/disulfide exchange. As can be seen from the below equations, this reaction
consists of two steps of nucleophilic substitution with the formation of a mixed disulfide
at the intermediate stage.

RISSR1+ RSH---- RlSH + RISSR Equation 3.1
RISSR + RSH ---- RISH + RSSR Equation 3.2

Among the various disulfides proposed, Ellman’s reagent, 5,5’-dithio-bis-(2-
nitrobenzoic acid), DTNB, had received the widest application (8). The reaction of
thiolate anion with excess Ellman’s reagent at pH 7-8 is favored toward the
stoichiometric formation Of TNB thiolate (5-thiO-2-nitrobenzoate, TNB') and a mixed

disulfide (RS-TNB) (Figure 3.2 ).

NO 2 NO 2
COOH COO H
¢ at
SII'I SH
Crz r + i —+ + o
—NH —CH —C — | OOH
O CIHZ fl NC’2
OOH —NH —CH —C —
NO2
(m NB) (RS- TNB) (TNB)

Figure 3.2 The reaction of Ellman’s reagent, 5,5’-dithiO-bis-(2-nitrobenzoic acid), with

cysteinyl residues in proteins

45

More work was done by Vanaman and Stark (9) on this reagent. They proposed
the use of enzyme-thionitrobenzoate mixed disulfides as intermediates in the preparation
of specific derivatives of enzyme -SH groups and the potential of the use of enzyme
thiocyanates in specific chemical cleavage of polypeptide chains at cysteine and cystine
residues. They found that the reaction of the sulfhydryl group with DTNB and other

symmetrical disulfides are different. (Figure 3.3).

S—S
_ co2 002
002

002

DTNB 3-carboxypropyldisulfide

Figure 3.3 Structure of DTNB and 3-carboxypropyldisulfide

Even the interatomic distances and bond angles in the above two compounds are slightly
different, DTNB and 3-carboxypropyldisulfide can achieve similar orientations. So the
failure of 3-carboxypropyldisulfide to react with sulflrydryl group reflects the large
intrinsic difference between the reactivity of an aliphatic disulfide and DTNB (9). The
reaction of DTNB with free -SH in proteins can be summarized as follows: it forms a
mixed disulfide with concomitant release of 1 mole of TNB per mole of -SH as shown in
Figure 3.2 (8). The position of equilibrium for this disulfide interchange reaction is very

much in the direction of the mixed disulfide, since the aromatic mercaptide anion has the

46

low pKa of 5.1 (8). The use of enzyme-TNB mixed disulfide to achieve specific and
reversible chemical modification of -SH groups in proteins seems to be very promising.

Reaction of disulfides with cyanide has been used by Wood and Catsimpoolas
(10) as a route to specific chemical cleavage Of peptides which contain cystine. The
unsymmetrical disulfide enzyme-TNB is cleaved by cyanide in only one way, i.e., to
release TNB. For specific cleavage at cystine and cysteine residues, one rrright (a) reduce
the protein disulfides with reducing agent, (b) allow all Of the -SH groups to react with
excess DTNB, and (c) add excess cyanide at pH 7. The product is an N-terrninal
iminothiazolidine ring.

Later on, another reagent 2,2’-dipyridyl disulfide (2-Py-S-S-2-Py) (Compound I)
is more often used than 5,5’-dithio-bis-(2-nitrobenzoic acid) because its reactions with
thiols are essentially stoichiometric over a wide range of pH (Grassetti & Murray, 1967;
Brocklehurst & Little, 1973) (11,12,13). The particular value of pyridyl disulfides as
reactivity probes and labelling reagents is that they are two-protonic-state electrophiles,
i.e., they markedly increase their electrophilicity when protonated on a pyridyl nitrogen
atom (e.g.,compound 11). Mixed disulfides containing the 2-pyridyl moiety (compound
111) have found application in the purification of thiol enzymes by covalent
chromatography. The mixed disulfides retain the ability to act as two-protonic—state
electrophiles and thus permit selective isolation or labelling of low-pKa thiols in acidic
media. The range of such electrophilic reagents is being extended to provide variation in
solubility, ionization properties and spectroscopic characteristics and in the nature of the
electrophilic center such as the cyanylating reagent 2-thiocyanopyridine (Compound IV).

This reagent provides thiol derivatives that are sterically relatively undemanding so that it

47

may possess advantages relative to 2-nitro-5-thiocyanatobenzoic acid (NTCB) analogous
to those that compound I possess relative to 5,5’-dithio-bis-(2-nitrobenzoic acid) (DTNB)

(11,12).

H +
@—s—s—@ @‘rs—s—‘ce

(Compound I)

(Compound II)
N N
CO,_._._. @r—s—m
(Compound Ill) (Compound IV)

Figure 3.4 Cysteine Modification Reagents

C. NTCB (2-nitro-5-thiocyanobenzoic acid) as a reagent:

The 2-nitro-5-thiocyanobenzoic acid mentioned above is a widely used reagent
for quantitative cyanylation of the cysteine residues (14). Its reaction is shown in Figure
3.5. NTCB is shown to be a potential tool for the selective chemical cleavage of peptide
chains at the N-terrninal of the cyanylated cysteinyl residues. Earlier reports (Wood and
Catsimpoolas, 1963; Catsimpoolas and Wood, 1966) (10,15) showed that B-
thiocyanoalanine residues when generated by excess cyanide treatment of cystine
disulfide bonds, underwent cyclization to form 2-iminothiazolidine rings, with

consequent cleavage of N-peptide bond. (For a review, see Spande et al., 1970) (16).

48

S-

SCN

RSH+ QC ——> RSCN + o
I oo’ oo'
no2

NO 2
Figure 3.5 Reaction of NTCB with SH

D. CDAP (l-cyano-4-dimethylamino-pyridium salt) as a reagent:

Another reagent called 1-cyano-4-dimethylamino-pyridinium (CDAP) perchlorate
or fluoroborate is also used as a cyanylation reagent ( Wakselman, 1976) (17). Its
reaction is shown in Figure 3.6. According to Wakselman, CDAP reacts rapidly with
thiols in neutral and acidic medium: in 11 nrinutes at pH 3.6, 98% of the catalytic activity
of papain is inhibited; Cys 7 and Cys 19 of the reduced B-chain of bovine insulin are
quantitatively cyanylated at pH 3.5, then the N-peptide bonds are selectively cleaved at
pH 9.5. This water soluble reagent was also found to react with the protein sulflrydryl
group in a wide pH range, from pH 2-7, and it can lead to isolable thiocyanates without

oxidative Side reactions. Then the protein chain may be selectively cleaved in alkali.

49

NM62 NMe

RSH + (AU ——> RSCN + 0
II“ II“
CN H

CDAP

Figure 3.6 Reaction of CDAP with a generic free sulflrydryl group(RSH)

This chapter will focus more on other cyanylation reagents such as 2-
thiocyanopyridine and 4-(dimethylamino) phenyl thiocyanate and compare them with the
traditional reagents such as NTCB and CDAP. Although other reagents mentioned above
can modify the free sulfhydryl group in proteins and peptides, they either can only be
used under alkaline conditions which may cause disulfide bond scrambling or there is no
specific cleavage reaction for the modified cysteine residues in order for us to use mass
spectrometry to analyze them. Our laboratory uses both NTCB (18) and CDAP (19) to
modify cysteine residues, but there are limitations Of these two reagents. NTCB can only
be used at pH 8 to 9, so that disulfide scrambling is very likely to happen. CDAP seems
to be a good choice since it can be used at pH 3, and the reaction is fast and specific.
However, for some peptides and proteins which have very tight structures, their cysteine
groups may be difficult to cyanylate because of their poor accessibility. In other cases,
we might want to increase the pH of the reaction or raise the reaction temperature so that
poorly accessible cysteine groups can be cyanylated. CDAP is very unstable above pH 4,
having a very short half life; for this reason, it is necessary to make a flesh CDAP

solution for the cyanylation reaction. Unlike CDAP, PYRSCN(2-thiocyanopyridine) and

50

4-(dimethylamino) phenyl thiocyanate are very stable and they have similar structures to
NTCB. We expect PYRSCN and 4-(dimethylamino) phenyl thiocyanate can be used in a

wide pH range. More research was done on these two NTCB series of reagents.

11. Materials and Methods:
A. Chemicals:

Guanidine hydrochloride was a product of Boehringer-Mannheim Biochemicals
(Indianapolis, IN), citric acid, sodium citrate, l-cyano-4-dimethylaminO-pyridinium
tetrafluoroborate (CDAP) were purchased from Sigma and used without further
purification. Acetonitrile and TFA were of HPLC grade. 2-nitro-5-thiocyanobenzoic acid
(N TCB) and Trizrna buffer were purchased from Sigma and used without further
purification. The 0.01M solution of NTCB was prepared in either 0.1M Tris-HCI or the
buffer made in 6M guanidine-HCI solution and the pH of the solution was adjusted to pH
8.0. The 0.10M CDAP solution in pH 3.0, 0.1M citrate buffer-4M guanidine-HCI was
freshly prepared prior to use. Tris(2-carboxyethyl)phosphine (TCEP) hydrochloride was
purchased from Pierce Chemical co. (Rockford, IL). The TCEP solution in 0.1M citrate
buffer at pH 3.0 was prepared as 0.01 M stock solution and stored under N2 at ~20C for
weeks with little deterioration. Rabbit muscle creatine phosphokinase was purchased
from Sigma and used without firrther purification.

Peptide YEKPLQNFTLCFR (MW 1657.1), peptide RGPCRAFI (MW 919.1) and

peptide RY V VLPRPVCFEKGMNYTVR (MW 2427 .9) were purchased from Bachem

51

California Inc.. The lnmol/ml protein and peptide solutions were prepared in 0.1M

citrate buffer, pH 3.0, containing 6 M guanidine-HCI.

B. Synthesis of 2-thiocyanopyridine (PYRSCN)

2 grams of 2-mercaptopyridine and 2.9-3.6 grams of NaCN was suspended in 30-
40 ml methanol saturated with KBr (20). This mixture was cooled to ~5C and stirred
while a solution of 1.32 ml Br; in 7.4 ml methanol was added dropwise. The reaction
mixture was then poured into ice-water and the product was extracted with ethyl ether.
The upper layer of the extract was collected in a 50 ml flask, and a rotovaporator was
used to vaporize the ether. The residual solution in the flask is the PYRSCN reagent; the
color is dark brown.

Purification of PYRSCN was done by first adding some anhydrous sodium sulfate
to absorb water that was left in the solution and removal by filtration. Then 10ml of
methylene dichloride was added to dissolve it. A column was prepared, with some glass
wool on the bottom, then 0.3cm anhydrous sodium sulfate, then 5 cm silica gel was added
on the top. 10 ml of methylene dichloride was used to wet the column, and then
PYRSCN solution was poured onto the column. The first 10 ml of the eluent was
discarded; the following eluent was collected. The impurities elute slower than the pure
compound, so they were left on the top part of the column. Use nitrogen stream to
evaporate the methylene dichloride from the solution. After purification, the color of

PYRSCN is light yellow. It is not soluble in water, but soluble in 2:1 actonitrilezwater.

52

C. Detection of PYRSCN by GC/MS:

GC/MS was done on a Varian Saturn 2000. GC conditions were: 35C for 3
minutes, then ramp to 260C at 15C/min. Solvent delay was set for 5 nrinutes. An aliquot
of 0.5ul PYRSCN reagent was dissolved in 24ml of chloroform, and a lul sample was
injected. A molecular ion peak and a fragment peak corresponding to loss of SCN were

both detected, identifying the PYRSCN reagent.

D. Mass Spectrometry:

MALDI mass spectra were obtained on a Voyager Elite time-of-flight (T OF)
mass spectrometer ( PerSeptive Biosystems Inc., Framingham, MA) equipped with a
model VSL-337ND nitrogen laser (Laser Science, Newton, MA). The accelerating
voltage in the ion source was set to 25 KV. Data were acquired in the positive or
negative linear mode of operation. Time-to-mass conversion was achieved by external
calibration using standards of bradykinin (m/z 1061.2), bovine pancreatic insulin (m/z
5734.5) obtained from Sigma Chemical Co. ( St. Louis, MO). All experiments were
performed using or-cyano-4-hydroxycinnamic acid (Aldrich Chemical Co., Milwaukee,
WI) as the matrix. Saturated matrix solutions were prepared in a 50% (v/v) solution of
acetonitrile/aqueous 0.1% TFA, and mixed in equal volumes with peptide or protein
samples, and applied to a stairrless-steel sample plate. The mixture was allowed to air dry

before being introduced into the mass spectrometer.

53

E. Cyanylation of Nascent Sulfhydryls:

For CDAP: to the partially reduced protein mixture was added a 10-fold molar
excess of CDAP solution over the total cysteine content. Cyanylation of the nascent
sulflrydryl groups was accomplished by incubation at room temperature for another 10-15
minutes.

For NTCB: to the partially reduced protein rrrixture was added a 10-fold molar
excess of NTCB solution over the total cysteine content. Cyanylation of the nascent
sulflrydryl groups was accomplished by incubation at 37C for 30 minutes.

For PYRSCN and 4-(dimethylarnino) phenyl thiocyanate: to the partially reduced
protein mixture was added a 50-fold molar excess of PYRSCN solution or 4-
(dimethylamino) phenyl thiocyanate solution over the total cysteine content. Cyanylation
of the nascent sulflrydryl groups was accomplished by incubation at room temperature for

30 minutes.

F. HPLC Separation of Partially Reduced and Cyanylated Protein Isomers:
Partially reduced and cyanylated species were separated by reversed-phase HPLC
with linear gradient elution using Waters model 6000 pumps controlled by a PC
computer. UV detection was at 215nm. Mobile phase are acetonitrile and water. The
columns were a Vydac C18 (#214TP54, 10-um particle size, 300-A pore, 4.6 x 250mm) .
The major HPLC fractions were collected manually and the masses of the collected
protein isomers were determined by MALDI-MS. Appropriate fractions were then dried

for further use. The sample size of 10 nanomoles was used for convenient detection from

54

conventional HPLC columns; the use of microbore columns should allow the use of

much smaller sample sizes.

G. Cleavages of singly reduced and cyanylated protein isomers
To the dried HPLC fraction was added 1 M NH40H. Cleavage of the peptide
chain was performed at room temperature at pH 11 for one hour. Excess ammonia was

removed in a vacuum system.

H. Complete reduction of remaining disulfide bonds
Truncated peptides, still linked by residual disulfide bonds, were completely
reduced by reacting with 2ul of 0.1M TCEP solution at 37C for 30 minutes at pH 3-5 to
minimize the possibility of reoxidation. Samples were diluted with 100ul of 50% (v/v)

acetonitrile/0.1% TFA solution prior to analysis by MALDI-MS.

III. Results and Discussion:
A. Locate the free sulfhydryl groups in rabbit muscle creatine phosphokinase
using PYRSCN as the cyanylation reagent:
Creatine Phosphokinase is a protein containing 380 amino acid residues and 4 free
sulfhydryl groups at positions 73, 145, 253 and 282.
The cyanylation reaction was conducted on this protein using a 50-fold of excess
PYRSCN reagent at pH 7, room temperature for 30 minutes. The modified protein was
purified by RP-HPLC using gradient elution to remove excess reagent and solvent, then

cleavage reactions were conducted on HPLC fractions 1-3 using 1M ammonia at pH 11

55

for one hour. Sample was then dried and pH was adjusted to pH 3 using citrate buffer,
and then take this to MALDI for detection.

The HPLC chromatogram showed 3 peaks (Figure 3.7). The first peak gave no
signal above m/z 1000 in MALDI mass spectrum, so it might be due to some
interference from the PYRSCN reagent. The second and the third peak gave
complementary data. MALDI data are shown below (Figure 3.8) for the cleavage
products of the cyanylated material represented by peak 3 in the HPLC chromatogram.
From the mass spectrum of the cleavage products of peak 2 the protein with only its
cysteine 73 and 145 cyanylated, there are three peaks: m/z 8236.02 is the mass of
residues 1 to 72, m/z 8165.21 is the mass of the iminothiazolidine (ITC) derivative of
residue 73 to 144, and m/z 4117.95 is the doubly charged ion Of m/z 8236.02. This
indicates that the second HPLC peak is the cyanylation product with only cysteine 73 and
145 cyanylated. The mass spectrum of the third HPLC peak gave a fill] set of data, that
is, this cyanylation peak is the peak with all four cysteines cyanylated. On one hand this
tells us that the accessibility of each Of the four free sulflrydryl groups in creatine
phosphokinase is different, on the other hand since CDAP cyanylation of creatine
phosphokinase only gave us a single cyanylation peak on HPLC which suggests that
CDAP is a stronger cyanylation reagent than PYRSCN.

From the data shown below (Table 3.1) we can successfully deduce the free
sulfhydryls’ locations in creatine phosphokinase. Although PYRSCN is not as reactive
as CDAP toward free sulflrydryl groups, it still cyanylates the protein as indicated by the
MALDI spectrum in Figure 3.8. PYRSCN seems to be an effective cyanylating reagent

around pH 7.

56

mv

 

1800.00

1600.00-

1400.00.1
1200.00
1000.00-
800.00 3

600.00 %

 

400.00
1

200.00; I n
o_oo.;.__.l Ix..._.I

2463101214161820222426
Minutes

 

 

 

 

 

 

 

 

Figure 3.7 HPLC chromatogram of cyanylated creatine phosphokinase

73 145 253 282
1 ---GCV---HCS---FCV---TCP---3 80

 

 

 

 

 

 

 

Fragment Calculated m/z [M+H]+ Observed m/z [Mi-HP-
1-72 8234.2 8228.03
73-144 8161.0 8160.52
145-252 12625 12614
253-281 3414.0 3411.76
282-380 10719 10716.5
253-380 14055 13377

 

 

 

 

 

Table 3.1 Calculated and observed masses for cleavage products of creatine phosphokinase

57

 

 

 

 

 

I
soooo~ .7.
\I
N
N
25000- :3
9°
8
20000. 3
a»
E 3” s;
§ 15000- g»; t 3
o _. as oo ...
U 2" :3 g Q a N
a: s; a g t. a
a .. a s
5000 3: E 5
'5 Xi
\I
0.1
4o'oo 6000 sobo Ioooo' rzboo

 

Mass (m/z)

Figure 3.8 MALDI spectrum of cleavage product of creatine phosphokinase HPLC peak 3 in
Figure 3.7

B. pH effect on cyanylation reaction:
Cyanylation reactions were done on three model peptides : peptide 919, peptide
1657 and peptide 2428, according to molecular weight, calculated and observed values
are shown in Table 3.2. Experiments were conducted by taking aliquots of a 1 nmol/ul
peptide solution and adding to it a 50-fold molar excess of the cyanylation reagent
solution. React under room temperature for 30 minutes. Both pH 7 and pH 8 were tried
on the three model peptides. Then HPLC was used to separate the products. In each

cyanylation mixture; three HPLC peaks were observed, each peak was integrated and

percentage of each one was given below in Table 3.3, Table 3.4 and Table 3.5.

58

Table 3.2 Calculated and observed masses for cyanylation product of peptide 919,

 

 

 

 

1657 and 2427.9
Cyanylated Product Mixed Disulfide Dimer
[NI-Hr" [M+H1 [M+H]*
Peptide Sequence MW. Cal. Observed Cal. Obsewed Cal. Observe
RGPCRAFI 919.3 945.3 945.53 1029.3 1029.42 1837-6 1336-2
YEKPLQNFTIDFR 1657.1 1683.1 1683.18 1767.1 1767.33 3313.2 3312.24
RYWLPRPwFEKGMNYTVR 2427.9 2453.9 2454.43 2537.9 2539.7 4854.8 4657.3

 

 

For all the peptides reacting with PYRSCN shown in Figure 3.9, Figure 3.11 and Figure

3.13: HPLC peak 1 is the cyanylated product, peak 2 is the mixed disulfide product and

peak 3 is the dimer peak.

59

1. Results of Study on Peptide 2428:

Table 3.3 Product ratios of peptide 2428 after reacting with PYRSCN under both

 

 

 

pH 7 and pH 8.
pH Percent of Cyanylated Peptide Percent of Mixed Product Percent of Dimer
pH7 68% 14% 12%
pH8 50% 7% 43%

 

 

 

 

 

 

HPLC chromatograms of cyanylation products of peptide 2428.

Reaction products

 

 

Reaction products
at pH 7 at pH 8
. l
1
3
2 3
2
#A—J J—A—J

14.00 16.00 18.00

—v—.——v—w—-v——-r—r—v——v——v—-—
14.00 16.00 18.00 Minuws

Minutes

Figure 3.9 HPLC chromatogram of cyanylated peptide 2428, at pH 7.0 & 8.0, room
temperature, 50-foId PYRSCN for 30 minutes

60

The following figures show the MALDI spectrum obtained from the compounds
represented by each of the three peaks in Figure 3. 9

 

 

 

 

 

 

 

 

 

 

 

 

50000~ '3:
.59
it
40000 ~
8 30000-
C
8
D
20000-
{:3
— .3
100004 E: 51
’3 It 1.0.3
- i
0
1000 1500 2000 2500 3000
Mass (m/z)
Figure 3.10(a): MALDI spectrum of peptide 2428 HPLC peak 1
8
20000- S
M
£2
15000 9‘3
fl
5 10000-
O
U
5000 L
0- - ,

 

 

 

2000 2100 2200 2500 2400 2500 2600 2700 2800 2900
Mass (m/z)

Figure 3.10 ( b) MALDI spectrum of peptide 2428 HPLC peak 2.

61

Counts

15000-

100001

5000‘

 

9Z'0EPZ

€1.98?

 

J- -

 

 

 

2500 3000 3500 4000 4500 5000
Mass (In/Z)

Figure 3.10 ( c) MALDI spectrum of peptide 2428 HPLC peak 3.

62

 

 

2. Results of Study on Peptide 1657:

Table 3.4 Product ratios of peptide 1657 after reacting with PYRSCN under pH 7

 

 

 

and pH 8.
pH Percent of Cyanylated Peptide Percent of Mixed Product Percent of Dimer
pH7 60% 25% 15%
pH8 63% 12% 25%

 

 

 

 

 

HPLC chromatograms of cyan ylation products of peptide 1 65 7:

 

 

 

 

 

 

 

Reaction products . Reaction products
at pH 7 1 ‘ at pH 8
1
2
3 3
2
figo‘o‘ '26 0'0' '22' 0'0' grog, ' 26.00 78100 2600' 3.60 34100 160033100
. . Minuteis Minutes

Figure 3.11 HPLC of cyanylated peptide 1657, at pH 7.0 & 8.0, room temperature, SO-fold
PYRSCN for 30 minutes.

63

Counts

Counts

The following figures show the MALDI spectrum obtained from the compounds
represented by each of the three peaks in Figure 3.11.

50000‘

40000-

30000:

20000‘

100001

 

1683 24

 

 

0'1

 

3. MLM ALE

 

 

 

1400 1500 1600 1700 1800 1900
Mass (In/z)

Figure 3.12 (a) MALDI spectrum of the above peptide 1657 HPLC peak 1

 

70000

60000‘

50000‘

40000‘

30000‘

20000~

10000“

 

1768.28

1658.97

 

 

F 3315.48

 

JUL

 

1000 1500 2000 2500 3000 3500 4000
Mass (m/z)

Figure 3.12 ( b): MALDI spectrum of the above peptide 1657 HPLC peak 2.

64

 

Counts

 

70000

600001

50000“

40000‘

300007

200001

10000:

 

1658.32

1683.51

3315.21

 

L

 

1500 2000 2500
Mass (m/z)

Figure 3.12 ( c ): MALDI spectrum of the above peptide 1657 HPLC peak 3.

65

3000

3500

4000

 

 

3. Results of Study on Peptide 919:

Table 3.5 Product ratios of peptide 919 after reacting with PYRSCN under pH 7

and pH 8.

 

pH Percent of Cyanylated Peptide Percent of Mixed Product Percent of Dimer

 

pH7

50% 33% 17%

 

pH8

 

58% 27% 15%

 

 

 

 

HPLC chromatograms of cyanylation products of peptide 919:

 

 

 

 

 

 

 

 

 

Reaction products Reaction products
at pH 7 at pH 8
l
2
l
2 3
3
1000 ' 1200 ' 041.00 ' 16.00 ' 13.00 i - '101-05 T 1100' ' 14:00' ' 16.00 1 1300'
Minutes Mm
Figure 3.13 HPLC chromatograms of cyanylated peptide 919, at pH 7.0 & 8.0, room

temperature, 50-fold PYRSCN for 30 minutes.

66

Counts

Counts

The following figures show the MALDI spectrum obtained from the compounds
represented by each of the three peaks in Figure 3.13.

 

 

 

 

 

 

 

 

 

 

 

50000: g;
V?
W
V
0\
40000
30000
20000-
5 n
S a a
co N_ (x.
10000 °° $3 8
w W
0 .
850 900 950 1000 1050 1100
Mass (mlz)
Figure 3.l4(a) MALDI spectrum of the above peptide 919 HPLC peak 1
N
50000 ‘ 33 ' 34
§ 92‘
40000:
300001
20000:
_ O
c. a 3
10000- g 8 a
0 d
800 1000 1200 1400 1600 1300 “5'

 

 

Mass (m/Z)

Figure 3.14 ( b ): MALDI spectrum of the above peptide 919 HPLC peak 2.

67

Counts

 

 

 

 

 

 

40000 N.
\O
B
30000-
20000-
56,
100001 9.2
0 ~ A
800 1000 1200 1400 .3. 1600 1800 2000

Mass (in/z)

Figure 3.14 ( c ): MALDI spectrum of the above peptide 919 HPLC peak 3.

From the above data, we can see that for the PYRSCN reagent, as the pH goes from 7
to 8, the mixed disulfide products decrease dramatically for the three peptides studied.

The desired cyanylation product results from the cyanide group on the cyanylation
reagent replacing the hydrogen on the sulfhydryl group ( from -SH to —SCN), so there is
a mass shift of +25 Da for the cyanylated product. The mixed disulfide results from the
other side of the cyanylation reagent replacing the hydrogen atom on the sulflrydryl group
shown in scheme 3.1, so for our PYRSCN reagent, the mass shift for the mixed disulfide
is +109 Da.

Generally after applying the cyanylation reagent on peptides and proteins, three
reactions take place, the cyanylation reaction which has a mass shift of +25 Da for any

cyanylation reagent, and a side reaction which is called the mixed disulfide reaction as

68

was discussed above, for PYRSCN reagent it gives a mass shift of + 109 Da, for the 4-
dirnethylamino phenyl thiocyanate reagent it gives a mass shift of + 151 Da. Another
side reaction is the formation of a dimer of the original peptide. The dimer is usually
formed under more basic conditions, the thiolate ion reacts with itself and forms a dimer.
A dimer will not have much effect on the following cleavage reactions because the only
reaction that might take place is the dissociation of the dimer to the original peptide
which has no negative effect on our methodology in locating the sulflrydryl or the

disulfide bond positions. Hence, this reaction is not described further at the present time.

SCN N
| s
C 2 ('3' + I —> CH 2 O +
l
—NH—C —C— i _NH_c —|c— I
N H
Cyanylated Product
@
N S — 8
SH I I O
l s CH2 0
C 2 O + I ——> | H + HCN
C II —NH —CH —c —
-NH — — c — i
N Mixed Disulfide
Scheme 3.1 Formation of cyanylated and mixed product using PYRSCN reagent

69

 

(GSH + ArSCN —-> GSCN + p55

(5) t6

ArSSAr + CN '
/—
GSSG 455R
.1.

 

(3)

 

 

ArS
or \ GSSA + A 3'
CN - l' r

Scheme 3.2 Reaction proposed to occur in GSH and NTCB reaction by Degani and
Patchornik

C. Introduction to the reaction pathway:

Some research has been done on the NTCB ( 2-nitro-5-thiocyanobenzoic acid)
reagent by Degani (21). Since the structures of our two reagents are similar to NTCB in
that they all have a SCN group attached to a benzene ring, we Observed the same mixed
disulfide formation reactions with NTCB and the other two reagents. The proposed
reaction pathway of NTCB reacting with thiol groups could very possibly be used to
explain the cyanylation reactions of PYRSCN and 4-(Dimethylamino)phenyl thiocyanate.

Both Degani (21), and Price (22) Observed the mixed disulfide reaction with the
NTCB reagent. They found that when reacting NTCB with GSH (glutathione), formation
of GSTNB was seen. A proposed sequence of reactions, leading to the formation of both
GSTNB (i.e., GSSAr) and GSSG (oxidized glutathione), is shown in Scheme 3.2.
According to this scheme of consecutive competitive reactions, ArS' formed by primary

reaction 1, displaces CN' from still unreacted ArSCN to form ArSSAr (reaction 2). The

70

aromatic disulfide then reacts with free GSH to form GSSAr (reaction 3) which in turn
reacts in reaction 4 with still unreacted GSH to form GSSG. The latter can also arise
from GSCN and GSH through reaction 5 (23). All the secondary reactions 2-5 exist
independently.

From Price's work in 1976, he also observed the GSSAr formation when reacting
NTCB with a free sulflrydryl group, but he only suggested that reaction of 2-nitrO-5-
thiocyanatobenzoic acid (NTCB) with thiol groups can proceed via two possible
pathways, as shown in the following Scheme 3.3. The reacting thiol group is shown in

its ionized form.

(a) R-S-CN + Ar-S'
R-S' + Ar-S-CN

m R-S-S-Ar+ CN '

Scheme 3.3 Two proposed pathways for the reaction of NTCB with GSH by Price

Pathways (a) and (b) represent two competitive reactions; for some peptides and
proteins reaction (a) is dominant , and for others reaction (b) is dominant. This means
different peptides and proteins will follow different cyanylation reaction pathways.

Degani (21) found that pH had an effect on the reaction rates and hence the

percent of the mixed disulfide that is formed except reaction 2, which is pH independent

71

 

in this range. This is because all the nucleophiles can act as a leaving group and vise
versa and all of the nucleophiles in the system react primarily in their anionic form; their
effective concentrations at any given pH will therefore depend on the pKa's of their
respective conjugate acids. The pKa's of GSH and HCN are 8.56 (24) and 9.32 (25),
respectively, whereas that of ArSH is only about 5.1 (26). At pH 7 and above, ArSH is
therefore essentially fully ionized, and hence in this region, the rate of reaction 2 does not
depend on the pH.

From our data, we found that pH does have an effect on the reaction of PYRSCN
with all the three peptides that we studied. The general trend is as the pH goes higher, we
see less side reaction which means less mixed disulfide was formed. The reason could be
explained by the above theory. As the pH goes higher, more peptides are in their ionized
forms, thus more SH groups are in the S' form. Since it is the thiolate that attacks the
cyanylation reagent as a nucleophile, the more thiloate ion, the faster reaction 1 goes.
Under these conditions, less ArSCN is unreacted, so less ArSSAr is formed, and then less
mixed disulfide GSSAr is formed. While at pH 7, there is more peptide in its protonated
form, so less is reacted with the cyanylation reagent, and thus more ArSCN is unreacted,
so more mixed disulfide products are formed. So from our results we can see the better
pH for cyanylation of SH group is pH 8, but disulfide scrambling at pH 8 is unaviodable.
We used pH 7 for the following experiments.

We can also see from the above data that under same reaction conditions, using the
same reagent, a different peptide gives a different proportion of mixed and cyanylated
products. Price '3 work in 1976 found the same thing. This can only mean that the

reaction also depends strongly on the structure of the peptides.

72

D. Effect of reagent structure on the cyanylation reaction:

The cyanylation reaction was conducted using both reagents PYRSCN and 4-

(dimethylarnino) phenyl thiocyanate at pH 7. The percentage of each peak is compared

in Table 3.6.

Results of Study on Peptide 2428:

Table 3.6 Product rations of peptide 2428 after reacting with PYRSCN and 4-
(dimethylamino) phenyl thiocyanate at pH 7.

 

 

 

 

 

 

 

 

pH 7 % of Cyanylated Peptide % of Mixed Product % of Dimer
PYRSCN 50% 43% 7%
Dimethyl reagent 66% 23% 11%
l
3

 

WM

 

6.00

10.00 12.00 14:00 16.00 10.00 20.00 22100

Minutes

Figure 3.15 HPLC chromatogram of the cyanylation product of peptide 2428 reacted with 50-
fold molar excess of 4-(dimethylamino) phenyl thiocyanate at pH 7, room temperature for 30

minutes.

73

 

Unlike the cyanylation reaction with PYRSCN reagent (Figure 3.14), the mixed
disulfide for 4-(dimethylamino) phenyl thiocyanate is eluted later than the dimer, so now
peak 2 is the dimer peak, while peak 3 is the mixed disulfide peak.

From this experiment we found that a different reagent structure (PYRSCN and 4-
dimethylamino phenyl thiocyanate) also has an effect on the cyanylation reaction and it
can be explained by the reaction chain theory.

The reaction chain has some noteworthy properties (21) as described above that
all nucleophiles can act as leaving groups and visa versa. The order of decreasing
nucleophilicity is GSH > CN' > ArS', because the alkyl group is an electron donating
group, while an aromatic ring can stabilize the electrons. So when a fiee sulfhydryl is
attached to an alkyl group, the thiolate is a better nucleophile than when it is attached to
an aromatic ring. Electrons are better stabilized in the ring, so the GSH is a stronger
nucleophile than ArSH. So the order of increasing leaving groups is just the opposite:
ArS' > CN' > GSH. From this we can also see that CN' is thus both a moderate
nucleophile and a moderate leaving group. We can deduce that the reaction rate for any
reaction in the system depends on the specific combination of a nucleophile and a leaving
group, those reactions involving extreme combinations being the fastest ones. Thus a
sulfur substituent on a p-nitro-phenyl ring is more electrophilic than the sulfur atom of
the aliphatic side chain; k3 is therefore larger than k4 although both reaction 3 and 4
involve the same nucleophile (GSH) and the same leaving group (ArS'). A similar
difference in electrophilicity ( or thiophilicity) between carbon and sulfur may account
for the still lesser reactivity of ArSCN compared to that of either ArSSAr or GSSAr.

Also, slight conjugation effects within the aromatic thiocyanate molecule (27,28) would

74

tend to somewhat stabilize it. On the other hand, the symmetry of ArSSAr may further
contribute to the reactivity of this disulfrde, which is the highest in the system.

We assessed two cyanylation reagents with model peptide 2428. Both
experiments were conducted under the same reaction conditions at pH 7. The 4-
(dirnethylamino) phenyl thiocyanate reagent has much less mixed disulfide formed than
the PYRSCN reagent. Since the reaction conditions are exactly the same, so it must be
the different structure of the reagent that makes a difference. When reacted, 4-
(dimethylamino) phenyl thiocyanate has less electrophilicity than the PYRSCN reagent.
Because dimethyl amino is an electron donating group, the sulfur atom attached to the
dimethyl amino phenyl ring is more nucleophilic, thus less electrophilic. On the other
hand, the sulfur atom of the PYRSCN reagent is attached to a pyridine ring; the nitrogen
atom on the pyridine ring withdraws electron from the sulfur atom, thus it is electron
poor, so it has greater electrophilicity, and less nucleophilicity. Thus, when a thiol group
reacts with these two cyanylation reagents, since the thiolate ion is a stronger
nucleophile than the ArSCN, so it will attack the cyanylation reagent, also because ArS '
is a better leaving group than CN' as was described above, so GSCN will be formed as
indicated in Scheme 3.2 reaction 1. For the mixed disulfide reaction, because PYRS' has
a greater electrophilicity than 4-(dmethylamino) phenyl thiloate ion, so reaction 3 in
Scheme 3.2 with PYRSCN reagent tend to be easier , thus more mixed disulfide is
formed when using PYRSCN reagent. So the 4-(dimethylamino) phenyl thiocyanate
reagent is a better cyanylation reagent than PYRSCN when reacting with peptide 2428.

From the above discussion, an electron withdrawing group on the benzene ring of

the cyanylation reagent seems to cause more mixed disulfide to be formed, but the

75

function of different substituents on cyanylation reaction 1 which is the main reaction
still needs further investigation. Since this is a reaction chain, each reaction affects the

other, so the interaction among all these reactions must be considered.

IV. Summary:

All four reagents can give the cyanylation reactions as we expected, but in
different pH ranges. Experiments were done by conducting the cyanylation reaction
under specific pH conditions. Samples of the reaction mixture were dried and pH was
adjusted to acidic conditions using citrate buffer, then analyzed by MALDI-MS.

For comparison, a 50-fold molar excess of NTCB was used at pH 8-9, 37C for 30
minutes, and a single M+25 peak was observed for peptide 919, peptide 1657 and peptide
2428, while on peptide 2428 we saw some cleavage products. For all three peptides, we
saw a trace of B-elimination products as was observed by others (21).

The CDAP reagent can be used at pH 3-4. Only lO-fold molar excess of the
reagent at pH 3 for 15 minutes is enough to give a complete cyanylated product for all the
peptides tried. Since the reaction is completed at very acidic pH, disulfide bond
scrambling should be surpressed, and no B-elimination and cleavage reaction can take
place.

PYRSCN was first tried at pH 3, because from its stable structure we expected it
to be reactive in a wide pH range from acidic to basic. Later on, we found under very
acidic pH there is little cyanylation reaction; however as pH was increased, the

cyanylation reactions took place. At pH 6, there is already some cyanylated product,

76

while at pH 7 more is cyanylated. Since under acidic pH disulfide scrambling could be
largely prevented, we use pH 6.5-7 as our reaction condition; a 50-fold molar excess of
reagent was applied at room temperature for 30 minutes. The cyanylation reaction was
Observed on all three peptides studied, together with some B-elimination and a little bit of
cleavage.

4-(dimethylamino) phenyl thiocyanate was also first tried at acidic pH, but not
much cyanylation reaction was observed; however, at increasing pH (up to 8), more and
more cyanylated products were observed. At pH 7, there is already a significant amount
of cyanylated product formed. When using 4-(dimethylamino) phenyl thiocyanate
reagent to cyanylate cysteine, we also observed mixed disulfide and dimer peaks as we
did in the PYRSCN and NTCB cyanylation reactions. We found that the pH of the
reaction, peptide structure and substituents on the benzene ring of the cyanylation reagent
all have an effect on the proportion of cyanylated or mixed products that were formed.
Further research needs to be done to investigate the effect of substituents on the
cyanylation reaction.

From the above we can conclude that NTCB, PYRSCN, and 4-(dimethylamino)
phenyl thiocyanate have similar structures, all have SCN groups and benzene rings; these
are designated a series of SCN reagents. These SCN reagents can only be effective for
cyanylation under slightly acidic to basic pH. PYRSCN and 4-(dimethylamino) phenyl
thiocyanate have advantages over NTCB in that they can be used from pH 6-7, while
NTCB can only be used above pH 8, so disulfide scrambling can be prevented to some

degree when using PYRSCN and 4-(dimethylamino) phenyl thiocyanate. Among SCN

77

reagents, mixed disulfide reactions were reported for all of them (21); the effect of this
mixed disulfide on the following cleavage reaction still needs more study. CDAP is the
only reagent so far that can be used under low pH; reaction of CDAP with —SH is fast and
specific, but CDAP is very unstable, so for some difficult peptide or proteins with tight
structures that can only be cyanylated under more basic pH, we might need to try the
other two reagents. More work still needs to be done to test the potency of the two new
reagents.

So, for peptides and proteins that need to be cyanylated under relatively high pH,
such as pH 6-pH 7, then the two new reagents proposed have the advantages over both

CDAP and NTCB.

V. Future work

So far, it seems that CDAP is still one of the best choices for the cyanylation
reaction. Looking at its structure (Figure 3.16), we can see it is a salt, and the positive
charge is on the nitrogen atom directly attached to the CN, and since it only has a CN
group, so there is no mixed disulfide reaction, thus we can avoid this side reaction which
seems to exist in all SCN reagents. Also because the positive charge is close to the CN
group, it makes the CN group have greater electrophilicity toward the thiolate ion than
the CN group on the SCN reagents. From the reaction conditions for both the CDAP and
NTCB, we can see it only takes 10 to 15 minutes for the CDAP to react, while it takes 30
minutes for most of the SCN reagents to react. So a reagent with only a CN group, but
not a SCN group can avoid the mixed disulfide reaction , and also it might react much

faster than the SCN reagents.

78

The new reagents that we proposed are a series of imides such as N-cyanoimide

or N-cyanophthalirnide (Figure 3.16) (29).

l N—CN

Figure 3.16 Structure of N-cyanoimide salt

These reagents are like CDAP; the CN group is attached to a nitrogen atom and at
the ortho position on the ring there are electron withdrawing groups which will make the
CN group more electropositive, so the cyanylation reaction might take place easily.
Unfortunately, these reagents are not Commercially available now, so more research
needs to be done in this area. Also, the potency of the PYRSCN reagent still needs to be

investigated on more proteins and peptides.

79

VI. References:

1. Torchinsky, Y. M., in Sulfur in Protein; (Metzler, D., Translation Ed.), Pergamon,
Oxford, (1981)

2. Boyer, P. D., J. Am. Chem. Soc., 76, 4331 (1954)

3. Hellerman, L., Physiol. Revs, 17, 454 (1937)

4. Kochkina, V. M., and Torchinsky Y. M., Biochem. Biophys. Res. Communs., 63, 393
(1975)

5. Sakai, H., Anal. Biochem., 26, 387 (1968)

6. Yamuchi, T., Yamamoto, S., and Hayaishi, O., J. Biol. Chem, 250, 7127 (1975)

7. Gibson, Q. H.,J. Biol. Chem, 248, 1281 (1973)

8. Ellman, G. L., Arch. Biochem. Biophys., 82, 70 (1959)

9. Vanaman, T. C., and Stark, G. R., J. Biol. Chem, 245, 3565 (1970)

10. Catsimpoolas, N., and Wood, J. L., J. Biol. Chem, 241, 1790 (1966)

ll. Grassetti, D. R., and Murray, J. F., Arch. Biochem. Biophys., 119, 41 (1967)

12. Grassetti, D. R., and Murray, J. F., J. Chromatogr. 41, 121 (1969)

13. Stuchbury, T., Shipton, M., Norris, R., Malthouse, J. P. G., Brocklehurst, K., Herbert,
J. A. L., and Suschitzky, H., Biochem. J., 151, 417 (1975)

14. Jacobson, G. R., Schaffer, M. H., Stark, G. R., and Vanaman, T. C., J. Biol. Chem,
248, 6583 (1973)

15. Catsimpoolas, N., and Wood, J. L., J. Biol. Chem, 241, 1790 (1966)

16. Spande, T. F., Witkop, B., Degani, Y., and Patchornik, A., Advan. Protein Chem, 24,

98(1970)

80

17.

18.

19.

20.

21.

22.

23.

24.

25

26.

27.

28.

29.

Wakselman, M., Guibe-Jampel, E., Raoult, A., and Busse, W. D., J. Chem. Soc.
Chem. Commun., 1, 21 (1976)

Wu, J ., Gage, D. A., and Watson, J. T., Anal. Biochem., 235, 161 (1996)

Wu, J ., and Watson, J. T., Protein Sci., 6, 391 (1997)

Kottke, K., Friedrich, F., and Pohloudek-Fabini, R., Arch. Pharm, 300(7), 583 (1967)
Degani, Y., and Patchornik, A., Biochemistry, 13, 1 (1974)

Price, N. C., Biochem. J., 159, 177 (1976)

Aldrige, W. N., Biochem. J., 48, 271 (1951)

Friedman, M., Cavins, J. F., and Wall, J. S., J. Am. Chem. Soc, 87, 3672 (1965)

. Britton, H. T. S., and Robinson, R. A., Trans. Faraday Soc, 28, 531 (1948)

Ellman, G. L., Arch. Biochem. Biophys., 74, 443 (1958)
Bordwell, F. G., and Boutan, P. J., J. Am. Chem. Soc, 78, 851 (1956)
Campbell, T. W., and Rogers, M. T., J. Am. Chem. Soc., 70, 1029 (1948)

Stephens, R. W., and Domeier, L., Synthetic Commun., 21, 2025 (1991)

81

nrcnrcnn STATE UNIV. LIBRnRrEs
IIIIIIIIIIIIHIWWI“IIIIIIIIWIIIIWIHI
31293017665112