'11:,

T. :11 AL...

.2me :12.

.. .. 21...! .. .

.h
3...!

A ua$mwhm
A 3 GI 1 ~

.
v.2...

a. 3%.
be?“

Or. I.

. .

‘3‘

gas
93“
B?

i
31

E!
:3

ufiuvuqhn‘q‘
.M...‘H.P,n.a V

a .
2-3-9
-‘mkhmhnmfimm
.avxmnfithhvat...
‘§ 1!“!!!

L “a... . WENT? . 3.. ,w

.3

3.223.! e. .

 

\la...»l=l

, r3324! .1
.9545.

xii.
1.2,...

2. :N {
unun

LIBRARY
Michigan State
University

1006

This is to certify that the
dissertation entitled

STUDY OF PRIMARY AMINES FOR OPTIMUM
NUCLEOPHILIC CLEAVAGE OF CYANYLATED CYSTEINYL
PROTEINS DURING THE DEVELOPMENT OF AN ‘ON-LINE’

SYSTEM FOR THE DISULFIDE MASS MAPPING OF
CYSTINYL PROTEINS

presented by
JOSE LUIS GALLEGOS—PEREZ

has been accepted towards fulfillment
of the requirements for the

degree' In CHEMISTRY

will”

I ' Major Professor's Signature
W‘

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

A 0-— o

O~O ‘9-—'r<—.‘—0

coo—C O

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 afifiamfifeindd-pjs

STUDY OF PRIMARY AMINES FOR OPTIMUM NUCLEOPHILIC CLEAVAGE
OF CYANYLATED CYSTEINYL PROTEINS DURING THE DEVELOPMENT OF
AN ‘ON-LINE’ SYSTEM FOR THE DISULFIDE MASS MAPPING OF CYSTINYL

PROTEINS

BY

José Luis Gallegos-Perez

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

2005

ABSTRACT

STUDY OF PRIMARY AMINES FOR OPTIMUM NUCLEOPHILIC
CLEAVAGE OF CYANYLA TED CYSTEINYL PROTEINS DURING THE
DEVELOPMENT OF AN ‘ON-LINE’ SYSTEM FOR THE DISULFIDE MASS
MAPPING OF CYSTINYL PROTEINS

By

Jose Luis Gallegos Perez

Undoubtedly, modern mass spectrometry combined with chemical
derivatization is a very powerful tool in the analysis of biomolecules; currently, it
is the most important analytical tool used in proteomics. In this dissertation, the
use of mass spectrometry in the cyanylation-based disulfide mass mapping gives
us an excellent example of the application of mass spectrometry in biomedical
science and in the determination of the disulfide proteome. The cyanylation-
based disulfide mass mapping methodology relies on the chemical derivatization

and cleavage of the protein prior to analysis by mass spectrometry.

In this dissertation, the results of a quantitative and qualitative study of the
nucleophile-mediated CN-induced cleavage reaction is presented in the context
of the cyanylation-based methodology. The focus is on the use of primary
(methyl to butyl) amines as an alternative to ammonia as the cleavage reagent.
During optimization of chemical cleavage of cyanylated cysteinyl proteins and the
detection of the resulting cleavage products, several parameters such as type of

nucleophile, concentration, temperature, and reaction time were systematically

studied using Ribonuclease A as a model protein. High performance
liquid chromatography was used to monitor the yield of the expected cleavage

products as identified by mass spectrometry.

Also presented is the use of binary mixtures of homologous nucleophiles
to improve the confidence in distinguishing CN-induced cleavage products from
chemical by-products during analyses of cleavage reaction mixtures by MALDI-
MS. The use of two homologous nucleophiles for the cleavage reaction
provomotes the appearance of pairs of mass spectral peaks that facilitate
recognition of diagnostic CN-induced cleavage products, and increases
confidence in results when samples present weak signals or interfering chemical
noise. The analytical advantage that can be gained in the use of a binary mixture
of homologous nucleophiles for cyanylation-induced cleavage of partially reduced
isoforms of cystinyl proteins is evaluated, and the combined use of ammonia and
methylamine is proposed for labeling of cleavage products in a single chemical

step.

Finally, a preliminary development of an ‘on-line’ system for the
cyanylation-based methodology is completely evaluated as a mean of carrying
out all the chemistry ‘on-line’ to avoid sample manipulation and, by consequence,

possible sample losses associated with stepwise batch processing.

DEDICATION

A Dios y la Virgen de Guadalupe a quien me encomendé al inicio de esta etapa
en mi vida.
(To God and Our Lady of Gualaupe whom I entrusted when I started this stage in

my life.)

A mi madre Maria Gallegos Perez. Gracias por tus ensefianzas y amor.
(To my mother Maria Gallegos Perez. Thank you for your teachings and your

love.)

A quien sigo esperando llegue a mi vida.

(To the one for whom I still wait.)

ACKNOWLEDGMENTS

I want to express my gratitude to Michigan State University (MSU) not only
for the excellent academic graduate program in which I was involved but also for

the human values demonstrated.

My gratitude to the community founded by the Saint Marcelino
Champagnat, the Marist Brothers. Thank you also to Querétaro Autonomous

University and the Mexico National Autonomous University.

The fulfillment of this stage in my life would not have been possible
without the advice and support of Dr. Jack Throck Watson, my advisor. Thank
you very much for your understanding when I had some hard times in my life.
Thank you for the time you spent to teach me or just to listen when I needed it.
Special mention to Mrs. Shirley Watson, thanks for helping me with my English

writing.

Thank you to my committee members: Dr. Merlin Bruening, Dr. Marcus

Dantus and Dr. Greg Baker.

I also want to thank to people that always will remain in my heart: Tita
Carmen, Tita Amalia, Sandra Verdnica, Pbro. F. Carrillo, Sra. Carmen Ortega
and her husband, grandparents, godparents, aunts, cousins and the rest of my

family, especially Madrina Elena, Tla Remedios, Mago, Perla, Chava and wife.

I want to thank all my friends in Mexico and other countries. Thank you for
all your support and encouragement when I needed it. Thanks especially to the
families de la Guardia Herrera, Aguado Hernandez and Ibarra Montaflo for your

help before and through all my stay at MSU.

I also thank my friends that I met at MSU: Eliot, Sandra, and Sarhi; Kai-
Chun, Vitor and Sandra, Gabe, Eugenia, Tannia, Kiran, Paolo, Yongjae,

Arindom, Devi and family, Arun and family, Cchimee and Sanjeev.

Thanks to previous and current Watson group members: Theresa Burks,
Stephen Robert Bowman, Jianfeng (Frank) Qi, Wei Wu, Chad Borges, Yingda
Xu, Laura Rangel Ordéfiez, Charles Ngowe, Dr. Robin J. Hood, Yi-Te (Roberto)
Chou, Xue Li (now Dr. Shelly), Jamie Dunn and Olivia Guerre; to mass spec

facility personnel, Susi, Rhonda, Mrs. Bev, Lijun for your friendship and support
To anyone else not mentioned thank you.
Last but not least to Laura, you have shared as a friend all my stay at
MSU since I applied to come to USA. I have cherished all your support, first from

Mexico, then during your stay at MSU, and now even from Germany during good

and bad times.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ x
LIST OF FIGURES ......................................................................................................... xi
ABBREVIATIONS ........................................................................................................ xvii
CHAPTER 1 DISULFIDE MASS MAPPING OF PROTEINS ........................................... 1
1.- PROTEINS ............................................................................................................. 1
1.1 Biological importance of proteins ....................................................................... 1

2.- PROTEOMICS AND THE DISULFIDE PROTEOME .............................................. 2
2.1 Introduction ........................................................................................................ 2

2.2 Proteomics and Biology ..................................................................................... 4

2.3 The Disulfide Proteome and biological importance of disulfide bonds ................ 6

3.- MASS SPECTROMETRY AND PROTEIN CHARACTERIZATION ........................ 9
3.1 Introduction ........................................................................................................ 9

3.2 Ionization techniques ....................................................................................... 14
3.2a Fast Atom Bombardment (FAB) ................................................................. 14

3.2b Matrix-Assisted Laser/Desorption Ionization (MALDI) ................................ 16

3.20 Electrospray Ionization (ESI) ...................................................................... 20

3.2d Atmospheric Pressure Chemical Ionization (APCI) .................................... 27

3.2e Electron Capture Dissociation (ECD) ......................................................... 28

3.3 Mass analyzers ................................................................................................ 33
3.3a Quadrupole Mass Analyzer (Q) .................................................................. 34

3.3b Quadrupole Ion-Trap (QIT) or Quistor (Quadrupole Ion Storage) ............... 40

3.3c Linear (Quadrupole) Ion Trap (LIT) ............................................................ 45

3.3d Time of Flight (TOF) ................................................................................... 49

3.3e Fourier Transform Ion Cyclotron Resonance (FT-ICR) ............................... 55

3.4 Some examples of instrumentation .................................................................. 60
3.4a Tandem Mass Spectrometry ...................................................................... 60

3.4b Types of MS/MS experiments .................................................................... 61

3.40 Tandem-in-time instruments ....................................................................... 62

3.4d Tandem-in-space instruments .................................................................... 64

3.4e Other hybrid instruments ............................................................................ 66

3.4f Applications of MS/MS ................................................................................ 67

4.- PROTEIN SEQUENCING AND DISULFIDE MASS MAPPING ............................ 68
4.1 Introduction ...................................................................................................... 68

4.2 Disulfide mass mapping ................................................................................... 73
4.2a Importance of disulfide bonds .................................................................... 73

4.2b Strategies for locating of disulfide bonds in cystinyl proteins ...................... 74

5.- SUMMARY ........................................................................................................... 81
6.- REFERENCES ..................................................................................................... 84

vii

CHAPTER 2. OPTIMIZATION AND STUDY OF EXPERIMENTAL CONDITIONS FOR
NUCLEOPHILIC CLEAVAGE OF CYANYLATED CYSTINYL PROTEINS BY PRIMARY

AMINES IN DISULFIDE MASS MAPPING METHODOLOGY ....................................... 95
1.- INTRODUCTION .................................................................................................. 95
1.1 Chemical cleavage of proteins ......................................................................... 95
1.2 Chemical cleavage in the cyanylation-based disulfide mapping methodology 100
2.- MATERIAL AND METHODS .............................................................................. 106
2.1 Chemicals ...................................................................................................... 106
2.2 Mass spectrometry ........................................................................................ 107
2.3 Chromatography ............................................................................................ 108
2.4 Partial Reduction of RNase-A ........................................................................ 108
2.5 Cyanylation of single-reduced isoforms of RNase-A ...................................... 108
2.6 Cleavage of single-reduced and cyanylated RNase-A isoforms ..................... 109
2.7 Quenching of the cleavage reaction ............................................................... 109
3.- RESULTS AND DISCUSSION ........................................................................... 112
3.1 Assessment of nucleophilic cleavage reactions at room temperature ............ 112
3.2 Optimization of experimental conditions for the cleavage reaction using
ammonia and methylamine as nucleophiles ......................................................... 116
3.2a Evaluation of concentration, pH and the use of microwaves at room
temperature ..................................................................................................... 1 16
3.2b Effect of low temperature and reaction time on the cleavage reaction ...... 119
3.2c Effect of nucleophile concentration on the cleavage reaction ................... 121
3.2d Repeatability of cleavage reaction yields ................................................. 122
3.3 Evaluation of the optimized experimental conditions ...................................... 127
3.3a Comparison of MALDI-mass spectra of cleavage reaction mixtures prepared
under modified conditions at low temperature .................................................. 127
3.3b Application of the modified methodology to the analysis of a-lactalbumin 131
4.- CONCLUSIONS ................................................................................................. 135
5.- REFERENCES ................................................................................................... 136

CHAPTER 3. USE OF HOMOLOGOUS NUCLEOPHILES FOR CHEMICAL CLEAVAGE

AND LABELING IN CYANYLATION-BASED DISULFIDE MASS MAPPING ............... 140
1.- INTRODUCTION ................................................................................................ 140
1.1 Chemical Tagging Methodology for Quantitative Proteomics ......................... 140
1.2 Application of labeled methodologies in determination of disulfide bonds ...... 145
2.- MATERIAL AND METHODS .............................................................................. 148
2.1 Chemicals ...................................................................................................... 148
2.2 Mass spectrometry ........................................................................................ 149
2.3 Chromatography ............................................................................................ 150
2.4 Partial Reduction of RNase-A ........................................................................ 150
2.5 Cyanylation and isolation of single-reduced isoforms of RNase—A .................. 150
3.- RESULTS AND DISCUSSION ........................................................................... 152

3.1 Application of “external” labeling to locate CN-induced cleavage fragments ..152
3.2 Optimization of the conditions for a combined ammonia-methylamine reagent

for “internal” labeling ............................................................................................ 155
4.- CONCLUSION ................................................................................................... 161
5.- REFERENCES ................................................................................................... 162

viii

CHAPTER 4. DEVELOPMENT OF AN ON-LINE SYSTEM FOR THE CYANYLATION-

BASED DISULFIDE MASS MAPPING ........................................................................ 164
1.- INTRODUCTION ................................................................................................ 164
1.1 On-line methods in analytical chemistry and proteomics ................................ 164
1.2 On-line method for the Partial Reduction-Cyanylation-CN Induced Cleavage
(PRCC) Mass Mapping of Cysteinyl Proteins ....................................................... 168
2.- MATERIAL AND METHODS .............................................................................. 170
2.1 Chemicals and Enzymes ............................................................................... 170
2.2 Columns and reactor for the cleavage reaction .............................................. 171
2.3 Stationary phases .......................................................................................... 172
2.4 Preparation of traps with stationary phases ................................................... 172
2.5 Mass spectrometry ........................................................................................ 173
2.6 Partial Reduction, Cyanylation and Isolation of Single-Reduced lsoforms of
RNase-A .............................................................................................................. 174
2.7 Chromatography ............................................................................................ 174
3.- RESULTS AND DISCUSSION ........................................................................... 175
3.1 One-column approach for the on-Iine PRCC methodology ............................. 175
3.1a Proposed hardware for the one-column on-line system ........................... 175
3.1b Evaluation of the one-column on-Iine approach ....................................... 177
3.2 Two-column approach for on-Iine PRCC methodology ................................... 183
3.2a Proposed hardware for the two-column on-line system ............................ 183
3.2b Trap 2 evaluation: Capture of cleavage products .................................... 185
3.2e Trap 1 evaluation: Resin selection, load capacity efficiency, feasibility of
partial reduction and cyanylation chemical reactions. ...................................... 189
3.2d Trap 1 evaluation: Desorption efficiency ................................................. 193
3.2e Reactor Evaluation: Protein Transfer and CN-Protein Cleavage ............. 195
3.2f Overall Evaluation of the on-line system: The on-line PRCC methodology198
4.- CONCLUSIONS ................................................................................................. 202
5.- REFERENCES ................................................................................................... 203
CHAPTER 5. CONCLUSIONS AND FUTURE WORK ................................................ 207
1.- CONCLUSIONS ................................................................................................. 207
2.- FUTURE WORK ................................................................................................ 208
3.- REFERENCES ................................................................................................... 212

LIST OF TABLES

Table 1. Overview of the selected ionization techniques used in mass spectrometry.
(Adapted from reference [33] with additions from [3740]). ............................... 12

Table 2. Nomenclature for the single-reduced isoforms of RNase A ............................ 110

Table 3. Calculated m/z values for protonated cleavage products of IF4 produced using
ammonia or methylamine as nucleophile. ...................................................... 114

Table 4. Representative data for cleavage product yields for IF4 obtained at level 15x
under different experimental conditions. In parenthesis, the coefficient of
variation (CV) is expressed as a percentage, for experiments under the same
treatment. ...................................................................................................... 1 26

Table 5. Relative intensities from MALDI mass spectra for protonated cleavage products
produced using ammonia, methylamine, or ethylamine as cleavage reagents for
single-reduced/cyanylated a-Iactalbumin isoforms (lF1L, IF2L, |F3L, IF4L).
Calculated m/z values are for protonated cleavage products produced by

nucleophilic attack by ammonia. .................................................................... 134
Table 6. Nucleophile concentrations for the reagent mixtures tested. .......................... 156
Table 7. Calculated m/z values for the protonated cleavage products using methylamine
or ammonia as nucleophiles. ......................................................................... 157
Table 8. Checking list of experimental conditions tested for cleavage of IF4 on-column.
...................................................................................................................... 181
Table 9. Evaluated parameters to test the feasibility of the two-column approach for
developing an on-Iine system for the PRCC methodology. ............................ 184
Table 10. Flow program to desorb isoforms from trap 1 ............................................... 194

Table 11. Summary of the cleavage products assigned for an ‘on-line’ experiment using
ammonia. Fragments indicated with * were produced by double or triply
reduced and cyanylated isoforms of RNase-A. ............................................ 201

Table 12. Comparison of sample processing time required for in-vial and on-Iine
approaches. ................................................................................................. 202

LIST OF FIGURES

Figure 1. Protein-Structure Paradigm and the Disulfide Proteome (from [5]). .................. 2

Figure 2. Different levels of measurement in a system approach. (Adapted from [11]) ....4

Figure 3. Biochemical context of genomics, proteomics and the disulfide proteome

(adapted from [6]). ........................................................................................... 9
Figure 4. Fast atom bombardment process. In the scheme Xenon gas is used to desorb
sample M. ...................................................................................................... 15
Figure 5. Sketch of the MALDI desorption process. ....................................................... 17
Figure 6. Scheme of an electrospray source (adapted from reference [65]) ................... 21
Figure 7. Schematic of Taylor cone formation (from [66]). ............................................. 22
Figure 8. A droplet in the throes of what is known as Rayleigh Instability, and sometimes
referred to as Coulomb explosion (from reference [68]). ................................ 23
Figure 9. Droplet evolution scheme due to evaporation at constant charge and Coulomb
fissions close to the Raleygh limit. Adapted from reference [71] ..................... 25
Figure 10. Conceptual representation of Atmospheric Pressure Chemical Ionization. ...27
Figure 11. Ions produced by cleavage of the backbone chain In a peptide or protein.
Adapted from [82] and [83] ........................................................................... 30
Figure 12. Disulfide bond cleavage using ECD-MS. ...................................................... 32
Figure 13. Quadrupole mass filter. ................................................................................ 35
Figure 14. Quadrupole stability diagram. (Adapted from [38, 88, 92]) ............................ 36
Figure 15. Quadrupole stability diagram when working in rf-only mode (a = 0). (Adapted
from [38, 88, 92]) .......................................................................................... 39
Figure 16. Schematic diagrams of an ion-trap; Z0 is distance from the center of trap to

Figure 17.

Figure 18.
Figure 19.
Figure 20.

apex of the end-cap, ro is the internal radius of the ring electrode. (Adapted
from [93]). .................................................................................................... 41

A) Simulation in two- and three- dimensions of the trajectory of an ion with a
m/z value equal to 500 (Diagram from reference [94]). B) Stability diagram
for the ion-trap, subscript 2 represent axial motion between the endcaps. U is
the DC potential on the endcap electrodes, V is the rf potential applied to the
ring electrode, r0 is the radius of the ring electrode, to is the rf angular
frequency, 2 is the charge and m is the mass. Parameters B are additional
trapping parameter to q and a, to describe the ion stability in the diagram

(From reference [95]). .................................................................................. 42
Linear Ion Trap from Thermo-Finnigan [93]. ................................................. 46
Comparison of focusing in a QIT and a LIT, from reference [93]. .................. 47
Scheme of a Time Of Flight (TOF) mass spectrometer ................................. 51

xi

Figure 21.

Figure 22.
Figure 23.

Figure 24.
Figure 25.
Figure 26.

Figure 27.

Figure 28.
Figure 29.
Figure 30.
Figure 31.

Figure 32.

Figure 33.
Figure 34.
Figure 35.
Figure 36.
Figure 37.

Figure 38.
Figure 39.

Phenomena that affect resolving power in a TOF mass spectrometer. 1)
Temporal distribution 2) Spatial Distribution 3) Kinetic energy distribution 4)
Turn around time effect. ............................................................................... 52
Schematic diagram of ion trajectories in a reflectron-Time-of- Flight mass
analyzer. ...................................................................................................... 53
Scheme of a linear oa-TOF. to, t1 and t2 represent the flight time of ions after
the orthogonal pulse ..................................................................................... 54
Precessing of an ion in a magnetic field ........................................................ 55
Schematic diagram of a FT -ICR mass analyzer (adapted from [110]) ........... 57
Sequence of excitation (left) and detection in FT-ICR-MS. The ionic

micromotion is indicated by the small circles of ions. (Adapted from [66, 110])
.................................................................................................................... 58

Processing of data from an ICR analyzer using a Fourier transform algorithm,
converting time-domain data into frequency domain and then calibrating in
terms of m/z values (adapted from [109]. ..................................................... 59

Representation of a MS/MS (MSZ) experiment in a QIT. Idea and QIT
drawings from reference [95] ........................................................................ 63

Triple quadrupole scheme for the Quattro LC-MS from Micromass UK Limited
[111] ............................................................................................................. 64

Scheme of the commercial Q-TOF Premier from MicromassNVaters, with an
orthogonal TOF analyzer [112] ..................................................................... 65

Scheme of the commercial LTQ-FTMS from ThermoFinnigan (It is a LIT-ICR-
FT mass spectrometer) [114]. ...................................................................... 66

Tryptic map chromatogram of methionyl-human growth hormone and human
growth hormone showing the ability of the tryptic map to differentiate the two
materials. The N-terminal peptide differences are indicated by the arrows.
Slight differences in retention times of peaks eluted early in the
chromatogram are attributed to problems with the HPLC-pumps. (From

reference [1 18]) ............................................................................................ 69
Conceptual diagram of protelytic mass mapping using mass spectrometry.
(adapted from reference [36]) ....................................................................... 7O
Separation, degradation, and identification techniques used to characterize
proteins (adapted from [122]). ...................................................................... 71
Conceptual scheme illustrating the strategy for determination of disulfide
bonds using the enzymatic-based disulfide mass mapping approach. ......... 75
Conceptual scheme of the step-wise partial reduction/alkylation-sequencing
method to determine disulfide bond linkage. ................................................ 77
Conceptual representation of the key steps in CN-based methodology for
disulfide mass mapping ................................................................................ 79
Cleavage sites for a polypeptide chain. ........................................................ 96
Hydrolysis of amides by acid or base (adapted from [4]). ............................. 97

xii

Figure 40.
Figure 41.
Figure 42.
Figure 43.
Figure 44.
Figure 45.
Figure 46.

Figure 47.

Figure 48.

Figure 49.

Figure 50.

Figure 51.

Figure 52.

Figure 53.

Neighboring group effects associated with chemical cleavage in a protein
(adapted from [4]). ....................................................................................... 98

Mechanism of cleavage of methionine peptide bonds with cyanogen bromide
(adapted from [1] and [4]) ............................................................................. 99

Chemical cleavage of cyanocysteine by means of a nuclephilic attack in
alkaline media [8, 9]. .................................................................................. 100

Conceptual representation of the procedure of cyanylation-based disulfide
mass mapping ............................................................................................ 101

Cleavage of a cyanylated cysteine-containing peptide yields two fragments.
.................................................................................................................. 102

Chemical cleavage in the cyanylation-based disulfide mapping methodology.
.................................................................................................................. 104

HPLC chromatogram (see text for details) of RNase-A (IP = intact protein)
and its single-reduced and cyanylated isoforms (lF1...IF4). ....................... 111

HPLC chromatograms of cleavage products from single-reduced and
cyanylated isoforrn 4 (IF4) of RNase-A treated at room temperature with: Top,
1 M ammonia for 60 min; middle, 1 M ammonia for 10 min; bottom, 1 M
methylamine for 10 min. Cleavage products A, B, and C are defined in
Figure 43 and in Figure 45. It is also shown the retention time for the intact
IF4. ............................................................................................................ 1 13

MALDI mass spectra of the cleavage reaction mixture from cyanylated IF4 of
RNase-A, using ammonia (60 min) or methylamine (10 min) as the
nucleophile. A, B, C are the cleavage products according to Figure 43 and

Table 3 ....................................................................................................... 115
Summary of representative experiments at room temperature using |F2.
Squares point out the conditions for the control experiments. .................... 117
Comparison of HPLC chromatograms of cleavage reaction mixtures of IF4

using 1 M methylamine at room temperature or at 2 °C for 10 minutes. The
peaks labeled A, B, and C represent the cleavage products as defined in
Figure 43 .................................................................................................... 120

Relative yield of cleavage products using methylamine for IF4 from RNase-A
as function of the concentration at 2°C. Labels B, C, and A correspond to the
cleavage fragments according Table 3 and Figure 43. ............................... 122

A) Mean diamonds graphs. The middle line in the diamond represents the
mean for a group of data. The vertical endpoints form the 95% confidence
interval for the mean. Overlap marks are useful for comparison of means, but
it is better to use of comparison circle plots. B) Two hypothetical cases
where means of two sample data groups are not significantly and are
significantly different. Comparison circle plots for each group are evident
each case. ................................................................................................. 124

Distribution analysis and mean comparison for yield of cleavage product 8.
M10 = Experiments with methylamine, 10 minutes at 2 °C. N10 =
Experiments with ammonia, 10 minutes at 2 °C. N60 = Experiments with
ammonia, 60 minutes at room temperature. ............................................... 126

xiii

Figure 54.

Figure 55.

Figure 56.

Figure 57.

Figure 58.

Figure 59.

Figure 60.

Figure 61.

Figure 62.
Figure 63.
Figure 64.
Figure 65.
Figure 66.
Figure 67.

Figure 68.
Figure 69.

Figure 70.

MALDI mass spectra of the reaction mixture resulting from the cleavage of
cyanylated IF4 from RNase-A by the indicated nucleophiles. Peaks labeled
A, B, and C represent segments of the sequence as defined in Figure 43. 129

MALDI-MS spectra of the cleavage product mixture resulting from the
processing of IF4 from RNase-A with primary amines at low temperature (2
°C) and a concentration of 2 M. From back to the front: methylamine,

ethylamine, propylamine, or butylamine. .................................................... 130
Overview of MCAT peptide sequencing and quantitation, from reference [2].
.................................................................................................................. 142
Structure of an ICAT reagent. The reagent exists in two forms, heavy
(containing 8 deuteriums, X = D) and light (contains no deuteriums, X = H).
.................................................................................................................. 143
ICAT strategy for quantifying differential protein expression, based on thiol-
specific ICAT reagents (adapted from reference [5]). ................................. 144

Impact of different nucleophiles on the cleavage reaction. Species produced
using ammonia are designated wit “0" and cleavage fragments based on
methylamine are designated with subscrip “14". ......................................... 147

Expanded section of an overlay of three mass spectra processed in silico for
one isoform of a-lactalbumin. Ao represents the cleavage fragment produced
using ammonia as nucleophile, A14 when methylamine was the nucleophiles

and A28 in the case of ethylamine. .............................................................. 154
Ideal relative intensity ratio (rir) close to unity is represented for a MALDI
signal for a pair of labeled cleavage fragments produced by treatment of the
sample with a mixture of ammonia and methylamine. ................................ 155
Relative intensity ratios for labeled cleavage products A and B for RNase A
isoforms. .................................................................................................... 157
Top: Mass spectrum of labeled cleavage products for lF1. Bottom: Mass
spectrum of labeled cleavage products for lF2. .......................................... 159
Top: Mass spectrum of labeled cleavage products for IF3. Bottom: Mass
spectrum of labeled cleavage products for IF4. .......................................... 160
Simple on-line system for cleaning up and concentrating of samples. The
system has two positions; read the text for details. .................................... 165
Schematic representation of an on-line system with sample preparation and
two-dimensional HPLC. From reference [14] .............................................. 167
Comparison of the in-vial and on-line steps for the Partial Reduction
Cyanylation and Cleavage methodology. ................................................... 170
Lab-made setup to fill traps with stationary phase. ..................................... 172
Single-trap approach for an on-line method for PRCC methodology. The

instrumental setup in A) is used for the PRCC methodology (steps 1-8), after
this, T1 is connected to a HPLC system (as in B) for elution and collection of
cleavage products (step 9). ........................................................................ 176

Chemical structure of PRP-3: poly(styrene-divinyl)benzene resin ............... 178

xiv

Figure 71.

Figure 72.

Figure 73.

Figure 74.

Figure 75.
Figure 76.

Figure 77.

Figure 78.

Figure 79.

Figure 80.

Figure 81.

Figure 82.

Figure 83.

MALDI-MS spectrum for lF3 of RNase-A after being loaded into a small
column packed (Trap1) with PRP-3 resin. .................................................. 179

MALDI-MS spectrum for IF4 of RNase-A obtained using in-vial or on-column
approaches. During the in-vial approach, two of the cleavage fragments
(pointed out with arrows) are not observed. Peaks corresponding to i)
cleavage fragments of minor cyanylated isoforms that coeluted with IF4 or B-
elimination products are designated with (*) and ii) those some unknown side-
reaction products are designated with t ...................................................... 180

Kyte & Doolittle hydrophobicity plot for designated single reduced isoforms of
RNase-A. The location of the cysteine contained in each isoform is indicated
after the numeral identifying the isoform ..................................................... 182
The two-trap and reactor approach for an on-Iine method for PRCC
methodology. ............................................................................................. 183
Diagram of a perfusion particle. .................................................................. 185

Scheme of the instrumental setup used to evaluate the capture of cleavage
products by Trap 2. .................................................................................... 187

Chromatogram of cleavage product 1-25 from I F2 for RNase-A. The fact that
the peaks for the 'standard' and the 'sample' cleavage product are almost
overlapped indicates that recovery of the sample is close to 100% (in this
example, 95.2%). ....................................................................................... 189

Comparison of chromatograms of a standard sample of partially reduced
RNase-A as trapped by different stationary phase types. Intact RNase-A and
its isoforms (IF1.. IF4) are labeled. ............................................................. 190

Comparison of HPLC chromatograms of partial reduction reaction mixture of
1 nmol of RNase-A on microbore column filled with Poros R2 .................... 192

Chromatograms for cyanylation on-line with the two-column system. About 1
nmol of RNase-A was treated with CDAP at different concentrations during
15 minutes. From comparing the chromatograms, it is possible to observe
that using a concentration of 0.2 M of CDAP drives the cyanylation to
completion. ................................................................................................ 193

Conceptualized HPLC chromatogram (see A) showing that after desorption
experiments, when T1 is connected to a HPLC system (see B) there is an
absence of residual proteins in T1, which indicates a successful desorption
step. ........................................................................................................... 195

Top: Setup for the PRCC methodology using a two-column on-line system.
Bottom: Schematic diagram indicating that trap 2 can be connected to a
separation system for analysis of the cleavage products formed in the on-line
system. ...................................................................................................... 196

Comparison of chromatograms showing cleavage products for lF3 with in-vial
and on-line approaches under similar experimental conditions (experimental
conditions were similar to the described in section 3.2d). Table inset shows
the m/z values for cleavage products for IF3. ............................................. 198

XV

Figure 84. Cleavage products produced using an on-Iine system for PRCC with NH3 as
cleavage reagent. Top: Peaks representing cleavage products are identified
with an arrow. Bottom: Cleavage products identified for IF4 with different

nucleophiles, ammonia (Ao, Bo) or methylamine (A14, B14). Experiments with
different nucleophiles were carried out independently. ............................... 200

AC
APCI
BDK
CAD
CDAP
CID
CN
CNBr
Cys
Da

DC
ECD
ESI
FAB
FSA
FT
HPLC
ICAT
ICR
|F1 ...IF4
IF 1 L. . .IF4L
itz

LIT
m/z
MALDI
MCAT
MeNHz

ABBREVIATIONS

Alternating Current

Atmospheric Pressure Chemical Ionization
Bradykinin

Collisionally Activated Dissociation
1-cyano-4-dimethylamino-pyridinium
Collision-Induced Dissociation

Cyano group

Cyanogen Bromide

Cysteine residue

Dalton

Direct Current

Electron Capture Dissociation

Electrospray Ionization

Fast Atom Bombardment

Fragment Sets Algorithm

Fourier Transform

High Performance Liquid Chromatography
Isotope-Coded Affinity Tag

Ion Cyclotron Resonance

Cyanylated single-reduced isoforms of Ribonuclease A (numbered)
Cyanylated single-reduced isoforms of a-Lactalbumin (numbered)
lminothiazolidine-4-carboxyl terminus
Linear Ion Trap

Mass-to-Charge ratio

Matrix Assisted Laser Desorption/Ionization
Mass-Coded Abundace Tagging
Methylamine

xvii

MH+
MRM
MS
MS/MS

MSn

NSA
oa-TOF
PD
PRCC

Q

q

QIT
QUISTOR
RNase-A
SIN

SIM
TCEP
TFA
TOF
a-CHCA

I3
u-HPLC

Protonated molecule

Multiple Reaction Monitoring

Mass Spectrometry

Mass Spectrometry /Mass spectrometry (tandem mass
spectrometry)

When used to describe the mass spectrum that results from the
controlled dissociation of a precursor ion of a series of 13‘, 2"“, etc.
generation product ions (various mass spectrometry/mas
spectrometry techniques) (from reference 39 in Chapter 1)
Negative Signature Algorithm

Orthogonal Acceleration Time Of Flight

Plasma Desorption

Partial Reduction, Cyanylation and Cleavage (Methodology for
Disulfide Mass Mapping)

Quadrupole Mass Analyzer

Quadrupole used as collision cell no for mass separation
Quadrupole Ion Trap

Quadrupole Ion Storage

Ribonuclease-A

Signal-to-Noise ratio

Selective Ion Monitoring

Tris(2-carboxyethyl) phosphine

Trifluoroacetic acid

Time Of Flight

a-cyano-4-hydroxy cinnamic acid

B-elimination of HSCN from cyanylated cysteine residue

Micro-High Performance Liquid Chromatography

xviii

CHAPTER 1 DISULFIDE MASS MAPPING OF PROTEINS

1.- PROTEINS

1.1 Biological importance of proteins

Proteins are essential constituents of all organisms. Some proteins can
store and transport a variety of particles ranging from macromolecules to
electrons. In some cases, membrane proteins guide the flow of electrons in the
vital process of photosynthesis and transmit information between specific cells
and organs in complex organisms. Virtually every property that characterizes a
living organism is affected by proteins [1, 2]. Some proteins provide the structure
that helps give cells their bioactive shape; others serve as hormones to carry
signals from one cell to another. Another protein, hemoglobin, carries oxygen
from the lungs to the rest of the body. And perhaps most important of all,
proteins serve as enzymes that catalyze thousands of chemical reactions
necessary to sustain life. The diversity of protein structures and functions is as

wide as the diversity of the life itself [3].

The structure-function paradigm states that a specific function of a protein
is determined by its unique and ordered three-dimensional structure; this is one
of the cornerstones in protein biology and biochemistry (see Figure 1). These

days, scientists believe in the existence of “disordered” proteins, which in their

purified state at neutral pH lack an ordered structure [4, 5]. In summary, the

knowledge of the protein structure is an important issue in the protein study field

[5]-

 

Proteln as
Allosterlc BIO-Machines
Interactions

    

Enzyme

d
Ca talysls Disulfl e

Bonds
Determination

Dlsulflde

Proteomlcs {Proteome

      
   
 

30 Structure Protein Assisted Proteln
. FoIdIn
”a'y’" Structure-Functlo °
Parad'gm Bio-Technology

 
 
  
  

De Novo Proteins

Proteln Self-Organizatlon Proteln Englneorlng

Blomedlclne

Proteln Misfoldlng
and Dlseases

 

 

 

Figure 1. Protein-Structure Paradigm and the Disulfide Proteome (from [5]).

2.- PROTEOMICS AND THE DISULFIDE PROTEOME

2.1 Introduction

All the information necessary to make a complete human being is inside
each of our cells. However, not all genes are expressed in all cells. All cells
express some genes whose protein products provide unique cell-specific
functions. A “proteome” is the entire PROTEin complement expressed by a
genOME, or by a cell or tissue type; it represents the collection of proteins under

specific conditions. As an extrapolation of the concept of the “genome project”, a

“proteome project” is research that seeks to identify and characterize the proteins
present in a cell or tissue and define their patterns of expression [6, 7]. In other
words, proteomics is the study of multiprotein systems, in which the focus is on
the interplay of multiple, distinct proteins in their roles as part of a larger system
such as a cell line, tissue, or complete organism [6, 8]. It is said that protein
chemistry and proteomics involve protein identification, but the substantial
difference between them is that the goal of proteomics is to characterize the
behavior of a system rather than the behavior of a single component such as in

protein chemistry.

Science is moving from genomics to proteomics in order to get insight into
the functional network of gene expression. Actually, proteomics is older than
genomics. The first studies that can be called proteomics began in 1975 with the
introduction of a two-dimensional gel by several research groups who began
mapping proteins from Escherichia coli, mice, and guinea pigs. At that time,
around 300 proteins could be separated, but due to technical limitations, they
could not be completely identified. Currently, with more refined techniques, up to
10,000 protein spots have been separated under highly reproducible conditions.

[9, 10].

Recently, the concept of systems biology has been introduced in the study of
living systems from a holistic perspective based on the profiling of a multitude of

biochemical components (see Figure 2). From this point of view, proteomics and

genomics are a part of a whole in which the final goal is the study of a living

 

 

   
  

system.
DNA ——> Genomlcs
mRNA —> Transcriptomlcs
I Cell level
Protelns —> Proteomics Tissue level SYStems
* Organ level BIOIOQY
Organism level
Proteln-complex ——-> Functlonal Proteomics
Metabolltes —> Metabolomlcs

 

 

 

 

 

Figure 2. Different levels of measurement in a system approach. (Adapted from [11])

2.2 Proteomics and Biology

The accumulation of complete sequences of genomes is not sufficient to
elucidate biological function. From the study of genes solely, not all types of
information can be obtained. For example, proteins are responsible for the
phenotypes of cells, not genes. It is impossible to elucidate mechanisms about
diseases, aging and environmental effects solely by studying the genome. Thus,
the importance of proteomics is a direct result of advances made in large-scale
nucleotide sequencing of expressed sequence tags and genomic DNA. It is only
through the study of the proteins that questions about biological functions can be

answered [9, 12].

Four principal applications constitute protemics:

1) Mining: This is the simple identification of as many of the proteins in

a sample as possible.

2) Protein-expression profiling: This is the identification of a protein in a
sample in a particular state of the organism or cell, e. g., in a disease state or as
a function of exposure to a drug or other stimulus. Normal and diseased cells
can be compared to determine which proteins are expressed differently in

different states.

3)Protein-network mapping: This is the proteomics approach to

determining how proteins interact with each other in living systems.

4) Mapping of protein modifications: In this part of proteomics, the
task consists of identifying how and where proteins are modified.
Postranslational modifications usually govern the structure and function of

proteins, which is why it is important to know how a protein has been modified.

In recent years, attention has shifted to “functional proteomics” where the
proteins are purified with a specific function in mind. An example of this Is the

purification of an organelle or of a multi-protein complex with a particular function

[8]-

2.3 The Disulfide Proteome and biological importance of disulfide

bonds

The disulfide proteome is the subset of proteins containing disulfide bonds
in a proteome. The thiol redox status adds a new dimension to biology as well as

medicine and any field related with protein science [13].

In 1961, the work of Anfinsen et al., showed that disulfide bonds are
formed spontaneously, in RNase-A, upon oxidation in air. This work showed that
a polypeptidic chain has to be folded before it becomes a biologically active
protein. This experiment was the first to show that a protein could fold into its
correct tertiary structure with only the information provided in its aminoacid
sequence [14]. Since then, the two major facts, which are central to
understanding the mechanism of protein folding are being studied. They are: 1)
the development of the theory of folding, especially for predicting structure from
sequence, and 2) analysis of the relation between sequence homology and

evolutionary relatedness [15, 16].

Disulfide bonds are one of the key aspects related to folding among
cystinyl proteins and, consequently, they are important for the understanding of
the chemical structure and function of such a protein. Disulfide bonds are
important for the stabilization of the native structures of cystinyl proteins,

emerging as the primary mechanism by which the activity of a protein is

established [13, 17]. Without disulfide bonds, cystinyl proteins may be
disordered, weak and non-functional [18]. Two classes of disulfide-bonded
proteins are known: the first, in which the inter-cysteine linkage is a stable part of
the final folded structure, and the second, in which cysteine pairs alternate

between the reduced and the oxidized states.

In organisms from Escherichia coli to humans, disulfide bond formation is
important to proper folding of secreted protein, such as bacterial virulence factors
and mammalian glycoproteins [19]. In toxicology, the knowledge of cysteine
linkage is of great importance especially for biologically active peptides derived
from venomous animals and toxin-producing bacteria, where peptide toxins can
possess multiple disulfide bonds. One good example of toxins with disulfide
bonds is the neurotoxins derived from marine snails of the genus Conus, which
belong to a family of bioactive peptides called conotoxins or conopeptides.
These short, disulfide-rich peptide neurotoxins are produced in the venom of
predatory marine cone snails. It is generally accepted that an estimated 100,000
unique conotoxins fall into only a handful of structural groups, based on their
disulfide bridging frameworks. Conopeptides have defined tertiary structures
formed and stabilized by their disulfide bonds [20, 21]. Toxins from spiders,
snakes, or scorpions also show polypeptidic structures containing disulfide bonds

[22-24].

In some diseases, such as the prion transmissible spongiform
encephalopathies, the key event seems to be the conversion of the prion protein
(PsP) from its normal cellular form (PrP°) to an abnormal scrapie isoform (PrPs°)
that may involve a covalent reaction of the sole intramolecular disulfide bond of
PrP° [25]. Involvement of disulfide bonds in stabilization of structure and function
of the Newcastle disease virus hemagglutinin-neuraminidase has also been

published [26, 27].

The disulfide proteome is also gaining importance because of the
evidence suggesting that redox regulation plays a critical role in cell biology. The
redox cycling of the disulfide bond, its formation and dissociation, may be central
to an enzyme’s activity or to protein activation allowing for a “biological switch”
that regulates the protein activity [28—30]. For example, in plants, reduction of
regulatory disulfide bonds using reducing power generated by light, activates the
regulatory photosynthetic enzymes and deactivates glucose-6-phosphate
dehydrogenase (G6PDH) —an enzyme involved in the oxidative pentose

phosphate pathway [31].

Therefore, we can conclude that in the genomic era, when the proteomes
of hundreds of different organisms are known, it is particularly interesting to be
able to highlight the “disulfide proteome” (see Figure 3) for the comprehensive

analysis of the cell function and regulation [19].

 

Genome
DINA I “Genomlcs”

mRNA

I

Proteom . Disulfide
PiOtOIns —"' “PrOtOOMICS” Proteome

Cell Functions

 

 

 

Figure 3. Biochemical context of genomics, proteomics and the
disulfide proteome (adapted from [6]).

3.- MASS SPECTROMETRY AND PROTEIN CHARACTERIZATION

3.1 Introduction

During the last two decades of the last century, the need for structural and
quantitative information from biological samples led to the development of
ionization methods capable of handling fragile, nonvolatile, high-mass molecules,
avoiding, if possible, any derivatization step [32]. Currently, the determination of
the mass of biological compounds is one of the most important pieces of

experimental data in the analysis of their structure.

There is not doubt that mass spectrometry is the most accurate method to
determine the molecular mass of biological molecules. For a long time, chemists
and biochemists were searching for a simple method to determine accurately and
sensitively the molecular masses of blopolymers such as proteins or

polysaccharides [33]. The application of mass spectrometry to the study of

biology was realized in the early 1980s with the advent of reliable ionization
methods such as Fast Atom Bombardment (FAB) and Plasma Desorption (PD).
This made possible the ionization of small peptides and proteins. However, the
two developments in mass spectrometry (MS) instrumentation that revolutionized
protein chemistry were the introduction of electrospray ionization (ESI) and
matrix-assisted laser desorption/ionization (MALDI) [34, 35]. These methods
solved the problem of generating ions from biomolecules. Recently, Electron
Capture Dissociation (ECD) has emerged as a new technique that can be

applied to peptide and protein sequencing.

All mass spectrometers have four essential components:

 

 

 

 

SAMPLE ION MASS/CHARGE DETECTOR
INTRODUCTION 1 SOURCE __~ ANALYZER _.

 

 

 

 

 

 

 

 

 

 

 

 

The sample introduction, depending on the ionization technique, could be
gas chromatography, liquid chromatography, electrophoresis (e.g capillary
electrophoresis, isotacophoresis, and other related techniques) or a direct
insertion probe. Molecules of the sample are converted to ions in the ion source,
and the produced ions are separated according to their mass/charge ratio (m/z).
The ion beam coming from the mass analyzer is detected and transformed into a

usable signal by a detector.

10

Currently, many different ionization techniques and mass analyzers are
used in mass spectrometry, but in general the instrumentation layout is the same

[36].

11

 

.mcmmcE Lena... Lo. b.2323

.m<u. .0. am 2.5% «one
..o comma. mac. “.0 Econ
355.29... a 5.? 29:8
2.8 a “.o EcEEmnEon

$.25. Echbocam

 

 

 

 

 

 

 

 

 

 

 

on. coo? ... .m<n. o. .w__E_m Loco... .m<n. Lo. mm 95% 59.95 2:80 coszEo. mmmE cowbmucooom
.wcmcmco ,

N cm... EOE 5.3 mcSmcm Econ Eofi 355.59:

.5. coco .02 6529.03 .w.:ch.:mmoE 5.5.82-5... 5.3 EoEEmnan .5

.coEcco :9... .xEmE 0.... L2 .28: on coo 85 Ectao 258. can x59: 2:9.
mm. coco ... c. 033.9... on 63E 2292 cc. macaw .c.aE.m .cfimm. m. c. 82036 m. 0.955 EcEEmnEom E03. .mmn.
.coEEmnEon .20.th

.mco. Sacco

823:. o. c.:oc.oE

0.296 .950: 5.2, fine.

25. 93.3.9 can Econ

c2805 cm .5 conic.

m. 85:35. .mEoEEm

com: mom Somme. Lo cab :93 mam Emcee.

e... :0 9.2.8 9.0ch mews. .8238 < .cozommc c.:oc_oE\co.

.c...m_o> can can...” >..mEe£ coszoEmmc 22.. 639998 cm :0 =39 mm 938
no o2. ... on $2: “.5an00 5:0 m. o.=oc_oE 3.9585 228m 9.. .o cozmuco. :0. 838.5. .moEcco

.Eozmg
coszmEomt c_n.o:noacm

.2990ch .o: cfamcmofiEoEO woo 62.3.5. 95.055

mcE..cE0m a. co. 5.28.0.2 mm Loam. 330.500. eczmcmaom omega mam 5 3.28.9:

638m new 9.6.? 2.2.55 9 8.38 23mm .35an8 5.3 892:. 90
an. com ... on “maE ccaanoO 2.5.? __m Ewe: 2 m_nno._aa< ct «69.er 355.59... :9 cozmuco. .5..er
cozmnEo. ommcahmw
.22... mg m¥0<m>><¢o mm0<hz<>o< ..(mmsz DOIhmE

 

 

.covfim. ES. 303.com 5.? 8m. 8:939 Eot ceases .EchLGcam mmmE c. com: «03.5.02 8.65%. 3.8.3 on. “.0 30.290 é min...

12

 

20202008 00.200020
22.00.0200. 9.2.002

 

 

 

 

 

 

 

320002.02. 020202.002 .020..0_020..-.000 02. 022.00 2.0.0.0
00>0.0E. 0002020000 0002. .0 20222000 02. 00.0....00. >2 00.0.000 0.0 A>0
5.3 02.000. 00202002 22.0.0.8 20. 20.22, 2.020 02.000200 0 ... No v. 022.020 3.020
2002 002 00 220.020.... .0200“. 0. 00.0000 0000.0 0.50. 02. 020 0.0.02.0 .30. 202? 000.2200. 30w.
ooo we 0. 0: 2002 .220 002 .. 02.22.00 20020000 02. 0020020 Dom. 22.0220. ....00. 2.02 < 22.200005 0.0.000 22.00.m.
30.00202 .0 >0.0m20.0.0._2
$0.0m020z 00 2.0.0.0228
2.505. 0. ._ 0.0002
0.002820 02. .0 0.. 02.
20 .0..c0.00 02. :0 .20020000
0. 30.. 0.02.00 02. 202.5
.2028 00.00.02 0 000 .02
0.0002 0 E2. .0292... 0000.02. 000.2200.>0.002.00_0
002.0. 30. 0. 0.0... 2200.0 2.02.2.0. 0 0. 0502.00.02.02.
0.0 00.0000 000.020 2020.2 020200.002 20.2 0 0.2. 02.0.00
00 ooo ooN ... .0200 0. 0>...0..00 20> 2.0. .0. 0. 0.0000 >00w 0. 22.0.00 0.0200 02... >0.000..00_w
22.0220. 0.00020 0..02d00E.<
.x...0E 02. E2.
02......000202 020. 02. .0 20220000
30. 00022200. 22.0220. 02. 020 22.0220. 02. .0.
000.00 2...... 0.2.0028 3.020 002>20 .000. 02 .r
0. .02. .0N>_020 000E 02.02. 2.00 0 ... 00x.E 0.
0 02.002 .0. 2...... 02.0000 002.002 .02. 0.2020 20 E2. 020. 20..0~.20_\20..0.000n.
0o ooo oom .. >000 02 2.002.... 0. 92.9... .0 002.2000 0.0.0002 00.0.0.0 .r 00000.0 0. 0000 0. .000. < .000.. 00.0.00<-x...02
22.0.0000 .000..
...—s...— mm<2 wxo<m>><mo wm0<hz<>o< ..(mmsz DOE—Ms.

 

 

 

 

 

 

13

3.2 Ionization techniques

3.2a Fast Atom Bombardment (FAB)

During the 19703, several techniques were introduced to obtain mass
spectra of compounds with negligible vapor pressure. However, the technique
with the greatest impact on the analysis of nonvolatile and/or thermally labile
compounds was FAB. FAB was developed by Barber and co-workers, who
published the novel procedure of introducing the analyte in a viscous liquid
(matrix). Since that time, FAB has been used to analyze a wide variety of
molecules ranging from oligopeptides to organometallics and

buckminsterfullerenes. [38, 41]

In FAB, a small amount of sample (matrix + analyte) is placed on a target,
which is subsequently bombarded with a fast atom beam (for example, 5-8 keV
of neutral xenon or argon atoms). The impact desorbs protonated molecules
(ions) and fragments from the analyte (see Figure 4). Cluster ions from the liquid
matrix are also desorbed and produce a chemical background that varies with
the matrix used. There are several liquid matrixes, the more common of which
are glycerol, thioglycerol, m-nitrobenzyl alcohol, and diethanolamine [37]. There
is a variation of the technique termed “liquid secondary ion mass spectrometry"
(LSIMS). In this variation, cesium ions (05") are used instead of a neutral atom

beam; it seems that the charge of the primary bombarding beam is of little

14

importance in the production of sample ions. What does seem to be of major
importance is the momentum of the bombarding particle. The momentum can be
varied by either increasing the acceleration energy of the particle or by increasing

the mass of the bombarding particle (Xe > Ar) [41].

The mechanism of FAB is not well understood, but the FAB mass spectra
are likely to be dominated by cluster ions from the matrix. Cationized adducts of
the sample are observed: [M+H]*, [M+Na]*, [M+K]+ as well as negatively
charged ions, [M-H]'. For monovalent ionic salts, in a positive FAB spectrum the
cation C+ is observed as well as cluster ions [Cn + X”? and in the case of a
negative FAB mass spectrum, the presence of ion X' and clusters [CM + X..]' are

evident.

 

 

 

 

   

Primary gee lone Momentum transfer and

6" '°"'”"°" acceleration creation of a neutral ion beam

Sam ple (M) deeorptlon

 

 

 

 

 

Figure 4. Fast atom bombardment process. In the scheme Xenon gas is used to desorb sample
M.

15

3.2b Matrix-Assisted Laser/Desorption Ionization (MALDI)

MALDI was developed by Hillenkamp and Karas during the late 19803 and
constituted a milestone in the analysis of biomolecules [35, 42-44]. Currently, it
is nearly impossible to open a recent issue of any journal in the field of protein
science or biochemistry without finding at least one paper on MALDI (original
paper on application in biomolecules has been cited at least 1369 times
according to SciFinder Scholar© accessed July 31, 2005 [35]). A successful
MALDI analysis encompasses several crucial steps: sample preparation,
excitation of the sample and disintegration of the condensed phase, generation
and separation of charges and ionization of analyte molecules, and, finally,
extraction and separation according to the m/z value of the ions in the mass
spectrometer and detection. The technique makes use of the absorption of laser
light by a solid sample layer. In MALDI, the analyte is mixed with a 100-50,000-
fold molar excess of small, UV-absorbing molecules (the matrix) to form a solid-

phase deposit on a sample plate.
MALDI matrices must meet a number of requirements:
- be capable of forming a homogeneous and fine crystalline solid during
the sample preparation.

- be able to embed and isolate analytes (0.9., by co-crystallization)

- be soluble in solvents compatible with analyte

16

- be vacuum stable
- have a high absorptivity for the laser radiation
- cause co-desorption of the analyte upon laser irradiation

- promote analyte ionization [45]

The matrix absorbs and transfers the energy from the laser pulse to the
sample producing a plasma that allows in vaporization and ionization of the
analyte. The ionization process produces both positive and negative ions. The
analyte molecules usually combine with just one proton to form single-charge
ions [M+H]*; however, some double— and triple-charge ions may be observed
(e.g., [M+2H]2*, [M+3H]3*). The presence of sodium and potassium adducts is

also possible (Figure 5).

 

mam” main”
[M+2H]2* [M+H] [M+H]+
[WHY [WHY
[M’rfll‘ [WI-'1’
[M+2H]2*

 

0 Matrix
0 Sample

 

 

 

Figure 5. Sketch of the MALDI
desorption process.

17

MALDI is a complex chemical event, where one laser shot/one mass
spectrum is obtained in few nanoseconds. After the laser shot, a plume of
material containing neutral matrix molecules, as well as reactive species (matrix
radicals, electrons, and hydrogen atoms), is formed. Within this plume,
regardless of a low average degree of ionization, suitable analytes can be

ionized efficiently.

Models for MALDI were developed considering three different steps:

1) Incorporation and isolation of analyte molecules in the host matrix

2) Release of molecules into the vacuum upon disintegration of the matrix-
analyte solid after laser-energy deposition (energy absorption)

3) Ionization by ion-molecule reaction in the laser plume including matrix

reactive species.

These different steps frequently have been often considered and studied
separately isolated, i.e., the transition from the solid state to the gas phase from
a physical point of view or the ionization from a chemical or thermodynamic one.
One major consequence of the separation in single mechanistic step was that the
details of the solid-gas phase transition were not considered with respect to their
relevance for ionization. As a consequence, a “photochemical” ionization
mechanism was favored in the initial MALDI papers, considering the generation

of neutral analyte species as the first step and then ionization by proton-transfer

18

coming from matrix species in the gas phase matrix plume. Seldom were other

possible suggested mechanisms taken into account [4648].

A major advance in models of MALDI was to include cluster formation and
generation of charged clusters by either charge separation (deficit/excess of
anions or cations) or photochemical processes as an important primary ionization
step. This model (the “lucky survivor" model) tries to be comprehensive and
general for MALDI-MS ionization [47]; the key point of this model is to consider
ionization during desorption as a dominant ionization and charge-separation
pathway. Recently, a quantitative model of ionization in ultraviolet MALDI was
presented. This model was focused initially on primary ionization in a MALDI
sample containing only matrix. Later, it was extended to include ion-molecule

reactions [49, 50].

However, even after all these proposals for MALDI models, the

mechanism for ionization in MALDI is still under investigation [38].

From a practical point of view, MALDI sample preparation techniques are
vital to forming a homogenous solid sample that yielded imporved laser shot-to-
shot reproducibility. Sample preparation and recent improvements are given
elsewhere [45, 51]. Development of free-matrix solid surfaces where the sample
is directly deposited for analysis have also been developed [52, 53]. Although

MALDI is not easily coupled to separation methodologies, this technique offers

19

several advantages: 1) Biological samples can be examined without extensive
purification because MALDI is somewhat tolerant of buffers and salts, (although
in order to allow the desorption of proteins and DNA in a large excess of salt or
surfactant, different MALDI probes have been recommended [52-561). 2) It has a
low sample requirement, usually 1-10 pmol of analyte is enough to obtain a good
quality mass spectra. 3) Proteins with masses ranging to greater than 100 kDa

can be analyzed.

3.2c Electrospray Ionization (ESI)

The development of electrospray ionization (ESI) started with the research
of Dole et al. in the 19603 who described the idea that electrospraying a liquid
containing analyte molecules might liberate these as ions and make them
amenable to mass spectrometry [57, 58]. Fenn et al., in 1984 [59-61]
demonstrated the potential of ESI as a mass spectrometry technique for small
and large biomolecules [34, 62]. Independently and almost at the same time,
Aleksandrov et al, coupled a liquid chromatography micro-column to a mass
spectrometer where the ions were extracted from the eluent solution by spraying
into a non-uniform electric field. They observed protonated and cationized

products of small molecules [63, 64].

Electrospray ionization (ESI) is a soft ionization technique in which ions

that are in solution are put into the gas phase through a mechanism of ion

20

desorption or ion evaporation. The sample solution is sprayed across a high
potential difference (a few kilovolts, 3-8 W) from a narrow-bore needle (0.1-0.3
mm) into an orifice in the interface. In conventional ESI, heat and gas (N2) flows
are used to desolvate the ions existing in microscopic droplets of the sample

solution (See Figure 6).

 

 

 

 

 

 

 

 

N2 (80°)
+ 3 to 8 kV
I Nozzle [I Nozzle
‘ I
Skimmer
Spray
_ , f Analyzer

1

 

plll

Vacuum
Pumps
N. (80°)

Needle

 

 

 

 

 

 

 

 

 

Figure 6. Scheme of an electrospray source (adapted from reference [65])

The emerging liquid at the tube exit starts forming a cone under the
influence of increasing field strength. When a certain field strength is reached,
the equilibrium between surface tension and electrostatic forces becomes
independent of the radius of curvature; theoretically, the radius eventually could

become zero. However, in a real system, at the moment the critical electric field

21

is reached, a conical meniscus that is known as a “Taylor cone” is formed.
Immediately, a fine jet of liquid ejects from the cone towards the counter
electrode (See Figure 7). The jet carries a large excess of ions because it
emerges from the cone’s tip that is the point of highest charge density. Such a

jet cannot remain stable for an elongated period, and breaks up into small

 

 

   

droplets [66].
Reduction _-
o
0
6
Taylor cone 0 0 G
0
a
$3?
5%;

  

Electrons
0
C3

 

 

 

19
High Voltage
Power Supply ———->
~ _ Electrons

TDC

 

 

 

 

 

 

Figure 7. Schematic of Taylor cone formation (from [66]).

The initial fine droplets formed at atmospheric pressure in the spray are
enormous at the macromolecular scale; these droplets possess an excess of
positive or negative charges depending on the capillary bias polarity.
Qualitatively speaking, the charged droplets evaporate to a point where the

number of electrostatic charges on the surface becomes so large relative to the

22

droplet size that an explosion occurs (close to the point that is known as the
Rayleigh limit) to produce a number of a smaller droplets. These droplets also
contain electrostatic charges [39, 67]. This phenomenon is called Coulomb
explosion or Rayleigh instability, and it was captured on film by R. Gomez and

Tang in an ESI source (see Figure 8).

 

   

 

 

 

Figure 8. A droplet in the throes of what is known as Rayleigh Instability, and sometimes
referred to as Coulomb explosion (from reference [68]).

From a quantitative point of View, for fission to occur, it is necessary that
the solvent in the parent droplet evaporate until its radius reaches a value R that
leads to the first fission. Droplets close (~80%) and not exactly at the Raleygh
limit, with sizes in the 1-pm range, undergo fission because the droplets are
deformed mechanically. Typically, the droplets are observed to vibrate
alternately from oblate to prolate shapes and these elastic vibrations stimulate
disruptions in which the “parent" droplet emits a tail of much smaller offspring
droplets (see Figure 8). When relatively volatile solvents such as methanol,
acetonitrile, and water are involved and the droplets are a few micrometers in

diameter, the evaporation rate follows the surface evaporation limit law, which

23

leads to a simple dependence of the radius droplet with time t (in s), for
methanol: R (m) = R0 (m) — 1.2 x 10'3 (m/s) t(s), (where R0 is the radio at time
zero and R after a time t) [69]. Within ~ 100 ps, the radii are sufficiently close to
the Raleigh limit where the Coulombic repulsion between the charges overcomes
the cohesive force of the surface tension. In this process, a stream of smaller
offspring droplets is emitted by the parent. The offspring radius is about one-
tenth the parent radius. The offspring carry off 2% of the mass and 15% of the
charge of the parent droplet. The decomposition scheme, shown in Figure 9,
illustrates the cascade of fissions of a parent droplet. The offspring droplets, on
solvent evaporation, undergo Coulomb fissions, which leads to second
generation offspring which have only a few elementary charges. However, this
decomposition scheme is only partially valid because the data of Gomez and

Tang were obtained for larger droplets and different solvents [70, 71].

The next step in the process is the production of gas-phase ions. The
mechanism is not completely understood, but there are two main proposed
models, both concerning the production of small droplets by means of the
Coulomb fissions. However, they have substantial differences. The first model,
the Charged Residue Model (CRM) originally proposed by Dole, holds that the
sequence of Coulomb fissions produces ultimate droplets, each of which
contains only one molecule of solute. That molecule becomes a free gas-phase

ion by retaining some of its droplet’s charge as the last of its solvent evaporates

[72].

24

 

Parent droplet created near

   

the ES capillary tip with radius . solvent . . Coulomb +
(R)I1.5mmand numberof ? , evaporation. - sson 9909"-
cha s N 51250 '
'9' H. "46216 hump", RI0.939um R80.09um
".51250 NI43560 NI384
solvent
evaporation
Consecutive solvent " 74 "‘3
evaporation and 4— eeee. . . ... . % Consecutive solvent
Coulomb fissions R a 0 08 pm R _ 0 844 m R _ 0 848 um evaporation and
' ' " ' C l b fi l
NI326 NI37026 NI43560 “mm ”a”
solvent
evaporation
tI 70 us
.. + o «—222 e
R I 0.07 pm R I 0.756 m R I 0.761 um
N I 278 N I 31472 N I 37026

I 1

Consecutive solvent Consecutive solvent
evaporation and evaporation and
Coulomb fissions Coulomb fissions

 

 

 

Figure 9. Droplet evolution scheme due to evaporation at constant charge and Coulomb fissions
close to the Raleygh limit. Adapted from reference [71].

The second model, the Ion Evaporation Model (IEM) originated by lribame
and Thomsom [73] assumes the same sequence of evaporation, but argues that
before a droplet reaches the ultimate stage contemplated by the CRM (R z 10 —
20 nm), the field on its surface becomes strong enough to overcome solvation
forces and lifts a solute ion from the droplet surface into the ambient gas [72].
Other papers just assert that the gas phase ions produced by ESI are simply the
same ions that are in the bulk solution from which the droplet was formed, but

that is really a simplistic approach.

25

Most reported applications of ESI-MS involve molecules in which the
charge is most often generated by protonation of basic sites (e. 9., amino groups)
or deprotonation of acidic sites (e.g., phosphates). Thus, ESI mass spectra in
general correspond to a statistical distribution of consecutive peaks characteristic
of multiply charged molecules obtained through protonation (M + zH)2+ or
deprotonation (M — 2H)". Unfolding the proteins, due to reduction of disulfide
bonds in a cystinyl protein, could expose additional basic sites to protonation,
which results in higher average charge states in the corresponding ESI spectrum.
Because these multiply charged species differ only by a consecutively
decreasing number of charges as m/z increases, it is possible to deduce the
mass of the intact molecule. Examples of how to calculate the mass of the intact

molecule from ESI-MS spectra are given in detail in textbooks [38, 65-67].

Low-flow electrospray techniques that do not use a make-up solvent have
been developed successfully. Microelectrospray, the first low-flow-rate ESI
technique, was described using flow rates between 300-800 nL/min [74].
Currently, however, the most common flow-rate technique is nano—ESI; which
differs from micro-ESI in the diameter of the needle; flow rates are typically in the
range of 20-40 nL/min [64, 75]. The advantages of nano-ESI in comparison to
the high-flow rate techniques can be rationalized based on the fundamentals of
the ESI process. A lower solution flow rate is known to reduce the size of the
charged droplets initially produced in the spraying process. The smaller initial

droplets require fewer droplet fission events and less solvent evaporation prior to

26

ion release into the gas phase. As a consequence, a larger portion of the analyte
molecules present in the primary droplets is made available for MS analysis
whereas the increase in concentration of contaminants is significantly reduced.
Thus, nano-ESI may be considered to be more than just a miniaturized technique

[46].

3.2d Atmospheric Pressure Chemical Ionization (APCI)

Like electrospray, APCI creates gas-phase ions at atmospheric pressure.
As a consequence, the vacuum and gas dynamics of the interface for APCI are
identical to those for electrospray. The sample is sprayed into a heated chamber
(~400°C), the ion source. The analyte molecules are volatilized. Nitrogen and
oxygen molecules in the ion source are ionized with a corona discharge and
these ions react with solvent molecules, which are in the gas phase, to form
reagent ions. These reagent ions react with the analyte molecule to produce the

analyte ions [39, 76].

 

    

Discharge .. - 2

 

. corona
.’ e . . .
0.. . z: . ' ’r. | —-> to Mass Analyzer
. d
e 9 '. e 0 .e
' e

   

 

 

 

Figure 10. Conceptual representation of Atmospheric Pressure Chemical Ionization.

27

3.2e Electron Capture Dissociation (ECD)

Electron capture dissociation is a new fragmentation technique used in
Fourier transform ion cyclotron resonance mass spectrometry, and it is

complementary to traditional tandem mass spectrometry techniques.

This technique was introduced by Zubarev, Kelleher and McLafferty in
1998 [77]. A multiply charged cation captures an electron to produce cationic
dissociation products. The cations are generated by electrospray ionization,
transferred to the vacuum system of a mass spectrometer, and trapped by
electrostatic and magnetic fields in an ion-cyclotron resonance cell. Polypeptide
polycations, upon exposure to low-energy thermal electrons (< 0.2 eV), initially
capture an electron in a high orbit, which is followed by charge neutralization,
leading to an excited radical species that rapidly (10'11 s) undergoes bond
cleavage. A fraction of ions are reduced to cation-radicals that in part dissociate
by cleavage of the N-Ca bonds of the peptide backbone, Ca-alkyl side chain
bonds, or cystine S-S bonds [78, 79]. in general terms, the presence of a radical
makes this fragmentation very specific, giving cleavages between most of a

protein’s amino acids.
These days, the mechanism of ECD is poorly understood. Zubarev et al.

postulated that this specific and unusual reactivity is due to the fact that ECD is a

nonergodic process where the aminoketyl radical dissociation occurs before the

28

excitation energy (relatively large ~ 6 eV) is randomized, in other words
dissociation happens before the reactants undergo internal redistribution of
vibrational energy, minimizing the effect of differences in the backbone
dissociation energies [77]. This hypothesis for peptide-radical dissociations was
based in part on density functional theory calculations in an aminoketyl radical
model corresponding to a hydrogen atom adduct to N-methylacetamide, where
N-CH3 bond dissociation was predicted to dominate the spectrum. However,
concerns exist about this postulate because non-ergodic dissociation has been
observed only for small molecules; thus, other research groups claim that
electron capture produces a labile free radical species for which backbone

cleavage is the lowest energy reaction [79-81].

From ECD of peptides and proteins it is possible to distinguish different

cleavages:

1) Backbone Cleavage: This process produces c (peptide fragment with
an enolimine group) and 2' fragments (containing an a-radical site) and ~ 10% of

a' and y ions [79, 82] (see Figure 11).

29

 

 

 

m’ (n-m)+
R1 0
T )2 ,, l
m. ‘fir OH
0 R4
2‘ ion
or
R. on "" mm". ”'2
H Residue1'i—Residuez 327100003
1n" 'so-‘lliam
N4
0 R2 b-c=-l7.03Da
y—z'=+16.02Da
6 Ion

 

 

 

 

 

[1+
R1 H 0 R3 H 0 ‘I
N N
0 R2 0 R4

m+ O R O _| ("NH
R] 3
HZN

CEO
Z:

12

HzN

a' ion y ion

 

 

 

Figure 11. Ions produced by cleavage of the backbone chain in a peptide or
protein. Adapted from [82] and [83].

30

There are no favored sites of cleavage within peptides, except that amine
cleavage within proline is rarely observed. The extent of the cleavage in the
original ECD seemed to be dependent on the size and conformation of the
protein, for example, the examination of the ECD spectra of five 12- to 17-mer
peptides showed complete sequence coverage. Backbone cleavage at 16/20
sites was observed for a 21-mer peptide, but only 33 of 152 residue pairs were
cleaved in a 17 kDa myoglobin, and none in several larger proteins examined. It
has been proposed that the lack of backbone fragmentation in large proteins (>
20 kDa) is due to the presence of non-covalent interactions within the ion, i.e.,
electron capture results in cleavage of backbone bonds, but does not disrupt the
non-covalent interactions. In order to overcome this limitation, “activated ion" (Al)
ECD was developed. In this modification, the ions are heated while they undergo
electron capture. Activation methods include collisional activation with a
background gas, infrared irradiation and blackbody irradiation [83]. Al conditions
were sought to denature the center portions of the large ions and help the
electron capture dissociation process. Proteins as large as 45 kDa have been
analyzed with Al-ECD obtaining acceptable results showing some of the
expected ions due to the polypeptide cleavage, and in some cases after
optimization, noticeable improvement in the backbone cleavage was found [84-

86].

2) Disulfide Bond cleavage: ECD seems to be a good alternative for the

analysis of disulfide linkages. The capture of an electron preferentially cleaves

31

one disulfide bond in proteins. This approach is interesting because intact

disulfide bonds can be analyzed in peptides or proteins. [78, 83, 87]

 

"1+
8.1

/
R1 R2

1

(m-1)+ . (m-1)+
/SH_| s\_|
R1 R2

 

 

 

Figure 12. Disulfide bond cleavage using
ECD-MS.

3) Side-chain cleavages: These kinds of cleavage are not minor, and in
some cases, they show abundances comparable to those coming from backbone
cleavages. All peptides containing arginine, histidine, asparagine or glutamine
show side-chain losses associated with those residues, regardless of the residue
position within the peptide. For example, arginine side-chain cleavages result in
losses of CH4N2 (44.037 Da), CH5N3 (59.048 Da), and C4H11N3 (101.095 Da).
Losses associated with lysine and methionine were observed, but because these
losses have not been found in all peptides containing those residues, so the
losses cannot be considered of absolute diagnostic value. Other side-chain
cleavages with possible diagnostic value have been reported, but currently it is

quite early to know if those losses are always present in the ECD process.

32

3.3 Mass analyzers

Ionized matter can be affected by electric and magnetic fields. Thus, after
matter is ionized in an ionization source, ions are subjected to electric and
magnetic fields in order to separate them. Acceleration, energy and momentum
induced by magnetic and electric fields are dependent on the charge on the ion,
thus mass analyzers separate ions according to mass-charge (m/z) ratio, not

according to their mass only.

In principle, all ionization techniques could be coupled to any mass
analyzer, but some couplings work better than others because of the particular
characteristics of the mass analyzers, the ionization techniques, and the type of
information under investigation. Mass analyzers have some common
characteristics: 1) mass range (maximum m/z detectable), 2) resolving power
(ability to separate ions of adjacent W2), 3) accuracy of mass measurement, and

4) ion transmission, sensitivity [38, 88].

Next, there are brief descriptions about mass analyzers commonly found

in the analysis of biomolecules.

33

3.3a Quadrupole Mass Analyzer (O)

This instrument separates ions based on radiofrequency (rf) oscillations in
an electric field (the quadrupole field) superimposed on direct current (DC) _
potentials. This mass spectrometer employs a combination of DC and rf to filter
out all ions except those of a given m/z value at a given set of rf and DC
amplitudes. A scan of the amplitudes of rf and potentials allow ions of a
sequential range of m/z values to reach the detector to generate a complete

mass spectrum.

Quadrupoles consist of four metal rods held in strict alignment with one
another. Opposite rods are connected in pairs to both radio frequency and direct
current potentials. According Figure 13, ions coming form an ionization chamber
are accelerated at the entrance of the quadrupole by means of an acceleration
potential in the 2 direction. Within the quadrupole, ions experience no forces in
the Z axis, however, ions traveling along the z-axis are subjected to a total
transverse electric field resulting from the application of the potentials upon the
rods:

¢o=+(U-Vcoswt) and -¢o=-(U-Vcoswt)

4)., represents the potential applied to the rods, 0 is the angular frequency
(in rad/s) = 21: v, where v is the frequency of the rf, U is direct potential and V is

the “zero to peak" amplitude of the rf voltage [38, 67, 89, 90].

34

 

 

 

 

 

 

Figure 13. Quadrupole mass filter.

Depending of the potential applied, ions within a small range of m/z values
have stable paths through the quadmpole and all others ions are not transmitted.
Ideally, for specific values for V and U, only ions of one m/z value should be able
to have stable trajectories along the X-axis and along the Y-axis, thus those ions
could travel in the Z-direction and by consequence, be able to reach the detector
after exiting of the quadrupole mass filter. On the other hand, ions with unstable
trajectories either in the X-axis or Y-axis, will strike the rods in the quadrupole,

become discharged and not detected.
Equations of motion for an ion can be depicted in the form of the Mathieu
equation (which was established in 1866 by the French physicist E. Mathieu):
— + (au - 2qucos 2§)u = 0

where u stands for either x or y. These equations are quite complex and

derivation implies solution of differential equations that produce a set of variables

a, q, and 2;:

35

 

 

From these variables, we can see that it is possible to establish a
relationship between the coordinates of an ion and time using the Mathieu
equation; detailed integration and analysis of the motion equations can be found
elsewhere [88, 91, 92]. For ions of a given mass, certain values of a and q lead
to stable oscillations in the x and y directions. For a given quadrupole, ro is
constant, 0 is maintained constant and U and V are the variables. Since a, q, U
and V are interrelated, m/z, varies with either U or V. The relationship between
all of these variables and the determination of m/z can be understood in terms of

a plot of 3 vs q (See Figure 14).

 

0.2 -

x stability boundary

0.| '-

 

 

 

\
\/ \ / \
/ \4 /\ \
\/1 \\/ 1 \4/ i\./ l \1/ \ l
0.2 0.4 0.6 0.8 l.0

y stability boundary

 

 

 

Figure 14. Quadrupole stability diagram. (Adapted from [38, 88, 92])

36

This graph is the collection of points in a-q space that correspond to stable
solutions of the equation of motion (shadowed area in Figure 14) in physical
terms; this diagram is the reduction of a six-dimensional problem (involving 2, m,
U, V, r0, and 0) to a two-dimensional problem involving the reduced parameters a

and q.

We can describe the operation of a quadrupole as a narrow bandpass

mass filter or as rf-mass filter in terms of the a-q stability diagram.

In the first case, let’s consider that, in principle, it is possible to operate a
quadrupole in a manner in which the parameters a and q are, at all times,
independent from one another. In practice, quadrupoles are operated in a
manner that the values of those parameters are always related by a simple ratio

regardless of the actual magnitude of either a or q:

2U

a _

If " 7
In terms of the stability diagram, the above equation represents a straight

line with a zero intercept that is called operating line (see Figure 14). This line

can be visualized from a practical point of view like a collection of different

masses that might be stable if they are inside the stabilization diagram, for

example, in the diagram, masses m+1 and m+2 are stable (they are inside the

region A), on the contrary; masses m and m+3 are not stable. If the slope of the

operating line is increased, the UN ratio is higher, the area A will become smaller

37

and it would represent the case in which only a few masses (maybe only m+1)
will have a stable trajectory, meaning that the resolution of the quadrupole is
increased. From equations for a and q, it is known that U oc (m/z)a and V oc
(m/z)q, then if U or V are increased (keeping constant 0 and r0), m/z values
should be increased also. Thus, the most convenient method of scanning the
bandpass region is by sweeping the voltage applied; consider that if potentials U
and V are doubled, the masses that appear in the area A must be doubled,
therefore, a potential increase is equivalent to sliding the operation line. The
complete mass spectrum is obtained by scanning the magnitude of V and U
through increasing or decreasing values; by scanning the magnitude of rf and DC
potential from a low to high value, ions of increasing mass will sequentially take
on stable values of a and q space and be transmitted to the detector. For

example, for ions in Figure 14, the order would be: m —> m+1 —> m+2 —+ m+3.

For the case of using the quadruople in the rf-only mode,.the stabilization
diagram is very useful, because this operation mode is represented when U = 0,

it means a = 0 (see Figure 15).

We can observe in this way that a large portion of the operating line falls
within the stability region; it predicts that a large number of ions with different m/z
values have stable trajectories within the quadrupole. Ions of all m/z ratios
higher than a particular lower limit will have a stable trajectory (in this case,

higher than m-1). In this mode, the quadrupole causes all ions to be brought

38

back systematically to the center of the rods, even if they were deflected by a
collision. This focusing effect is important to increase the transmission of ions

after collisions [65, 67, 91, 92].

 

0.2 L

 

 

 

m+1
m+2

............ m+3

 

 

 

Figure 15. Quadrupole stability diagram when working in rf-only mode (a = 0). (Adapted
from [38, 88, 92])

Quadrupole mass spectrometers are symbolized by “Q” and rf-only
quadrupoles by “q”; these letters are commonly used in hybrid mass
spectrometers where quadrupoles are an essential part of the instrumentation
setup. Quadrupole-based instruments are among the most sophisticated
instruments for structure elucidation, and for this reason, it is necessary to have a
good understanding of the theory associated with the design and operation of

quadrupoles [91].

39

3.3b Quadrupole Ion-Trap (QIT) or Quistor (Quadrupole Ion Storage)

The invention of the ion trap (in 1953) by Wolfgang Paul and Hans
Steinwedel, was recognized by the award of the 1989 Nobel Prize in Physics. In
1983, the mass-selective instability scan by George C. Stafford, Jr. was
announced. On these two landmarks rests the entire field of ion trap mass

spectrometry [93].

In the quadrupole ion trap, the ions are trapped in a small volume between
a ring electrode and two end-cap electrodes in a three-dimensional space. Like
the quadrupole mass filter, it uses oscillating electric fields (rf) to trap ions in a
controlled manner. The common operation mode of an QIT is with the potential
applied to the ring, while the end-cap electrodes are grounded except when

auxiliary potentials are applied to either or both end-cap electrodes [94].

In quadrupole instruments, the potentials are adjusted so that only ions
with a selected mass go through the rods, in the case of QIT, ions within a range
of m/z values are stored in a field that is created with a fixed-frequency rf applied
to the cylindrical ring electrode (see Figure 16). As the amplitude of this fixed-
frequency rf is increased, the trajectories of ions on increasing m/z values will
become sequentially unstable and move toward the exit end cap electrode, which
has holes to allow ions to leave the QIT, and go to the detector to be recorded

[39, 94].

40

 

Helium
(dampening gas)

Ion

Ion
Injection '

Injection

 

RF generator

    

Entrance Exlt
endcap endcap
Ion Injection .................... p ................... * Ion Election
20
Ring
Electrode

L AC or DC generator-J

 

 

 

Figure 16. Schematic diagrams of an ion-trap; 20 is distance from the center of trap to
apex of the end-cap. r0 is the internal radius of the ring electrode. (Adapted from [93]).

There are two types of QlTs, 1) instruments where the primary ionization
takes place in the space where ions are stored, 2) instruments where the primary
ionization takes place outside the QIT; these last types of QlTs are currently the

most common in mass spectrometry instrumentation applied to biomolecules.

The mathematical analyses using the Mathieu equations allow us to locate
areas wherein ions of given masses have a stable trajectory. The equations are
very similar to those used for quadrupoles, however, in the ion trap, the motion of

the ions occurs in three dimensions (see A in Figure 17).

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

‘“ l l’
—3-DTrajectory
E
:5
Q
3
“U
C
2‘.
N
0-2 -— 1.0
0.0
B‘ B!
0.0 — .............................
-0.2 "‘
d“ ..
-0.4 —-
_ a _ 16zU
Z _ ' 2 2 2
-0.6 _ m(ro +2z0 )0)
q __ 82V
2 m(r02+22§)al)2
-0.8 I I l I a I t I I 1 I , I
0.0 0.2 0.4 0.6 0.8 1.0 1.2
9:

 

Figure 17. A) Simulation in two- and three- dimensions of the trajectory of an ion with a
m/z value equal to 500 (Diagram from reference [94]). B) Stability diagram for the ion-
trap, subscript 2 represent axial motion between the endcaps. U is the DC potential on
the endcap electrodes, V is the rf potential applied to the ring electrode, ro is the radius
of the ring electrode, 0 is the rf angular frequency, 2 is the charge and m is the mass.
Parameters B are additional trapping parameter to q and a, to describe the ion stability

in the diagram (From reference [95]).

42

 

In the stability diagram presented in Figure 17, it is interesting to note that
the stability region is only a small area in the a-q field. The parameter [3 is an
additional trapping parameter to q and a to describe the ion stability in the
diagram. It comes from the complete integration of the Mathieu equation which
requires the use of a function e‘°‘””’. The behavior of an ion can be depicted by
means of the parameter [3 value. As the value of [3 approaches to zero, the ion
is not affected by the oscillating fields; when [3 is equal to 1, the ion oscillates in
resonance with the rf field and absorbs kinetic energy to the point that its
magnitude of oscillation increases so that the ion escapes the trap or is
discharged by collision with the end cap surfaces. For B values between 0 and 1,
the ion oscillates in a periodic mode caused by the oscillating fields with an

average kinetic energy related to the value of B [38].

A relatively high pressure (10‘2-10'3 Torr) of inert gas, usually He, is used
to dampen the motions of the ions, confining their coordinates to the center of the
trap. Under normal operation, ions will have a sufficient [3 value so as not to
have enough kinetic energy to fragment upon collision with the inert gas.
Increasing the value of B for specific ions, it is possible to establish appropriate
conditions for ion/molecule experiments or collisionally activated dissociation

(CAD or sometimes called CID, from collision-induced dissociation).

43

It is possible to distinguish four different ion traps functions:

Collection: Ions are injected and cooled by means of the inert gas in the
center of the trap.

Isolation: For tandem experiments (MS") or for Selected Ion Monitoring
(SIM), a selected ion of a m/z value is stored in the trap, all the others ions are
ejected from the trap.

Excitation: For tandem experiments, after ions of one specific m/z value
are isolated, they can be excited and collided with the inert gas to obtain product
ions.

Ejection: Ions in the trap are ejected and detected to get the mass

spectrum.

These functions are programmed depending on the desired experiment.

There are two main drawbacks for the QIT, 1) Limited storage capacity
and 2) Inefficient trapping. In the first case, there is a maximum number of ions
that can be stored within a trap at any one time. When the QIT is filled beyond
this maximum, “space charge" occurs. This phenomenon is observed when
repulsive forces cause leakage of the excess ions in a trap. Concentration of
ions above the space-charge limit leads to poor performance of the instrument,
lowering resolution and sensitivity, and sometimes affecting calibration (mass

accuracy). In a 3D trap, the presence of a 3D rf field is unfavorable for ion

44

trapping, for example, during the positive cycle of the field, positively charged
ions are repelled from the trap by the positive electrical field, and only when this
field is negative, positively charged ions are pulled through the trap. However,
the possibility of isolation and fragmentation of specific ions (tandem in time
mass spectrometry) make the QIT a versatile tool of ample application in mass

spectrometry [38, 39, 93].
3.30 Linear (Quadrupole) Ion Trap (LIT)

Linear ion traps have found new applications in mass spectrometry
recently, either combined with other mass analyzers such as time-of-flight (TOF),
quadrupole ion traps (QlTs), or ion cyclotron resonance (ICR), or as stand-alone
mass spectrometers with radial or axial mass-selective ejection. In a LIT, ions
are confined radially by a two-dimensional radiofrequency field and axially by
stopping potentials applied to end electrodes. In comparison to QIT, LIT have
higher injection efficiencies and higher ion storage capacities. Their use is not
limited to simply storing ions, they can be operated as stand-alone mass

spectrometers [96].

Many of the techniques used to manipulate QITs can be applied to linear
quadrupole traps. These include resonant ion excitation for ion activation,
isolation, or ejection, and mass-selective instability scans with ejection radially

through a rod (or rods, see slot in Figure 18) or axially.

45

 

 
 
 
  
      

R0 = 4 mm 12 mm
Hyperbolic Rods
X0 > Yo

Back
Section

  
   

Center
Section

2 Front Slot: 30 mm x 0.25 mm
Section

 

 

 

Figure 18. Linear Ion Trap from Thermo-Finnigan [93].

The advantages associated with trapping ions in a LIT are several-fold,
because they have no quadrupole field along the z-axis. Ions admitted into a
pressurized LIT undergo a series of momentum-dissipating collisions effectively
reducing axial energy prior encountering the end electrodes thereby enhancing
trapping efficiency. The larger volume of the LIT relative to a QIT means that
more ions can be stored prior to undergoing charge space effects. Radial
containment of ions within a LIT results in strong focusing along the centerline of
the trap in contrast to the QIT in which the fields tend to focus the trapped ions to
a point (see Figure 19). Line rather than point-focusing properties may have
some influence on the relative susceptibilities to space charge phenomena [97],

in addition, conventional 3-D ion traps have a slower scan rate.

46

 

Ring Electrode

   

 

: §:3::2::. . 3.2::

._T__

 

 

3D—Trap: Ions need to be 2D Trap. 10115 can
confined to a small spread out in the axial
volume at center of trap dimension Without

for optimum performance degradmg performance

 

 

 

Figure 19. Comparison of focusing in a QIT and a LIT, from reference [93].

These problems with 3D ion-traps may contribute toward a reduced
number of peptides identified as well as limited quality fragmentation spectra
during proteomics experiments. Functional enhancements of the LIT include 15x
higher ion capacity, 3x faster scan rate, up to 100% detection efficiency, and up

to 70% trapping efficiency.

In a LIT, mass-selective radial ion ejection involves trapping the ions
within the four-rod quadrupole structure and resonantly ejecting the ions radially
through a slot cut in one of the quadrupole rods. In the commercial LTQ (from
Finnigan), the ions are moved from within the stability diagram to a position
where they become unstable in the x-direction and leave the trapping field for
detection through double slot cuts; from there, ions are accelerated into two high
voltage dynodes where ions produce secondary electrons. This signal is

subsequently amplified by two electron multipliers; the analog signals are then

47

integrated and digitized. This dual detection helps to increase sensitivity in the

process.

Mass-selective ejection from a linear trap is a less obvious technique for
ion extraction, but it is analogous to the rf-only transmission in a quadrupole
mass spectrometer. Because axial ejection linear ion trap operates in the low 10’
5 torr regime and ions emerge from the end of the device; it has proven to be a

good match with the ion path of a QqQ instrument [98].

Most other features and operations of HTS are similar to those of the
QlTs, and have been described in detail in several articles [96-101]. A
systematic comparison of LITs and QlTs in proteomics experiments found a 4-6
fold increase in the number of peptides and proteins identified on the LIT mass
spectrometer and, interestingly, more than 70% of the double and triply charged
peptides, but not singly charged peptides, showed better quality of fragmentation

spectra on the two-dimensional ion trap [98].

48

3.3d Time of Flight (TOF)

This instrument uses no external force to separate ions of different m/z
values. Ions are accelerated into a flight tube at values from hundreds to
thousands of volts (typically 20 W). The operating principle of a TOF mass
spectrometer involves measuring the time required for an ion to travel from an
ion source to a detector located 1-2 m from the source. All ions receive the same
kinetic energy during instantaneous acceleration (e. 9., 3000 eV), but because
they have different m/z values, they separate into groups according to velocity
(and hence m/z). The ions sequentially strike the detector in order of increasing

m/z value.

The simplicity of the physical principles of TOFMS is part of its current
success. The mass-velocity relationship for constant energy ions and the
electrostatic force on a charge when combined with the Newtonian mechanics

provide the foundations of all TOF mass spectrometers:

 

where: q = charge (Coulombs)
qV = lmuz V = electrical potential
F = Eq m = ion mass
F = ma u = ion velocity
= E F = force experienced by charge
m E = electric field strength
a = acceleration

 

 

 

49

An ion with mass m and total charge q = ze has a kinetic energy:

The time (t) needed to fly the distance (d) is giving by:

Thus,

 

 

mu2

E: = V=zeV
2 Cl

 

 

e = charge of an electron

z = charge state of the ion

 

 

 

 

d 2
M = qv =Zev

2
m d 2_
3%) —zeV

 

 

 

:Io.

 

 

 

 

This equation shows that m/z can be calculated from a measurement of t,

because d, and V are kept constant. As example, the transit time at 3000 eV of

an ion of m/z 800 is approximately 70 psec through a 1-m flight tube [38, 67,

102].

50

 

Ion source . 7 ,d-O

 

....C”.
”ONO”!
0. m0!

Drift tube ’ .

d >
Flight Tube

 

 

Detector

 

 

 

DO enema
raccoon...
”mo“.

 

 

 

Acceleration
gflds

 

 

 

Figure 20. Scheme of a Time Of Flight (T OF) mass spectrometer.

In principle, the upper mass range for a linear TOF has no limit. The most
important drawback of a linear TOF analyzer is its low resolving power; four basic
phenomena are responsible for this (see Figure 21) [76]:

1) Temporal distribution: Two ions of the same mass formed at different

times in the source, thus, they arrive at the detector at different times.

2) Spatial distribution: Two ions of the same mass formed at different

locations in the source arrive at the detector at different times.

3) Kinetic energy distribution: Two ions of the same mass formed with

different initial kinetic energy arrive at the detector at different times.

4) Tum-around time effect: Two ions of the same mass with the same

kinetic energy. but initial velocities in opposite directions arrive at the detector at

different times

51

 

 

1)

2)

 

4)

 

 

 

 

 

 

 

Ions formed at: t, > t,
rt1>toandt°=0 /—\ ta>t2
i t I 2 :
i @—> E ®—*
5 i g 5 * Detector
Ions formed at
different starting
locationsettoso t1>to ta":
, ' i i i I Detector
ions formed at
to = 0 with different t,> to '1 > t1,
: : E :
(la—4» @—>
i
7 i s a 5 * Detector
ions formed at the same
time to = 0 with equal
kinetic energy t, > to t, > t, 7
. s f a g
a} (39*
Ci, e—s, ®—+
, ' 5 3 E i ' Detector

 

 

 

Figure 21. Phenomena that affect resolving power in a TOF mass spectrometer. 1)

Temporal distribution 2) Spatial Distribution 3) Kinetic energy distribution 4) Turn around

time effect.

Several techniques have been developed to correct these problems,

spatial variations were improved by a unique two-field extraction and acceleration
source, energy variations and turn-around time were corrected by a unique
technique called time-lag focusing. These techniques can be found nicely
described in some textbooks [38, 66, 67] and articles [102-106]. Another major
improvement in resolving power involves the use of an electrostatic reflector, also
called a reflectron that was proposed by Mamyrim and Shmikk [107]. Let’s

considerer that isomass ions, with a distribution of energies, travel at different

52

velocities. After traversing the field-free region, they reach a series of grids and
ring electrodes of increasing potential (the reflectron) that create a retarding field
that acts as a mirror by deflecting the ions and sending them back through the
field tube; indeed, ions with more kinetic energy penetrate more deeply into the
reflector, spending more time, therefore they will reach the detector at the same
time as slower ions (see Figure 22). However, sensitivity of this mode of

operation is low in comparison to the usual linear mode [38, 66, 67].

 

 

Ion source

83
:0

“CO“....
.

 

 

8
0
£03..

0.

O O I O

......
......
......
......
IIIIII
A AAAAAA

 

 

 

 

 

 

 

Detector Reflector region

 

Figure 22. Schematic diagram of ion trajectories in a reflectron-Time-of- Flight mass analyzer.

A major breakthrough has been the development of an orthogonal
acceleration TOF (oa-TOF) analyzer, where “tum-around time” is not a problem.
In this mass analyzer, pulsed ions are extracted orthogonally (perpendicular)
from a continuous ion beam [66, 104, 106, 108]. Ions are sampled from a nearly
parallel ion beam (see Figure 23), the low-energy ion beam is allowed to fill an
ion accelerator region while it is held in a field-free state (in x-axis). Then, a
package of ions is pushed out orthogonally (in y-direction) by means of a sharp
pulse and it is accelerated into the TOF analyzer (y-direction). As the beam is

nearly parallel, the ions have nearly zero average velocity and minimal velocity

53

spread in the direction of the pulse applied. After the heaviest ions have reached

the detector, the next package is pulsed into the TOF-analyzer (Figure 23).

g Detector

 

 

 

 

 

y
c A
.9
E
1:
8
i— p x
Direction of beam
Beam optics V t,
ion [ml] 3:21;: Acceleration
Sourcg V ' ............ grids

—

   

 

 

Orthogonal
accelerator

 

 

 

Figure 23. Scheme of a linear oa-TOF. to, t1 and t2 represent the flight time of ions after the
orthogonal pulse.

Advantages of oa-TOF are: 1) high sensitivity, 2) high rate of spectra per

second, 3) high mass-resolving power, 4) mass accuracies up to 5 ppm, and 5)

compact design [66, 104, 106].

3.3e Fourier Transform Ion Cyclotron Resonance (FT-ICR)

The Fourier transform ion cyclotron resonance (ICR) mass spectrometer is
a magnetic ion trap. The analyzer employs a powerful magnetic field (3-7 Tesla).
Recall from physics that a charged particle precesses in a magnetic field at a
frequency inversely proportional to its mass, this motion is known here as the
“cyclotron motion”. Formerly, an ion of velocity v and charge q (where q = ze, e =
1.6021 x 1019 C), entering into a uniform magnetic field B perpendicular to its
direction will move on a circular path of radius r (see Figure 24), determined by

the equation:

 

 

 

 

 

 

 

Figure 24. Precessing of an ion in a magnetic field.

55

the angular frequency a) (in rad/s) about the z-axis is defined by:

 

 

a) = = erf
r
qB zeB zeB
then: a): — = _ =2”f
m
m m A
f: eB
27H"
2
where: f = frequency (Hertz, Hz)

B = magnetic field (Tesla. T)

2 = multiples of elementary charge

lDa

 

m= mass in Da, (m = m ““9"

From the equations above, we can notice that: 1) the cyclotron angular

A Fourier Transform Ion Cyclotron Resonance Mass Spectometer is also

56

1.66054x10'27kg

)

frequency (w) is independent of the ion’s initial velocity, but a function of its mass,
charge, and the magnetic field, 2) the frequency is also independent of velocity,
but depends on the ratio (q/m)B, 3) the radius of the trajectory increases, for a
given ion, proportionally to the velocity, thus if the radius becomes larger than

that of the cell where the ion is confined, the ion is expelled [38, 66, 67, 109,

an “ion trap”. It consists of a cubic cell (the trap) inside a strong magnetic field.

The cell has three distinct sets of plates: trapping, transmitter, and
receiver plates. For mass analysis, a continuous beam of ions is formed outside
the cell and pulsed into it. When ions enter the FT-ICR analyzer cell, they are
constrained by the strong magnetic field such that they have a cyclotron motion;
ions are also constrained by electric potentials applied to a set of trapping plates

perpendicular to the magnetic field (see Figure 25) [76].

 

 

 

 

 

d
o—‘I p Detection

 

 

 

Excitation

 

 

 

 

Trapping

I

Figure 25. Schematic diagram of a FT-ICR mass analyzer (adapted from [110]).

 

 

 

Thermal energy produces a circular micromotion of the ions in the xy-
plane as they enter to the magnetic field; however, the radii of these orbits are
too small to be detectable. Ions of each mass have their characteristic
cyclotronic frequency. It can be demonstrated that ions excited by AC irradiation

at their own frequency and with the same energy [the same potential (Vp-p/d),

57

applied over the same time, Tm] will have an orbit with the same radius, in other

VT

words, the orbital radius is independent of W2: 1' = ——g‘c’n;‘°

 

__l

 

e
3
IL
ll
0

 

 

 

 

__l

Excitation ON OFF

Detection OFF ON

 

 

 

Figure 26. Sequence of excitation (left) and detection in FT-lCR-MS. The
ionic micromotion is indicated by the small circles of ions. (Adapted from
[66, 110])

In summary, all ions present in the cell are accelerated with an rf sweep
(commonly called a chirp) that covers a range of frequencies corresponding to
the m/z values we wish to observe; after excitation, ions will have the same radial
trajectories, but at frequencies dependent on their m/z ratio provided that the

potential is the same at each frequency [37, 67].

As the excited ions pass near the detector plates, the frequency of their
passage is detected as an induced current in the plates; it is called “image”

current. This transient signal (known also as “time domain” data or FID, from

58

“free induction decay") is a composite of the cyclotron frequencies of all the ions
present in the cell. The signal slowly dies off with time as the ions relax and
return to their stable circular orbits in the center of the analyzer cell, while it is
digitized and stored in a computer. The Fourier Transform algorithm is used to
extract the frequency and amplitude for each component of the composite signal.
The frequency of the motion of an ion is inversely proportional to its mass. This
frequency plot is then converted to an m/z plot and displayed as the mass
spectrum (see Figure 27). The SIN ratio is enhanced by averaging many cycles
before transforming and storing data. Once the ions have been detected, a
radiofrequency pulse (a quench pulse) is applied to the cell to eject ions before

the next bundle of ions are collected and trapped into the cell [39, 76].

FT-lCR-MS is a very powerful tool for analytical proteomics because it can
be coupled to an ESI source achieving spectacular resolving power (resolution

between two adjacent peaks in a mass spectrum).

Transient
,. (time domain spectrum)

 

 

'— m-C_ .... .... . *Wimw —— m

l Fourier transform

“Mass spectrum" in Hz

 

 

 

 

 

 

l gonna.” to m,, Figure 27. Processing of data from an ICR
and calibration analyzer using a Fourier transform
.. algorithm, converting time-domain data into
-. frequency domain and then calibrating in
_ l l "Ma" shown!!!” In W: terms of m/z values (adapted from [109].

 

 

 

 

59

3.4 Some examples of instrumentation

3.4a Tandem Mass Spectrometry

Tandem mass spectrometry or MS/MS is a technique in which ion
formation/fragmentation and subsequent decomposition of the original ions is
carried out “in tandem”; it involves two stages of mass analysis separated by a

reaction or fragmentation step.

There are two different approaches to MS/MS experiments:

1) Tandem-in-time: Instruments are, in general, ion-trapping mass
spectrometers such as two-dimensional or three-dimensional quadrupole ion
traps (QIT, LIT) and FT-ICR mass spectrometers. In this case, the various
stages of mass spectrometry are conducted within the same physical trapping

volume, but at different times during the experiment [98].

2) Tandem-in space: Instruments for these experiments have two (in

theory, it could be more than two) mass spectrometers in physically different

locations in the instrument.

60

3.4b Types of MSIMS experiments

There are four types of MS/MS experiments as described next; conceptual

diagrams are also shown for each experiment.

a) Product-ion analysis: This technique involves

 

 

 

Mass analyzer 1
a

 

 

 

 

 

Selected mlz

Collision cell

->

 

 

 

 

 

CID

Mass analyzer 2

/'

 

 

Scanned

 

 

the isolation of a
precursor ion, the
collisional activated

dissociation of the

precursor ion and the production of a mass spectrum of the ions resulting from

this induced fragmentation.

b) Precursor—ion analysis: All ions formed by a primary ionization pass into

 

 

Mass analyzer 1

 

 

 

 

 

 

Scanned

 

Collision cell
->

 

 

 

Mass analyzer 2
fl

 

 

 

CID

Selected mlz

 

 

a collision cell, one
m/z value at a time.

The second mass

analyzer is set to allow only a single m/z value to reach the detector.

c) Common-neutraI-loss analysis: All of the precursor ions formed by a

 

 

 

 

 

 

 

 

Scanned at
M I a

 

Collision
-§

  

 

 

 

CID (looking for
loss of “x" Da)

Mass war 2

 

 

Scannedat
m/z=a-x

 

 

primary ionization
are allowed to
pass, one m/z

value at a time, into

the collision cell. The second mass analyzer scans at the same value of the first

mass analyzer, but with a constant m/z difference [39].

d) Selected Reaction Monitoring (SRM, sometimes called Multiple

Reaction Monitoring, MRM): Ions of specific m/z values are allowed to pass to

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mass analyzer 1 Collision cell Mass analyzer 2 the collision cell,
——> _, - - ->

—> + _ , > the second mass
Selected m/z CID Selected m/z analyzer is set to
precursor Ions product Ions

 

 

allow only product ions to pass to the detector [39, 98]. This operation mode

increases sensitivity and is a good choice for quantitative experiments.

3.4c Tandem-in-time instruments

As it was stated in 3.4a, instruments are, in general, ion-trapping mass
spectrometers such as two-dimensional or three-dimensional quadrupole ion

traps (QIT, LIT) and FT-ICR mass spectrometers.

The approach for a MS/MS experiment consists of trapping of the ions
coming from the ionization source, isolation of the precursor ions of interest by
ejecting unwanted ions, fragmenting the precursor ions and performing a final
analysis step. In principle, these devices are capable of an unlimited number (n)

of repetitive cycles of isolation-fragmentation-isolation of ions before the final

62

scanning step, thus it might be possible to perform MS/MS/MS....MS

experiments (MS").

 

 

I1- 1) Ionization and Ion trapping +I I1- 2) Selection of desired Ions -DI
(precursor Ions)

 

H-3) Stabilization of precursor Ions-H I1- 4) Collisional dissociation ->l

      
 

- Precursor Ion
I

 

I" 5) Sequential mass scanning and detection -DI

 

 

 

Figure 28. Representation of a MS/MS (MSZ) experiment in a QIT. Idea and QIT drawings from
reference [95].

Representation of the MS/MS experiment in Figure 28 can be

conceptually extended to a Linear Ion Trap or a FT-ICR.

63

3.4d Tandem-In-space instruments

Examples of tandem-in-space instruments more commonly used in the

analysis of biomolecules are:

Triple quadrupole (QqQ): In the case of the triple quadropole instrument,
the collison cell (q) can be a rf—only quadrupole or a hexapole, octapole or a ring

guide, however, all these instruments are referred as QqQ.

A very common mode of operation is the Multiple Reaction Monitoring

(MRM).

 

    

Sample Inlet E
Sampling Cone _
—~
‘ a W mm
% Quadrupole1
Collision Cell
KI .

   
 

4

l

 

I

  
 

Data System

 

Quadrupole 2
Detector

  

 

 

 

 

 

Figure 29. Triple quadrupole scheme for the Quattro LC-MS from Micromass UK
Limited [111].

64

Quadrupole-time-of-flight (QqTOF): This instrument can be thought in

terms of replacement of the second mass-selective quadrupole of a QqQ

platform with a time-of-flight spectrometer. Using a TOF for the final stage of the

mass analysis, the instrument provides high resolution, good mass assignment

accuracy, high sensitivity, and the capability to record a complete mass spectrum

for each pulse of ions injected into the device [98].

 

Lock mass

 
 
 
  

Lockspray Lock mass Reference spray
baffle Reference spray
T-wave® Tranfer
ion guide Lens optics

Resolving
quadrupole

  

   
 
 
 
  

collision cell I

Analyte sp ' ' £-

Isolatlon valve
and removable '
sample cone

Oil-free
Scroll pump Air-cooled turbomolecular . <-
pumps
Isolation valve
and removable
sample cone Reflectron

 

     
 

 

o-TOF

 

Figure 30. Scheme of the commercial Q-TOF Premier from MicromassNVaters, with an

orthogonal TOF analyzer [112].

One variation of a QqTOF instrument is that obtained by substituting the

Qq section with a Quadrupole Ion Trap, (QIT-TOF). The use of a linear ion trap

instead of a QIT in this assembly has been also proposed (QqLIT); this last setup

is interesting because this instrument retains the classical triple quadrupole scan

functions such as selected reaction monitoring, product ion, neutral loss, and

65

 

precursor ion scanning while also providing access to sensitive ion trap

experiments [1 13].

Linear Traps can be used to improve the performance of FTICR-MS
systems. Unwanted ions that can cause space charge problems in the FT-ICR-
MS can be ejected from the linear ion trap to improve the resolution, sensitivity,

and dynamic range of the system [109].

 

 

Linear Ion Trap

alpha

 

 

 

 

 

 

 

 

';,_ A“. " ‘ ‘ 1 .» ‘1‘ "I F '1 r ‘. . ,‘
‘- " : . "v-T 4“: ‘ 1&5”! $9”:
'1. , .. ' ' 95 >4 . " . 1‘ ' "‘5.“ i“
m . 3" t ‘3 . ~ . ‘ .
Mill Q fill mace-1M ...-
\ “I > ,. ?e- ‘.‘. “A “J" A u 75" 4
.q, ‘ '15 ”MM .‘t‘fiu U
600 m’lhr l5IJs 300 Us 400 Us 210m 210L/s ‘ 7t“ .; 3 = '1: “I .
. ‘ ‘ x ”Q ' )‘

 

 

 

 

Figure 31. Scheme of the commercial LTQ-FTMS from ThermoFinnigan (It is a LlT-ICR—FT mass
spectrometer) [114].

3.4e Other hybrid instruments

Thus far this introduction has described the most common instrumentation
that is used in bioanalytical mass spectrometry at this time. However, it is
possible to find reports about other possible couplings for MS/MS techniques
such as TOF/T OF instruments or others using magnetic (B) and electric sectors

(E) in combination with quadrupoles, ion traps or TOF mass spectrometers. In all

66

cases, the driving force to new designs is the desire to obtain powerful tandem
mass instruments with higher resolving power, efficiency, sensitivity, and mass

accuracy [67, 98].

During the last few years, new types of mass analyzers have been
described, for example, the rectilinear ion trap and the orbitrap invented by
Makarov. Recently, a commercial LIT-orbitrap has become available but at this
moment it is too early to evaluate the benefits of these new mass analyzers [115-

117].

3.4f Applications of MSIMS

The numerous tandem mass spectrometer techniques available have
increased the power of mass spectrometry. In general, we can distinguish three
types of applications: 1) structure elucidation, 2) selective detection of target
compounds, and 3) ion-molecule reactions. In proteomics, genomics, molecular
biology, natural products and related fields, MS/MS techniques are becoming
more common with the appearance of relatively inexpensive instruments. The
understanding of ionization methods, mass analyzers, and detectors is essential

in order to take advantage of the capabilities that mass spectrometry offers.

67

4.- PROTEIN SEQUENCING AND DISULFIDE MASS MAPPING

4.1 Introduction

Even though over the decades, the purpose of peptide sequencing has
changed, the main goal has remained intact: the understanding of the function of
a protein in a living system. Originally, the aim of peptide sequencing was to
determine only the sequence of a protein, wholly or in part, in order to be able to
know its structure and by consequence, if possible, its function. As a
consequence of the advances in the field of genomics, we approach the time
when all genes, and therefore, all protein sequences, will be known. However,
knowledge of the sequence of a gene does not give us a complete information on
matters that are significant to the function of a protein such as inter- and

intramolecular disulfide bonds.

Conventional peptide mapping can be used to determine the entire
primary structure of a protein. This method is used routinely to compare the
structure of a product to that of a reference material or to those of previous lots to
confirm primary structures. In the standard approach, the protein is cleaved at
specific sites by enzymes (most notably trypsin) or chemical reagents. The
resultant peptides can be easily separated by high performance liquid
chromatography (HPLC), two-dimensional thin layer chromatography, or gel

electrophoresis. A map or “fingerprint” is obtained, which allows for

68

differentiation between proteins of similar but not identical, primary structure
[118]. If necessary, Edman degradation can be used for the purpose of peptide
sequencing of the peptidic fragments. Figure 32 shows an example of peptide

mapping using trypsin for the enzymatic cleavage.

 

MethionyI-Human Growth Hormone

.I ILLLLIIJ .. -...
I

 

 

—-§
‘1
I

 

 

Human Growth Hormone

 

 

r ‘I f V I I I T T I j' T‘V Y T V ‘r r Y I W V V I rvrf Y Y I Tj' j Y T I‘V’ Y fifi V T Yj ‘ V 1 f I
0.0 50.0 100.0
Minutes

 

 

 

Figure 32. Tryptic map chromatogram of methionyI-human growth hormone and human growth
hormone showing the ability of the tryptic map to differentiate the two materials. The N-terrninal
peptide differences are indicated by the arrows. Slight differences in retention times of peaks
eluted early in the chromatogram are attributed to problems with the HPLC-pumps. (From
reference [1 18]).

After the development of Fast Atom Bombardment, it was realized that
determination of the molecular weights of peptides generated from proteolytic or
chemical digestion of a protein could constitute an identification of the original
protein by comparing an experimentally-obtained array of masses with an array
of predicted or calculated masses from in silico digestion of various sequences

listed in a protein database; this technique was called Peptide Mass Mapping

69

(PMM or peptide mass fingerprinting, PMF). PMM is an extension of the mass
mapping technique, the produced peptide fragments, after the protein cleavage,
are identified according to their mass. The experimentally obtained masses are

compared with the theoretical peptide masses of proteins stored in databases by

means of mass search programs [36, 119].

 

 

Enzymatic or chemical cleavage l

$05.0:

Analysis of Peptide Mixture by MS

 

Relative Intensity
@
@©®

 

m/z

 

 

 

Figure 33. Conceptual diagram of protelytic mass mapping using mass
spectrometry. (adapted from reference [36])
Since its description in the late 805 and early 90$, PMM has been used to

determine the entire primary structure of proteins, and it has proved to be very

useful to provide additional information about posttranslational modifications such

as disulfide bonds [118, 120-122].

70

As we can see from Figure 33, the specific cleavage of a protein yields a
specific number of peptides of specific length, sequence, and most importantly of
specific mass that are used as inputs for computer algorithms for identification.
Currently, there are a number of software intemet—based tools available to
facilitate protein identification by peptide mass mapping such as Mascot
(h_ttp://www.matrixscience.com), Profound (Qttpzllprowl.rockefeller.ed_u)
ProteinProspector (http://prospector.gcsfedu), and extensive sequence data
bases such as Expasy/SwissProt Ms.expasy.ogq[) or NCBI
(mg/lwwwncbinlmnihgg). MSIMS experiments are also used as a tool for

the sequencing and identification of peptide fragments (see Figure 34) [6, 36,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

122]
: .
E .. :' . C i
: e " ° .. = AM.
= \ .. .. f .. '.-.’ .............
1DE\ 2DE \ / HPLC
Chemical or enzymatic reaction
Pepfides
MS v MSIMS
Peptide mass mapping Peptide fragments Edman sequencing
Peptide mass search MSIMS search

 

 

 

 

 

 

 

 

 

Figure 34. Separation, degradation, and identification techniques used to characterize
proteins (adapted from [122]).

71

Effective PMM is based on several assumptions; the first is that the
published genome is accurate. Frame-shifts can arise from incorrect sequencing
of the genome, specifically the inclusion of an extra base or the deletion of a
base in a gene’s sequence. Even with an accurate genome, any of several
protein modifications can introduce complications into PMM. Protein or peptide
modifications may not be predictable from the genome. These modifications can
occur naturally in an organism as a result of deliberate chemical derivatization or
via inadvertent modification during sample handling. Another common
assumption in peptide mass mapping experiments is that proteolytic or chemical
specificity is perfect. Many commonly used proteolytic enzymes and chemical
digestion protocols cleave proteins via specific reactions, however, some of
these procedures can have side reactions due to impurities (e.g., the presence of
chymotripsyn in the trypsin enzyme reagent) or strong chemical conditions (e.g.
deamidation at high pH values). Finally, a common assumption in PMM is that
all mass spectral features arise from peptides; however, for example, when using
MALDI experiments, analyte-matrix or analyte-alkali clusters could produce some

confusion during data analysis.

72

4.2 Disulfide mass mapping

4.2a Importance of disulfide bonds

Analysis of the three-dimensional structure of proteins is important for
understanding their biological activity and function. Among the 20 protein amino
acids, cysteine has unique properties. It can stabilize the three-dimensional
structure of a protein by linking the sulfhydryl groups of two cysteines together to
form a disulfide bond. Disulfide linkages can play an essential role in
stabilization of protein tertiary structures. Thus, changes in the disulfide bond
structure of a protein may result in important changes in activity and function. In
addition, the disulfide bond structure of a cystinyl protein may change during
protein folding. Knowledge of these structures is a prerequisite to a successful
manipulation of complex folding processes, for example, as in biotechnological

procedures (e.g., gene transfer) [123].

Disulfide bridges between cysteine residues are a key structural element
of many secreted proteins and peptides, being especially abundant in some
hormones, enzymes, plasma proteins, inhibitors, and venom proteins. Cysteine
can also contribute to protein biological functions because its free sulfhydryl
group can play a role in the active site for enzyme catalysis and in chelating
metal ions. With some smaller molecules, biological activity may depend strictly

on correct pairing of the cysteines. Analyzing the connectivities is thus an

73

important facet of protein structure determination, but it can take a large

investment of time and material. [124, 125]

The advent of Fast Atom Bombardment mass spectrometry during the
19803 made the task of locating disulfide bond more tractable [126]. In the
19903, new developments in mass spectrometry instrumentation revolutionized
protein chemistry; the introduction of electrospray ionization and matrix-assisted
laser desorption/ionization [34, 35] solved the problem of generating ions from
biomolecules and, therefore, the application of MS in the determination of

disulfide bonds has become almost routine.

4.2b Strategies for locating of disulfide bonds in cystinyl proteins

In principle, we can distinguish three different approaches that are

currently used for locating of disulfide bonds: enzymatic-based disulfide mass

mapping, partial reduction/alkylation-sequencing analysis, and partial

reduction/cyanylation CN-induced cleavage mass mapping.

Enzymatic-based disulfide mass mapping

The determination of the disulfide bonds in insulin and ribonuclease-A

established the basic strategy that still applies today for the determination of

disulfide bonds arrangements using the enzymatic-based approach. The steps

74

involved in this strategy are: (1) fragmentation of a non-reduced protein of
interest into disulfide-linked peptides with cleavages between all half-cystine
residues; ideally, the enzyme will cleave the protein at least once between the
cysteines as is illustrated in Figure 35, (2) separation of disulfide-linked peptides,
(3) location-identification by MS of disulfide-linked peptides, (4) Reduction of
disulfide-linked peptides, and (5) isolation-characterization of half-cystinyl

peptides using MS [17].

 

 

 

 

 

 

 

 

 

 

. 5 . 5
gmmmmm é“ newsman;
illllm §T|lllllm

 

Figure 35. Conceptual scheme illustrating the strategy for determination of disulfide bonds using
the enzymatic-based disulfide mass mapping approach.

In this approach, lack of cleavage between each cysteine could
complicate the analysis because it will not produce separate proteolytic peptides

for each of the cysteines. The closer the cysteines are to one another in the

75

sequence, the less likely it will be to find a useful proteolytic site; in the case
where cysteines are adjacent, this strategy is almost useless. Most proteolytic
digestions are carried out in alkaline media causing the possibility of scrambling”

or rearrangement (interchange) of native and non-native disulfides [127].

Partial reduction/alkylation-sequencing analysis

The original approach was proposed by Gray [124]. The strategy consists
of progressive reduction and alkylation of disulfide bonds followed by
characterization by sequence analysis. As the multiple disulfide bonds are
reduced by partial reduction, the nascent sulfhydryls are modified with different

alkylating reagents to chemically tag the product at different stages of reduction.

This analysis involves 4 key steps: 1) partial reduction in acidic conditions
2) separation of products by HPLC, 3) alkylation of free thiols, and 4) sequence
analysis to determine the location of tagged cysteines. Variants of this method
were developed in which up to three alkylating agents were used to label
different pairs of thiols, allowing in some cases, a full assignment in one
sequence analysis (see Figure 36). The incorporation of mass spectrometry has
been a straightforward improvement to this methodology. In the original protocol,
the use of the water-soluble reducing agent tris-(2 carboxyethyI)-phosphine

(T CEP) at pH = 3 was proposed in order to minimize the scrambling problem.

76

This approach is an advantageous alternative for small cysteine-rich
proteins in which only certain disulfides are accessible to the reducing agent.
Using this methodology after the reduction of a single disulfide bond, the 3D-
protein structure may change and there might be a better accessibility for the

reductant during subsequent reduction stages.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 “'A 1--A
2 Partial reductionIHPLC 2 Partial reduction/HPLC
3 ._. Alkylatlon (reagent A) Alkylatlon (Reagent B)
> 3 > 3 "'-'B

4

"”A 4--A
5 ._—

5 5
6 Ill—I

5 s-—e

1 ‘l-A

Partial reduction/HPLC 2 ‘

“WING" (Reagenti) 3 “a I Sequencing step (Edman degradation
4 --A or Mass Spectrometry)

5 --C
6 “-8

 

 

 

 

 

Figure 36. Conceptual scheme of the step-wise partial reduction/alkylation-sequencing
method to determine disulfide bond linkage.

This approach has been applied successfully with several modifications
according to the specific requirements for a particular protein [128-132]. The
most common used alkylation reagents work under slightly alkaline conditions,
thus, maybe the most important drawback of this methodology is a possible
disulfide exchange during the alkylation step. Nevertheless it seems that the use
of maleimides such as N-ethylamaleimide and 4-dimethylaminophenylazophenyl-

4'-maleimide) under acidic conditions, can overcome this problem [131-133]

77

Partial Reduction/Cyanylation CN-induced cleavage (PRCC) based

mass mapping methodology

During the late ‘905, Wu and Watson established a cyanylation (CN)-
based chemical method [125] for disulfide mass mapping to overcome many of
the shortcomings of the conventional proteolytic approach. The method is based
on five main steps: 1) the cystinyl protein is partially reduced, 2) the nascent
cysteines produced during reduction of a given disulfide linkage are cyanylated,
3) the polypeptide backbone is then cleaved on the N-terrninal side of the
modified cysteines, 4) any residual disulfide bonds are reduced to release the
cleavage products, and 5) the reduced cleavage reaction mixture is analyzed by
mass spectrometry (MS). Data interpretation leading to the connectivity of
cysteines in a disulfide bond relies on recognizing mass spectral peaks that
represent CN-induced cleavage products, and mapping these cleavage products
to the sequence of the cystinyl protein. During cleavage of the peptide bond on
the N-terminal side of a cyano(CN)-modified cysteine, the nucleophile is added to
the carbonyl carbon. In this way, the two nascent cysteines in a single-reduced
isoform provide the loci of cleavage sites, leading to production of three cleavage
fragments: the N-terminal peptide (A), the middle itz-peptide (B), and the C-
terminal itz-peptide (C), in which itz represents the 2-iminothiazolidine-4-carboxyl

group that is formed during the cleavage step (see Figure 37).

78

 

 

I
i610 $ (1512

i S

#49
ssss
Il
2630 snoszz

T S

1

SH

110

84 95

‘ ‘ 124
s

I

58”,?
1

Nu:
26 “MSCN 65

i S

l

Nu:
3535041

L? f 110
l

72
l

 

84

‘ 124
s
I

4 .
F 1° “z.

I s ‘Nu58

65 72 84 9s 5,
' L124

s|__s :5 INum

l

B

l

5) Analysis by Mass Spectrometry

C
I

 

 

109
\
Nu

57 mg
Nu 58

ii:
‘— 124

no
C

 

 

1) Partial Reduction of RN ase—A
Using tris(2-cerboxyethyl)phosphine
(T CEP) pH = 3

2) Cyanylation of one of four possible
single-reduced isoforms (IF4)

Using l-cyano-4-
dimethylsminopyridlnlum
tetrafluoroborste (CDAP) pH = 3

3) Cleavage
Nu: = nucleophile

4) Total Reduction
Using TCEP, pH - 3

 

Figure 37. Conceptual representation of the key steps in CN-based methodology for disulfide

mass mapping.

The cyanylation (CN)—based mass

of disulfide bond connectivities has proved useful even in proteins containing
adjacent cysteines [125, 134, 135], which is one of the greatest challenges in

bioanalytical chemistry. Some studies of protein folding are also possible using

this methodology [136].

an algorithm based on this CN-based m

79

mapping methodology for assignment

In a step to automate and facilitate data interpretation,

ethodology has been developed and

 

applied successfully [137, 138]. In the course of applying the CN-methodology,
occasional disappointing yields of certain cleavage products have been
observed, the cause of which was apparently related to some complex function
of amino acid composition and sequence in the vicinity of the cysteines [139].

The analysis of the cleavage reaction is the subject of chapter 2.

Miscellaneous methods

There are other methods that have been proposed for determining
disulfide connectivity; in general, they are only variations of the methods
described above with modifications to minimize the risk of protein scrambling or
with the goal of improving yields and detection of proteolytic or chemical-induced
fragments. Maybe the only novel and promising approach for determination of
disulfide linkage is by using electron capture dissociation [78, 140]; however, at
this time it is too early to know the impact of this technique in the determination of

disulfide bonds in cystinyl proteins.

80

5.- SUMMARY

Undoubtedly modern mass spectrometry combined with chemical
derivatization is a very powerful tool in the analysis of biomolecules and
currently, it is the most important analytical tool used in proteomics. In this
dissertation, the use of mass spectrometry in cyanylation-based disulfide mass
mapping give us an excellent example of the application of mass spectrometry in
biomedical science and in the determination of the disulfide proteome. This
methodology relies on the chemical derivatization and cleavage of the protein

prior analysis my mass spectrometry.

In Chapter 2 of this dissertation is presented a systematic study of the
nucleophile-mediated CN-induced cleavage reaction for the cyanylation-based
methodology using primary (methyl to butyl) amines as an alternative to
ammonia as the cleavage reagent. Reaction conditions are systematically
studied using high performance liquid chromatography to monitor the yield of the
expected cleavage products as identified by mass spectrometry. The focus of
this study was to optimize cleavage of cyanylated cystinyl proteins and to

improve the detection of the cleavage products.

In Chapter 3 is investigated the use of binary mixtures of homologous

nucleophiles to improve the confidence in distinguishing CN-induced cleavage

products from chemical by-products during analyses of cleavage reaction

81

mixtures by MALDI-MS. Analytical advantage can be gained in the use of a
binary mixture of homologous nucleophiles for cyanylation-induced cleavage of
partially reduced isoforms of cystinyl proteins. The combined use of ammonia
and methylamine was studied in order to find the adequate concentration ratio for
labeling of cleavage products in a single chemical step. The appearance of pairs
of mass spectral peaks facilitates recognition of diagnostic CN-induced cleavage
products, and increases confidence in results when samples present weak

signals or interfering chemical noise.

Chapter 4 presents a preliminary development of an on-line system for the
cyanylation-based methodology. The goal of this development was to carry out
all the chemistry on-Iine in order to avoid sample manipulation and, by
consequence, possible sample losses. The on-Iine system consists of a first trap
filled with a polymeric resin in which the protein is bound, partially reduced and
cyanylated. After these steps, the protein is eluted from the first trap and
transferred into a reactor for the cleavage reaction. Finally, the protein cleavage
mixture is trapped onto a second polymeric column where the residual disulfide
bonds are reduced and the remaining salts are removed from the sample. The
second trap containing the CN-induced cleavage fragments can be connected to
3 HPLC or a LC-MS system. Cleavage fragments also can be eluted to be

analyzed by MALDI-TOF-MS.

82

In Chapter 5 of this dissertation are presented conclusions and future work

is sketched.

83

6.- REFERENCES

[1] T. E. Creighton, Proteins. Structure and Molecular Principles, W. H. Freeman,
New York, NY., 1993.

[2] M. Trudy, M. R. James, Biochemistry, McGraw-Hill, Beijing, People's Republic
of China., 1999.

[3] C. Cambillau, Proteins - reflecting the diversity of life, Curr. Opin. Struct. Biol.
13 (2003) 655-657.

[4] A. L. Fink, Natively unfolded proteins, Curr. Opin. Stmct. Biol. 15 (2005) 35-
41.

[5] V. N. Uversky, Natively unfolded proteins: a point where biology waits for
physics, Protein Sci. 11 (2002) 739-756.

[6] D. C. Liebler, Editor, Introduction to Proteomics: Tools for the New Biology,
Humana Press, Totowa, NJ., 2002.

[7] M. R. Wilkins, J. C. Sanchez, A. A. Gooley, R. D. Appel, I. Humphery-Smith,
D. F. Hochstrasser, K. L. Williams, Progress with proteome projects: why all
proteins expressed by a genome should be identified and how to do it,
Biotechnol. Genet. Eng. Rev. 13 (1996) 19-50.

[8] J. S. Andersen, M. Mann, Functional genomics by mass spectrometry, FEBS
Lett. 480 (2000) 25-31.

[9] P. R. Graves, T. A. Haystead, Molecular biologist’s guide to proteomics,
Microbiol. Mol. Biol. Rev. 66 (2002) 39-63;.

[10] K. Sperling, From proteomics to genomics, Electrophoresis 22 (2001) 2835-
2837.

[11] M. Wang, R. J. Lamers, H. A. Korthout, J. H. van Nesselrooij, R. F. Witkamp,
R. van der Heijden, P. J. Voshol, L. M. Havekes, R. Verpoorte, J. van der Greef,
Metabolomics in the context of systems biology: bridging traditional Chinese
medicine and molecular pharmacology, Phytother. Res. 19 (2005) 173-182.

[12] A. Pandey, M. Mann, Proteomics to study genes and genomes, Nature 405
(2000) 837-846.

[13] H. Yano, S. Kuroda, B. B. Buchanan, Disulfide proteome in the analysis of
protein function and structure, Proteomics 2 (2002) 1090-1096.

84

[14] A. Hiniker, J. C. Bardwell, Disulfide bond isomerization in prokaryotes,
Biochemistry 42 (2003) 1 179-1 185.

[15] R. L. Baldwin, Protein folding from 1961 to 1982, Nat. Struct. Biol. 6 (1999)
814-817.

[16] E. Welker, M. Narayan, W. J. Wedemeyer, H. A. Scheraga, Structural
determinants of oxidative folding in proteins, Proc. Natl. Acad. Sci. USA. 98
(2001) 2312-2316.

[17] J. J. Gorman, T. P. Wallis, J. J. Pitt, Protein disulfide bond determination by
mass spectrometry, Mass Spectrom. Rev. 21 (2002) 183-216.

[18] S. P. Gygi, B. Rist, S. A. Gerber, F. Turecek, M. H. Gelb, R. Aebersold,
Quantitative analysis of complex protein mixtures using isotope-coded affinity
tags, Nat. Biotechnol. 17 (1999) 994-999.

[19] P. L. Martelli, P. Fariselli, R. Casadio, Prediction of disulfide-bonded
cysteines in proteomes with a hidden neural network, Proteomics 4 (2004) 1665-
1671.

[20] J. P. Bingham, N. M. Broxton, B. G. Livett, J. G. Down, A. Jones, E. G.
Moczydlowski, Optimizing the connectivity in disulfide-rich peptides: alpha-
conotoxin Sll as a case study, Anal. Biochem. 338 (2005) 48-61.

[21] E. Fuller, B. R. Green, P. Catlin, O. Buczek, J. S. Nielsen, B. M. Olivera, G.
Bulaj, Oxidative folding of conotoxins sharing an identical disulfide bridging
framework. FEBS J. 272 (2005) 1727-1738.

[22] X. Chen, Y. Li, X. Tong, N. Zhang, G. Wu, Q. Zhang, H. Wu, Solution
structure of BmP08, a novel short-chain scorpion toxin from Buthus martensi
Karsch, Biochem. Biophys. Res. Commun. 330 (2005) 1116-1126.

[23] J. J. Calvete, C. Marcinkiewicz, D. Monleon, V. Esteve, B. Celda, P. Juarez,
L. Sanz, Snake venom disintegrins: evolution of structure and function, Toxicon
45 (2005) 1063-1074.

[24] Y. Xiao, S. Liang, Inhibition of neuronal tetrodotoxin-sensitive Na+ channels
by two spider toxins: hainantoxin-Ill and hainantoxin-IV, Eur. J. Phannacol. 477
(2003) 1-7.

[25] E. Welker, W. J. Wedemeyer, H. A. Scheraga, A role for intermolecular
disulfide bonds in prion diseases?, Proc. Natl. Acad. Sci. USA. 98 (2001) 4334-
4336.

[26] T. P. Wallis, J. J. Pitt, J. J. Gorman, Identification of disulfide-linked peptides
by isotope profiles produced by peptic digestion of proteins in 50% (18)O water,
Protein Sci. 10 (2001 ) 2251 -2271.

85

[27] J. J. Pitt, E. De Silva, J. J. Gorman, Determination of the disulfide bond
arrangement of Newcastle disease virus hemagglutinin neuraminidase.
Correlation with a beta-sheet propeller structural fold predicted for
paramyxoviridae attachment proteins, J. Biol. Chem. 275 (2000) 6469-6478.

[28] P. Eaton, M. E. Jones, E. McGregor, M. J. Dunn, N. Leeds, H. L. Byers, K.
Y. Leung, M. A. Ward, J. R. Pratt, M. J. Shattock, Reversible cysteine-targeted
oxidation of proteins during renal oxidative stress, J. Am. Soc. Nephrol. 14 (2003)
8290-296.

[29] H. P. Harding, Y. Zhang, H. Zeng, l. Novoa, P. D. Lu, M. Calfon, N. Sadri, C.
Yun, B. Popko, R. Paules, D. F. Stojdl, J. C. Bell, T. Hettmann, J. M. Leiden, D.
Ron, An integrated stress response regulates amino acid metabolism and
resistance to oxidative stress, Mol. Cell 11 (2003) 619-633.

[30] M. Xu, A. Arulandu, D. K. Struck, S. Swanson, J. C. Sacchettini, R. Young,
Disulfide isomerization after membrane release of its SAR domain activates P1
lysozyme, Science 307 (2005) 113-117.

[31] K. Lee, J. Lee, Y. Kim, D. Bae, K. Y. Kang, S. C. Yoon, D. Lim, Defining the
plant disulfide proteome, Electrophoresis 25 (2004) 532-541.

[32] E. de Pauw, Matrix selection for liquid secondary ion and fast atom
bombardment mass spectrometry, Methods Enzymol. 193 (1990) 201-214.

[33] J. Schiller, K. Arnold (2000) in Encyclopedia of Analytical Chemistry
(Meyers, R. A., Ed.), pp. 559-585, John Wiley & Sons Ltd, Chichester, NY.

[34] J. B. Fenn, M. Mann, C. K. Meng, S. F. Wong, C. M. Whitehouse,
Electrospray ionization for mass spectrometry of large biomolecules, Science
246 (1989) 64-71.

[35] M. Karas, F. Hillenkamp, Laser desorption ionization of proteins with
molecular masses exceeding 10,000 daltons, Anal. Chem. 60 (1988) 2299-2301.

[36] J. T. Watson, Introduction to Mass Spectrometry, Lippincott Williams &
Wilkins, Philadelphia, PA., 1997.

[37] IonSpec.com. FTMS tutorial part I,
http://www.ionspec.com/FTMSWo20Tutorial/tmial fs.h;rn, (accessed August 15
2005) ,

[38] A. S. Carr, S. R. Annan, Overview of Peptide and Protein Analysis by Mass
Spectrometry, John Wiley @ Sons, Inc., 1997.

[39] D. Sparkman 0., Mass Spectrometry Desk Reference, Global View
Publishing, Pittsburgh, PA., 2000.

86

[40] R. A. Zubarev, Reactions of polypeptide ions with electrons in the gas
phase, Mass Spectrom. Rev. 22 (2003) 57-77.

[41] W. E. Seifert, Jr., R. M. Caprioli, Fast atom bombardment mass
spectrometry, Methods Enzymol. 270 ( 1996) 453-486.

[42] F. Hillenkamp, U. Bahr, M. Karas, B. Spengler, Mechanisms of laser ion
formation for mass spectrometric analysis, Scanning Microscopy, Supplement 1
(1987) 33-39.

[43] R. Holm, M. Karas, H. Vogt, Polymer investigations with the laser
microprobe, Anal. Chem. 59 (1987) 371-373.

[44] M. Karas, D. Bachmann, U. Bahr, F. Hillenkamp, Matrix-assisted ultraviolet
laser desorption of non-volatile compounds, Int. J. Mass Spectrom. Ion
Processes 78 (1987) 53-68.

[45] Sigma-Aldrich (2001) MALDI-Mass Spectrometry in Analytix.

[46] A. Schmidt, M. Karas, T. Dulcks, Effect of different solution flow rates on
analyte ion signals in nano-ESI MS, or: when does ESI turn into nano-ESI?, J.
Am. Soc. Mass Spectrom. 14 (2003) 492-500.

[47] M. Karas, M. Gluckmann, J. Schafer, Ionization in matrix-assisted laser
desorption/ionization: singly charged molecular ions are the lucky survivors, J.
Mass Spectrom. 35 (2000) 1-12.

[48] R. Zenobi, R. Knochenmuss, Ion formation in MALDI mass spectrometry,
Mass Spectrom. Rev. 17 (1998) 337-366.

[49] R. Knochenmuss, A quantitative model of ultraviolet matrix-assisted laser
desorption/ionization, J. Mass Spectrom. 37 (2002) 867-877.

[50] R. Knochenmuss, A quantitative model of ultraviolet,matrix-assisted laser
desorption/ionization including analyte ion generation, Anal. Chem. 75 (2003)
2199-2207.

[51] M. Cadene, B. T. Chait, A robust, detergent-friendly method for mass
spectrometric analysis of integral membrane proteins, Anal. Chem. 72 (2000)
5655-5658.

[52] J. J. Thomas, 2. Sheri, J. E. Crowell, M. G. Finn, G. Siuzdak,
Desorption/ionization on silicon (DIOS): a diverse mass spectrometry platform for
protein characterization, Proc. Natl. Acad. Sci. USA. 98 (2001) 4932-4937.

[53] J. Wei, J. M. Buriak, G. Siuzdak, Desorption-ionization mass spectrometry
on porous silicon, Nature 399 (1999) 243-246.

87

[54] Y. Xu, M. L. Bruening, J. T. Watson, Non-specific, on-probe cleanup
methods for MALDI-MS samples, Mass Spectrom. Rev. 22 (2003) 429-440.

[55] Y. Xu, J. T. Watson, M. L. Bruening, Patterned monolayer/polymer films for
analysis of dilute or salt-contaminated protein samples by MALDI-MS, Anal.
Chem. 75 (2003) 185-190.

[56] Y. Xu, M. L. Bruening, J. T. Watson, Use of polymer-modified MALDI-MS
probes to improve analyses of protein digests and DNA, Anal. Chem. 76 (2004)
3106-31 11.

[57] M. Dole, R. L. Hines, L. L. Mack, R. C. Mobley, L. D. Ferguson, M. B. Alice,
Gas phase macroions, Macromolecules 1 (1968) 96-97.

[58] M. Dole, L. L. Mack, R. L. Hines, R. C. Mobley, L. D. Ferguson, M. B. Alice,
Molecular beams of macroions, J. Chem. Phys. 49 (1968) 2240-2249.

[59] M. Yamashita, J. B. Fenn, Electrospray ion source. Another variation on the
free-jet theme, J. Phys. Chem. 88 (1984) 4451-4459.

[60] M. Yamashita, J. B. Fenn, Negative ion production with the electrospray ion
source, J. Phys. Chem. 88 (1984) 4671-4675.

[61] M. Yamashita, J. B. Fenn, Application of electrospray mass spectrometry in
medicine and biochemistry, Koenchu-Iyo Masu Kenkyukai 9 (1984) 203-206.

[62] C. K. Meng, M. Mann, J. B. Fenn, Of protons or proteins, Zeitschrift fuer
Physik D: Atoms, Molecules and Clusters 10 (1988) 361-368.

[63] M. L. Aleksandrov, L. N. Gall, N. V. Krasnov, V. I. Nikolaev, V. A. Pavlenko,
V. A. Shkurov, G. I. Baram, M. A. Grachev, V. D. Knorre, Y. S. Kusner, Direct
coupling of a microcolumn liquid chromatograph and a mass spectrometer,
Bioorg. Khim. 10 (1984) 710-712.

[64] M. Wilm, M. Mann, Analytical properties of the nanoelectrospray ion source,
Anal. Chem. 68 (1996) 1-8.

[65] E. d. Hoffmann, J. J. Charette, V. Stroobant, Spectrométrie de masse,
Dunod Ed., Paris, 1999. ‘

[66] J. H. Gross, Mass spectrometry: a textbook, Springer, Berlin ; New York,
2004.

[67] E. d. Hoffmann, V. Stroobant, Mass spectrometry : principles and
applications, Wiley, Chichester; New York, 2001.

[68] J. B. Fenn, Electrospray wings for molecular elephants (Nobel lecture),
Angew. Chem. Int. Ed. Engl. 42 (2003) 3871-3894.

88

[69] L. Tang, P. Kebarle, Dependence of ion intensity in electrospray mass
spectrometry on the concentration of the analytes in the electrosprayed solution,
Anal. Chem. 65 (1993) 3654-3668.

[70] P. Kebarle, L. Tang, From ions in solution to ions in the gas phase - the
mechanism of electrospray mass spectrometry, Anal. Chem. 65 (1993) 972A-
986A.

[71] P. Kebarle, A brief overview of the present status of the mechanisms
involved in electrospray mass spectrometry, J. Mass Spectrom. 35 (2000) 804-
817.

[72] P. Kebarle, M. Peschke, On the mechanisms by which the charged droplets
produced by electrospray lead to gas phase ions, Anal. Chim. Acta 406 (2000)
11-35.

[73] J. V. lribame, B. A. Thomson, On the evaporation of small ions from charged
droplets, J. Chem. Phys. 64 (1976) 2287-2294.

[74] M. R. Emmett, R. M. Caprioli, Micro-electrospray mass spectrometry: Ultra-
high-sensitivity analysis of peptides and proteins, J. Am. Soc. Mass Spectrom. 5
(1994) 605-613.

[75] M. S. Wilm, M. Mann, Electrospray and Taylor-Cone theory, Dole's beam of
macromolecules at last?, Int. J. Mass Spectrom. Ion Processes 136 (1994) 167-
180.

[76] R. C. Willoughby, E. Sheehan, S. Mitrovich, A global view of LC/MS: how to
solve your most challenging analytical problems, Global View Publ., Pittsburgh,
Pa., 2002.

[77] R. A. Zubarev, N. L. Kelleher, F. W. McLafferty, Electron Capture
Dissociation of Multiply Charged Protein Cations. A Nonergodic Process, J. Am.
Chem. Soc. 120 (1998) 3265-3266.

[78] R. A. Zubarev, Electron-capture dissociation tandem mass spectrometry,
Curr. Opin. Biotechnol. 15 (2004) 12-16.

[79] F. Turecek, N-Ca Bond Dissociation Energies and Kinetics in Amide and
Peptide Radicals. Is the Dissociation a Non-ergodic Process?, J. Am. Chem.
Soc. 125 (2003) 5954-5963.

[80] E. A. Syrstad, F. Turecek, Toward a general mechanism of electron capture
dissociation, J. Am. Soc. Mass Spectrom. 16 (2005) 208-224.

[81] D. M. Horn, R. A. Zubarev, F. W. McLafferty, Automated de novo
sequencing of proteins by tandem high-resolution mass spectrometry, Proc. Natl.
Acad. Sci. USA. 97 (2000) 10313-10317.

89

[82] K. Breuker, H. Oh, C. Lin, B. K. Carpenter, F. W. McLafferty, Nonergodic and
conformational control of the electron capture dissociation of protein cations,
Proc. Natl. Acad. Sci. USA. 101 (2004) 14011-14016.

[83] H. J. Cooper, K. Hakansson, A. G. Marshall, The role of electron capture
dissociation in biomolecular analysis, Mass Spectrom. Rev. 24 (2005) 201-222.

[84] D. M. Horn, Y. Ge, F. W. McLafferty, Activated ion electron capture
dissociation for mass spectral sequencing of larger (42 kDa) proteins, Anal.
Chem. 72 (2000) 4778-4784.

[85] S. K. Sze, Y. Ge, H. Oh, F. W. McLafferty, Top-down mass spectrometry of a
29-kDa protein for characterization of any posttranslational modification to within
one residue, Proc. Natl. Acad. Sci. USA. 99 (2002) 1774-1779.

[86] Y. Ge, B. G. Lawhorn, M. EINaggar, E. Strauss, J. H. Park, T. P. Begley, F.
W. McLafferty, Top down characterization of larger proteins (45 kDa) by electron
capture dissociation mass spectrometry, J. Am. Chem. Soc. 124 (2002) 672-678.

[87] R. A. Zubarev, N. A. Kruger, E. K. Fridriksson, M. A. Lewis, D. M. Horn, B. K.
Carpenter, F. W. McLafferty, Electron Capture Dissociation of Gaseous Multiply-
Charged Proteins Is Favored at Disulfide Bonds and Other Sites of High
Hydrogen Atom Affinity, Journal of the American Chemical Society 121 (1999)
2857-2862.

[88] K. R. Jennings, G. G. Dolnikowski, Mass analyzers, Methods Enzymol. 193
(1990) 37-61.

[89] R. M. Smith, K. L. Busch, Understanding mass spectra: a basic approach,
Wiley, New York, NY., 1999.

[90] C. Steel, M. Henchman, Understanding the quadrupole mass filter through
computer simulation, J. Chem. Educ. 75 (1998) 1049-1054.

[91] J. J. Leary, R. L. Schmidt, Quadrupole mass spectrometers: an intuitive look
at the math, J. Chem. Educ. 73 (1996) 1142-1145.

[92] P. E. Miller, M. B. Denton, The quadrupole mass filter: basic operating
concepts, J. Chem. Educ. 63 (1986) 617-622.

[93] J. Pappas Quadrupole Ion-Trap Mass Analyzer, Thenno Electron, Somerset,
NJ”(2004)

[94] M. W. Forbes, M. Sharifi, T. Croley, Z. Lausevic, R. E. March, Simulation of
ion trajectories in a quadrupole ion trap: a comparison of three simulation
programs, J. Mass Spectrom. 34 (1999) 1219-1239.

[95] Ion trap demo, Thermo-Finnigan, 2002.

90

[96] D. J. Douglas, A. J. Frank, D. Mao, Linear ion traps in mass spectrometry,
Mass Spectrom. Rev. 24 (2005) 1-29.

[97] J. W. Hager, A new linear ion trap mass spectrometer, Rapid Commun.
Mass Spectrom. 16 (2002) 512-526.

[98] J. W. Hager, Recent trends in mass spectrometer development, Anal.
Bioanal. Chem. 378 (2004) 845-850.

[99] J. C. Schwartz, M. W. Senko, J. E. Syka, A two-dimensional quadrupole ion
trap mass spectrometer, J. Am. Soc. Mass Spectrom. 13 (2002) 659-669.

[100] M. Sudakov, D. J. Douglas, Linear quadrupoles with added octopole fields,
Rapid Commun. Mass Spectrom. 17 (2003) 2290-2294.

[101] V. Mayya, K. Rezaul, Y. S. Cong, D. Han, Systematic comparison of a two-
dimensional ion trap and a three-dimensional ion trap mass spectrometer in
proteomics, Mol. Cell. Proteomics 4 (2005) 214-223.

[102] W. C. Wiley, l. H. McLaren, Time-of-flight mass spectrometer with improved
resolution, Rev. Sci. Instr. 26 (1955) 1150-1157.

[103] M. Karas, \"Time-of-flight mass spectrometer with improved resolution,\" W.
C. Wiley and I. H. McLaren, Rev. Sci. lnstrum., 26, 1150 (1955), J. Mass
Spectrom. 32 (1997) 1-3.

[104] M. Guilhaus, V. Mlynski, D. Selby, Perfect timing: time-of-flight mass
spectrometry, Rapid Commun. Mass Spectrom. 11 (1997) 951-962.

[105] D. C. Muddiman, R. Bakhtiar, S. A. Hofstadler, R. D. Smith, Matrix-assisted
laser desorption/ionization mass spectrometry: instrumentation and applications,
J. Chem. Educ. 74 (1997) 1288-1292.

[106] R. J. Cotter, The new time-of-flight mass spectrometry, Anal. Chem. 71
(1999) 445A-451 A.

[107] B. A. Mamyrin, Laser assisted reflectron time-of-flight mass spectrometry,
Int. J. Mass Spectrom. Ion Processes 131 (1994) 1-19.

[108] J. H. J. Dawson, M. Guilhaus, Orthogonal-acceleration time-of-flight mass
spectrometer, Rapid Commun. Mass Spectrom. 3 (1989) 1 55-1 59.

[109] M. P. Barrow, W. l. Burkitt, P. J. Derrick, Principles of Fourier transform ion
cyclotron resonance mass spectrometry and its application in structural biology,
Analyst 130 (2005) 18-28.

91

[110] A. G. Marshall, C. L. Hendrickson, G. S. Jackson, Fourier transform ion
cyclotron resonance mass spectrometry: a primer, Mass Spectrom. Rev. 17
(1998)1-35.

[111] MicroMass UK Limited. Quattro LC User's Guide,
http://www.micromass.co.l&, (accessed August 2005).

[112] W. MicroMass UK Limited. Q-Tof premier mass spectrometer (brochure),
http://wwwwaterscom, (accessed August 2005).

[113] G. Hopfgartner, E. Varesio, V. Tschappat, C. Grivet, E. Bourgogne, L. A.
Leuthold, Triple quadrupole linear ion trap mass spectrometer for the analysis of
small molecules and macromolecules, J. Mass Spectrom. 39 (2004) 845-855.

[114] Thermo-Finnigan. LTQ-FTMS, (accessed August 2005).

[115] Q. Hu, R. J. Noll, H. Li, A. Makarov. M. Hardman, R. Graham Cooks, The
Orbitrap: a new mass spectrometer, J. Mass Spectrom. 40 (2005) 430-443.

[116] M. Hardman, A. A. Makarov. Interfacing the orbitrap mass analyzer to an
electrospray ion source, Anal. Chem. 75 (2003) 1699-1705. -

[117] Z. Ouyang, G. Wu, Y. Song, H. Li, W. R. Plass, R. G. Cooks, Rectilinear ion
trap: concepts, calculations, and analytical performance of a new mass analyzer,
Anal. Chem. 76 (2004) 4595-4605.

[118] R. L. Gamick, N. J. Solli, P. A. Papa, The role of quality control in
biotechnology: an analytical perspective, Anal. Chem. 60 (1988) 2546-2557.

[119] Bioinfomiatics Web. Genomic glossary,
mpzllwwgeocities.comgiioinformgticsweb/Mex.html, (accessed August

2005).

[120] W. J. Henzel, T. M. Billeci, J. T. Stults, S. C. Wong, C. Grimley, C.
Watanabe, Identifying proteins from two-dimensional gels by molecular mass
searching of peptide fragments in protein sequence databases, Proc. Natl. Acad.
Sci. USA. 90 (1993) 5011-5015.

[121] B. W. Gibson, K. Biemann, Strategy for the mass spectrometric verification
and correction of the primary structures of proteins deduced from their DNA
sequences, Proc. Natl. Acad. Sci. USA. 81 (1984) 1956-1960.

[122] B. Thiede, W. Hohenwarter, A. Krah, J. Mattow, M. Schmid, F. Schmidt, P.
R. Jungblut, Peptide mass fingerprinting, Methods 35 (2005) 237-247.

92

[123] S. Barbirz, H. P. Happersberger, M. Przybylski, M. O. Glocker, Selective
cyanylation of cysteinyl residues as an approach for the mass spectrometric
determination of protein structures, European Mass Spectrometry 5 (1999) 123-
131.

[124] C. N. Pace, F. Vajdos, L. Fee, G. Grimsley, T. Gray, How to measure and
predict the molar absorption coefficient of a protein, Protein Sci. 4 (1995) 2411-
2423.

[125] J. Wu, J. T. Watson, A novel methodology for assignment of disulfide bond
pairings in proteins, Protein Sci. 6 (1997) 391-398.

[126] D. L. Smith, Z. R. Zhou, Strategies for locating disulfide bonds in proteins,
Methods Enzymol. 193 (1990) 374-389.

[127] J. Wu, J. Qi, J. T. Watson, Encyclopedia of mass spectrometry, in press.

[128] J. E. Hale, J. P. Butler, V. Gelfanova, J. S. You, M. D. Knierman, A
simplified procedure for the reduction and alkylation of cysteine residues in
proteins prior to proteolytic digestion and mass spectral analysis, Anal. Biochem.
333 (2004) 174-181.

[129] V. Schnaible, S. Wefing, A. Bucker, S. Wolf-Kummeth, D. Hoffmann, Partial
reduction and two-step modification of proteins for identification of disulfide
bonds, Anal. Chem. 74 (2002) 2386-2393.

[130] S. Sechi, B. T. Chait, Modification of cysteine residues by alkylation. A tool
in peptide mapping and protein identification, Anal. Chem. 70 (1998) 5150-5158.

[131] U. Goransson, A. M. Broussalis, P. Claeson, Expression of Viola cyclotides
by liquid chromatography-mass spectrometry and tandem mass spectrometry
sequencing of intercysteine loops after introduction of charges and cleavage
sites by aminoethylation, Anal. Biochem. 318 (2003) 107-117.

[132] U. Goransson, D. J. Craik, Disulfide mapping of the cyclotide kalata B1.
Chemical proof of the cystic cystine knot motif, J. Biol. Chem. 278 (2003) 48188-
48196.

[133] C. R. Borges, J. T. Watson, Recognition of cysteine-containing peptides
through prompt fragmentation of the 4-dimethylaminophenylazophenyl-4'-
maleimide derivative during analysis by MALDI-MS, Protein Sci. 12 (2003) 1567-
1572.

[134] J. Wu, J. T. Watson, Assignment of disulfide bonds in proteins by chemical
cleavage and peptide mapping by mass spectrometry, Methods in Molecular
Biology (Totowa, NJ, United States) 194 (2002) 1-22.

93

[135] J. Qi, J. Wu, G. A. Somkuti, J. T. Watson, Determination of the disulfide
structure of sillucin, a highly knotted, cysteine-rich peptide, by
cyanylation/cleavage mass mapping, Biochemistry 40 (2001) 4531-4538.

[136] J. Wu, Y. Yang, J. T. Watson, Trapping of intermediates during the
refolding of recombinant human epidermal growth factor (hEGF) by cyanylation,
and subsequent structural elucidation by mass spectrometry, Protein Sci. 7
(1998) 1017-1028.

[137] J. Qi, W. Wu, C. R. Borges, D. Hang, M. Rupp, E. Torng, J. T. Watson,
Automated data interpretation based on the concept of "negative signature mass"
for mass-mapping disulfide structures of cystinyl proteins, J. Am. Soc. Mass
Spectrom. 14 (2003) 1032-1038.

[138] C. R. Borges, J. Qi, W. Wu, E. Torng, A. P. Hinck, J. T. Watson, Algorithm-
assisted elucidation of disulfide structure: application of the negative signature
mass algorithm to mass-mapping the disulfide structure of the 12-cysteine
transforming growth factor beta type II receptor extracellular domain, Anal
Biochem 329 (2004) 91 -1 03.

[139] J. Wu, J. T. Watson, Optimization of the cleavage reaction for cyanylated
cysteinyl proteins for efficient and simplified mass mapping, Anal. Biochem. 258
(1998) 268-276.

[140] O. A. Mirgorodskaya, K. F. Haselmann, F. Kjeldsen, R. A. Zubarev, P.
Roepstorff, Towards the standard-module approach to disulfide-linked
polypeptide nanostructures. l. Methodological prerequisites and mass
spectrometric characterization of the test two-loop structure, Eur. J. Mass
Spectrom. 9 (2003) 139-148.

94

CHAPTER 2. OPTIMIZATION AND STUDY OF EXPERIMENTAL CONDITIONS
FOR NUCLEOPHILIC CLEAVAGE OF CYANYLATED CYSTINYL PROTEINS
BY PRIMARY AMINES IN DISULFIDE MASS MAPPING METHODOLOGY

1.- INTRODUCTION

1.1 Chemical cleavage of proteins

Proteins and peptides may be cleaved at various amino acid residues by
the use of endroproteolytic enzymes (enzymatic cleavage) or by the addition of a
chemical reagent (chemical cleavage). Since the early 19603 chemical cleavage
methods have been reported for cleaving proteins at specific amino acid
residues. Ideally, chemical cleavage methods should have high specificity with
minimal degradation of the polypeptide chain. If a particular amino acid residue
occurs infrequently and is well positioned along the chain, chemical cleavage at
this amino acid residue would generate a relatively small number of peptides of

different length that can be purified easily [1].

Proteins are biopolymers composed of 20 different amino acids. The

 

repeating sequence 0 o

H H H H H H
HzN— —c-[-N— —c]—,;Nl-l— —coor+
i i i

 

 

 

95

is held together by amide bonds. Proteins are digested, (hydrolyzed) by
specific mammalian enzymes such as trypsin or chymotrypsin; and by less
specific plant and animal enzymes, like papain and elastase; or by very
aggressive bacterial enzymes (pronase, subtilisin) that hydrolyze all peptide
bonds indiscriminately. The chemical cleavage of the peptide group generally
requires strong acid or base and elevated temperatures. However, there are
some chemical reagents that are able to specifically cleave a protein. For
example, cyanogen bromide (CNBr) is able to cleave at the methionine residue
with a yield of over 70%; this cleavage reaction is probably the most popular and

efficient for proteins [1, 2].

The selective chemical cleavage of the backbone of proteins or peptides
can occur at three different sites [3]: the amino-terminal side of the a-carbon

atom (p, in Figure 38), the peptide bond (q), or the carboxyl side of the a-carbon

(r).

 

O
R

_ // | NH—
fiuécnficgo

Figure 38. Cleavage sites for a polypeptide
chain.

 

 

 

96

In order to understand the principle of selective chemical cleavage, it is
first necessary to understand the general rules of general of hydrolysis of amide

bonds (q in Figure 38).

The amide group is a dipole with the negative charge residing on oxygen

and the positive charge localized partly on the carbonyl carbon and partly on the

. . 5- e
amide nitrogen: (O o
y: <—-> (I:
\
/6+\N— /a u—
H

To this dipole can be added the hydroxyl anion of a basic medium or the
proton of an acid. In both cases, the unstable tetrahedral intermediate breaks
down into a carboxylic acid and another amino-terminal residue. There is
generally no selectivity in the attack of the H‘“ or OH' on an amide dipole in a

protein chain (Figure 39).

 

Alkaline (f? + - (If) 0
on _
hydrolysis /C\ "' '> /(|3\":‘_——'> /I(I3H T H2N
5+ fi— OHH
q (I)H
O C H
OHI
Acid ' - /6‘N— I Go
C + H —> —> bx —>
hydrolysis /0\N— X K'OH—H /C<¢_.\._H
H / H
H
OH 'x
I ,OH on" o
c ,/ ’5 (I OH II + H2N—
/ \fi— H X -> /C<+ —> /C
H2

 

 

 

Figure 39. Hydrolysis of amides by acid or base (adapted from [4]).

97

In selective chemical cleavage, the negative end of the dipole (the
imidolate anion) seeks an electrophilic center within the molecule. Such an
electrophilic center can be created by a suitable leaving group (X). The
departure of the leaving group X need not be in separate discrete steps, but may
be concerted. The preferred mechanism is 1,5-interaction, which leads to a

thermodynamically favored five-membered ring systems (Figure 40) [4].

 

...
KL |/ Hill—'—
c

A\oe

e c//
'3‘» E \o
NIL c\/\

 

O—

H
N
H

 

 

 

Figure 40. Neighboring group effects associated with chemical
cleavage in a protein (adapted from [4]).

The most common site for selective chemical cleavage is the methionine
residue [2, 5]. The group in a methionyl peptide that is being removed is the
methyl thio-ether. In the first step, the cyano-sulfonium intermediate is formed by
interaction with the pseudohalogen cyanogen bromide, and this reaction
facilitates removal of sulfur group. The attack of the imidolate anion leads to the
unstable imino-y-lactone, which hydrolyzes to homoserine (lactone) and a new
amino-terminal residue. Cleavage of tryptophan, tyrosine, histidine, and reduced

phenylalanine follows a similar mechanism (see Figure 41).

98

 

 

R
a, 2, _
7NH 13.7 N 33'
H c H20,H+

HC /0 o

’ ‘3 R I? l l!
2 1\C/ Y \O

H2c\c/

H2

 

 

 

Figure 41. Mechanism of cleavage of methionine peptide bonds with cyanogen bromide (adapted
from [1] and [4]).

Other examples of peptide bond cleavage (q in Figure 38) are numerous.
In one of these mechanisms, the Net-amide nitrogen atom can interact with an
electrophile position generated at the 8-position on the side chain to form a labile
N-acyI—pyrrolidone, —oxazolidone, or -thiazolidine with subsequent cleavage of
the acyl-nitrogen function. Another example is the chemical cleavage of the
peptide bond on the N-terminal side of cyanocysteine residues (q in Figure 38)
under alkaline conditions using ammonia or hydroxide ions as nucleophiles (see

Figure 42) [6, 7].

99

 

N
if
s
\
0 CH2 N 8 3
( )v I 0 Sc/ \cn2
R1\ « /cn\ /R, )K V I
cu ) c —> "N + .. cu n.—>
.. H II cu H2N/ \c/
R: Nu o I Nu II
R; 0
HM s
o %c/ \im2
R1 + \
——> \cHJK u/cu\c/ R’
I Nu H II
R, o
peptide itz-peptide
itz = 2-Imlnothlazolldlne-4-carboxyl group

 

 

 

Figure 42. Chemical cleavage of cyanocysteine by means of a nuclephilic attack in alkaline
media [8, 9].
1.2 Chemical cleavage in the cyanylation-based disulfide mapping

methodology

It was explained in Chapter 1, that during the late “903, Wu and Watson
established a cyanylation (CN)-based chemical method [10] for disulfide mass
mapping to overcome many of the shortcomings of the conventional proteolytic
approach. This method is based on five main steps: 1) the cystinyl protein is
partially reduced, 2) the nascent cysteines produced during reduction of a given
disulfide linkage are cyanylated, 3) the polypeptide backbone is then cleaved on
the N-terminal side of the modified cysteines, 4) any residual disulfide bonds are

reduced to release the cleavage products, and 5) the reduced cleavage reaction

100

mixture is analyzed by mass spectrometry (MS). Data interpretation leading to
the connectivity of cysteines in a disulfide bond relies on recognizing mass
spectral peaks that represent CN-induced cleavage products, and mapping these

cleavage products to the sequence of the cystinyl protein (see Figure 43).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I l
1¥W§i512i425§m
S 58 S S 110
i r U I l
l I) Partial Reduction of RNase-A
Using tris(2-eerboxyethyl)phosplllne (TCEP) pH - 3)
26 10 H is 7.2 i4 25 5"”,
l
S 53 S S 110
I I U I l
l 2) Cyanylation of one of four possible
single-reduced isoforms (1!“ 4)
Nu: Nu;
Using l-cysno-4-dimethylsminopyrldlnium
2‘ 10318981 ‘35 P I,“ 3’5 “C N 124 tetrsf'luoroborete(CDAP) pH -3
‘ a. s 58 s s s 3 no
1 3) Cleavage
Nu: = nucleophile
,r:°a errr‘_
s ‘Nuss s s [s 5 Nil 110
1 | B44 |
l 4) Total Reduction
57 itz‘ 109 124 Using TCEP, pl-l - 3
1 ‘Nu ss ‘Nu no
A B C
5) Analysis by Mass Spectrometry

 

 

 

Figure 43. Conceptual representation of the procedure of cyanylation-based disulfide mass
mapping.

The cyanylation (CN)-based mass mapping methodology for assignment
of disulfide bond connectivities has proven useful in characterizing these linkages

even in proteins containing adjacent cysteines [10-12], which is one of the most

101

challenging analytical problems; studies relating to protein folding have also been

carried out using this methodology [13-15].

The essence of the cyanylation-based mass mapping method is based on
selective cyanylation of the sulfhyde group of a cysteine residue, and
subsequent nucleophilic attack to promote cyanylation (CN)-induced cleavage of
the peptide backbone on the N-terminal side of the modified cysteine as
illustrated in Figure 44 [6]. Mass analysis of the cleavage reaction mixture
confirms the location of the affected cysteine; e.g., in Figure 44, detection of the
fragments consisting of residues 1-9 and itz-10-20 (itz = iminothiazolidine-
blocked N-terminus of a peptide) verifies that a cysteine residue is located at
position 10 in the hypothetical 20-residue cystinyl protein. This characteristic
cleavage, in particular, has been useful in recognizing the position of free thiols in

a study of disulfide isomerization after membrane release in a protein [16].

 

 

 

 

 

 

N
1 N IN 1
. CN-induced
Cyanylation Cleavage 9 Nu
10 ’CHz-SH -——-> 10 - CHZ-SCN »
CDAP, P“ = 3 Nucleophile, Nu: 10 “Z
20 20
C C 2°C

 

 

Figure 44. Cleavage of a cyanylated cysteine-containing peptide yields two
fragments.

102

In the CN-based methodology, chemical cleavage of the peptide bond on
the N-terminal side of the cyanocysteine is carried out under alkaline conditions,
using ammonia or hydroxide ions as nucleophiles [6, 9, 17]. For this cleavage
reaction, Jacobson [6] established the first protocol and, Degani and Patchomik
[17] studied the pH dependence of the cleavage and B-elimination reactions
using di- and tri-peptides as model compounds. Stark [8] summarized the
methodology and reviewed some examples and applications. In all this early
work, OH' ion was the nucleophile used under mildly alkaline conditions (pH
range: 8-12) with reaction times ranging from 18 to 80 hours. Later, Nakagawa
et al [18], applied the CN-cleavage reaction to human parathyroid hormone
[hPTH(1-84), containing one cysteine residue at position 35] in order to produce
hPTH(1-34). They reported the formation of some by-products during the
cleavage step, such as deamidation of asparagine or aspartic acid and

methionine oxidation [7, 18, 19].

Wu and Watson [9, 10] systematically studied the rate and yield of the
cleavage reaction as a function of pH using OH' ion or ammonia as nucleophiles.
They studied the effects of peptide structure on the production of the desired
cleavage products and elimination byproducts, and also found that the presence
of proline (and sometimes tyrosine) on the N-temiinal side of the cyanylated
cysteine plays an important role in promoting the B-elimination side-reaction (see
Figure 45). However, a systematic study of the cleavage reaction in a more

complex protein has not been reported so far, and some contradictory reports of

103

using low temperature together with low or high concentrations of the nucleophile

have been proposed to maximize the yield of the CN-induced cleavage reaction

 

 

 

 

 

 

 

[7.19]-
,---. Nit-MW
awvm J" \i“‘- l_|
"Jiaiaissasfzusisnom
Nu S S ““““ 5..
. . I _______________
a” f" ~~~~~~~~~~~~~~~~~~~
s\
it i: .i
Rafi" l2,
° "W
.. MI r .
A“ "KM' R1 a rii" E/
R: o
W W
Peptide It" Mom (pawl-Fm)
N ' SH 9‘ 1 i112; l railwalzL 1
'"2'3" “364053657284“ “pH-95mm '26405365728495110
SH SH SH SH SH SH SH SH
A B C
moi-impede mumm- C-temirlllu-pqitlde
m mumzpelknraimin
CleavaoaproalctsafterTotaiRechnion bothcleavagesites

 

 

 

Figure 45. Chemical cleavage in the cyanylation-based disulfide mapping methodology.

In the course of applying the CN-methodology, we have observed
occasional disappointing yields of certain cleavage products, the cause of which
was apparently related to some complex function of amino acid composition and
sequence in the vicinity of the cysteines [9]. Interestingly, in spite of the failure to

detect all predicted cleavage products for a given cystinyl protein, we continued

104

to find good success with the outcome of our analyses. In the interim, we have
validated the robust nature of the CN-based methodology through computational
analysis of patterns of CN-induced cleavage products [14]. These revealed that
certain subsets of the whole array of mapping data are uniquely related to a
particular member of the large set of theoretically possible isomeric disulfide
structures for a given cystinyl protein. Thus, in principle, detection of only one
subset of cleavage products is required to recognize the connectivity of cysteines
in a given disulfide [20]. Nevertheless, we would prefer to achieve maximal yield

of cleavage products in all cases.

We undertook a systematic study of the nucleophile-mediated CN-induced
cleavage reaction using high performance liquid chromatography (HPLC) to
monitor the yield of the expected cleavage products as identified by mass
spectrometry. The focus of this study was to optimize cleavage of cyanylated
cystinyl proteins and to improve the detection of the resulting cleavage products.
Primary amines were hypothesized to serve as effective alternative promoters of
the cleavage reaction in cyanocysteinyl proteins. This chapter describes results
of our study on the feasibility of using a series of primary amines as nucleophiles:
methylamine, ethylamine, propylamine, and butylamine. Ammonia was the
nucleophile used as a control. The new protocol was applied to analysis of or-
lactalbumin from bovine milk type III as well as RNase-A. For a given
nucleophile, three main conditions were evaluated: concentration, temperature,

and reaction time. Analysis of cleavage reaction mixtures by HPLC produced

105

chromatograms that made it possible to visualize and analyze changes in the
yield of the cleavage reaction under different experimental conditions. Matrix-
assisted laser desorption/ionization (MALDI)-MS was used to identify the CN-

induced cleavage products.

2.- MATERIAL AND METHODS

2.1 Chemicals

Guanidine hydrochloride was obtained from Boehringer Mannheim
Biochemicals, (Indianapolis, IN); 1-cyano-4-dimethylaminopyridinium
tetrafluoroborate (CDAP) and tris-(2-carboxyethyl)phosphine (TCEP)
hydrochloride were purchased from Sigma Chemical Co. (St. Louis, MO).
Methylamine, ethylamine, and propylamine were acquired as hydrochloride salts
and butylamine was acquired as a liquid from Aldrich. Ammonia was obtained
from EM Science as an aqueous solution containing approximately 14.4 M
ammonium hydroxide. Bovine pancreatic ribonuclease A type Ill-A (RNase-A)
and a-lactalbumin from bovine milk Type III were purchased from Sigma

Chemical Co. (St. Louis, MO).

106

Water grade I [21] or better was obtained from a Millipore system, HPLC
grade acetonitrile was purchased from EMP chemicals Inc. (Gibbstown, NJ), and
spectrophotometric grade trifluoroacetic acid +99% (TFA) was obtained from

Aldrich (Milwaukee, WI).

2.2 Mass spectrometry

MALDI mass spectra were acquired on a Voyager DE-STR time-of-flight
(TOF) mass spectrometer (Perkin-Elmer Biosystems Inc.) equipped with a 337-
nm nitrogen laser. Measurements were made in linear mode with the
accelerating voltage set typically to 25,000 V with grid voltage at 95%, guide wire
at 0.05%, and extraction delay time at 150 nsec. Time-of-flight to mass
conversion was achieved by external and/or internal calibration using standards
of bradykinin (MH‘average = 1061.22 Da), bovine pancreatic insulin (MH"alvfiraga =
5734.56 Da), horse skeletal myoglobin (MH‘average = 16,952 Da), and/or
cytochrome c (MH‘average = 12,361.1 Da) obtained from Sigma (St. Louis, MO).
Samples were prepared by spotting 0.5 pL of analyte + 0.5 pL of matrix solution
onto a stainless steel sample plate with or-cyano-4-hydroxy cinnamic acid (from
Sigma, St. Louis, MO) as the matrix (10 mg/mL) in the modified thin-layer method
as described by Cadene and Chait [22]. For some experiments samples were

desalted using Zip-Tips from Millipore (Billerica, MA).

107

2.3 Chromatography

HPLC was performed using a Waters 2695 separation module coupled to
a Waters 2487 Dual 1. Absorbance detector at 214 nm. The column for
purification of isoforms contained a C4 stationaty phase (Vydac # 214TP54, 5-pm
particle size, 300 A pore, 4.6 x 250 mm) and for p-HPLC separation of reaction
mixtures was the column contained a C18 (Vydac # 218TP5115, 5-pm particle
size, 300-A pore, 1.0 x 150 mm). HPLC fractions were collected manually and, if

necessary, dried for further use.
2.4 Partial Reduction of RNase-A

The protein sample was placed in a vial where the partial reduction was
carried out by adding TCEP in 0.1 M citrate buffer at pH 3.0 for the cysteine
content and keeping a concentration of 4 nmol/pL (e.g., 40 nmol of TCEP was
_reacted with 10 nmol of RNase-A in a total volume of 10 pL ) followed by

incubation at room temperature for 15 min.
2.5 Cyanylation of single-reduced isoforms of RNase-A

A 20-fold molar excess of 0.1 M CDAP (in 0.1 M Citrate buffer at pH = 3.0)
over the total cysteine content was added to the mixture of partially reduced

isoforms for the cyanylation step at room temperature for 15 min. Single-reduced

108

and cyanylated species were purified using HPLC, and then dried. 6M guanidine
hydrochloride solution was used to reconstitute the sample in order to obtain
(depending on the experiment) solutions between 0.3 and 1 nmol/pL; an accurate
concentration was determined, if necessary, by injecting an aliquot of this

solution into a p-HPLC system.

2.6 Cleavage of single-reduced and cyanylated RNase-A isoforms

The nucleophile solution was made using the water grade I or better [21].
Taking pKa values into account for each reagent, the pH was modified, if
necessary, with NaOH, to reach a value of pH = pKa + 1 to assure predominance
of the basic form for the amine (> 90%). Cyanylated isoforms were prepared in 6
M guanidine and treated with a given nucleophile under specified experimental
conditions (concentration, temperature, and reaction time). In this report,
conventional conditions for the cleavage reaction means that the cyanylated

isoform was treated with 1 M ammonia for 60 min at room temperature [10].

2.7 Quenching of the cleavage reaction

After the cleavage reaction was allowed to proceed for the specified time,

it was quenched by adding citric acid to establish pH = 3, followed by adding a

0.1 M TCEP solution and incubating at 37°C to completely reduce the remaining

disulfide bonds.

109

2.8 Isolation of single-reducedlcyanylated isoforms of RNase-A by

means of high performance liquid chromatography

RNase-A (m/z for MH‘average = 13683.3) contains 124 amino acids among
which eight cysteines form four disulfide bonds: Cys 26-Cys 84, Cys 40-Cys95,
Cys 58-Cys 110, and Cys 65-Cys 72. After partial reduction and cyanylation,
single-reduced/cyanylated isoforms of RNase-A can be isolated using reverse
phase-HPLC. Because the cyanylated isoforms elute after the intact protein (IP),
we refer to them according to their elution order (IF1, IF2, IF3, IF4). The
chromatogram in Figure 46 shows the order and nomenclature, which was
established during previously reported cyanylation-based mass mapping using
MALDI-TOF-MS [10]. Table 2 lists the cysteines that were connected by a
disulfide bond in the intact protein, i.e., a disulfide that was reduced to form the

particular single-reduced and cyanylated isoform of RNase-A.

Table 2. Nomenclature for the single-reduced isoforms of RNase A.

 

 

 

 

 

 

 

 

Single reduced/cyanylated Reduced Linkage
isoform

IF 1 Cys 65 - Cys 72

IF3 Cys 26 — Cys 84

IF3 Cys 40 — Cys 95

IF4 Cys 58—Cys 110

 

110

 

1.25-

1.00-

0.75-

Absorbance (AU)

0.25-

 

 

 

 

Tlme (min)

 

 

 

Figure 46. HPLC chromatogram (see text for details) of RNase-A
(IP = intact protein) and its single-reduced and cyanylated isoforms
(IF1...IF4).

2.9 Partial reduction and cyanylation of a-Iactalbumin

a-Lactalbumin (m/z for MH‘average = 14,176) contains 123 amino acid
residues among which eight cysteines are linked by four disulfide bonds: Cys 6-
Cys 120, Cys 28-Cys 111, Cys 61-Cys 77, and Cys 73-Cys 91. Partial reduction
and cyanylation of the protein were carried out following a procedure similar to
that described for RNase-A. Single-reduced and cyanylated species (IF1L, IF2L,
IF3L, IF4L) were separated and collected using p-HPLC. These species were

processed under the optimized cleavage reaction conditions using ammonia,

methylamine, or ethylamine as nucleophiles.

111

3.- RESULTS AND DISCUSSION

3.1 Assessment of nucleophilic cleavage reactions at room

temperature

Initial screening experiments showed that yields of cleavage product did
not increase significantly over time (from few minutes up to 120 min were
evaluated); on the contrary, after 60 min, the yield of some cleavage products
decreased, suggesting the existence of secondary reactions. The cleavage
reaction for isoform IF4 was observed to be in greater competition with side
reactions (data not shown) than those for the other three isoforms. For this
reason, we chose IF4 (isolation was explained in section 2.8) as the model
isomer for optimizing the cleavage reaction conditions. It was also observed that
reaction times less than 30 minutes seemed to be a good choice to avoid
decreasing the yield of cleavage products; thus, a series of experiments varying

the reaction time was carried out.

At room temperature, IF4 was reacted with 1 M methylamine or 1 M
ammonia for different reaction times. Cleavage reaction mixtures were analyzed
by p—HPLC and fractions were collected and analyzed by MALDI-MS for
identification. From these chromatograms, it was observed that yields of
cleavage products were maximal after 10 minutes of reaction time. In Figure 47,

the chromatogram for the cleavage reaction mixture using the conventional

112

methodology (60 min) is compared with chromatograms of reaction mixtures
prepared with methylamine or ammonia for 10 minutes. Ideally, 1 nmol of
cyanylated isoform produces 1 nmol of each of the cleavage products, and this
result should be reflected in the chromatograms. Using a UV-Vis detector at 214
nm, the signal is proportional to the number of amino acid residues in the
cleavage products [23, 24]. Thus, ideally, the peaks for cleavage products A and
B should be more intense than that for C, which contains fewer residues (see
Figure 43 and values in Table 3); however, it is evident from the chromatograms
in Figure 47 that the opposite result was obtained, apparently reaction conditions

at room temperature might promote competing side reactions in some cases.

 

  
  
     
  

NH‘OH so min

 

' I ' I U I I I 1 I—r-fi
30 35 40 45 50 55 50

NH‘OH 10 min

 

Absorbance (AU)
0

CHSNHz 10 min

  

 

0.2
0.1
0.0 .
30 35 40 45 50 55 60
Time (min)

 

 

 

Figure 47. HPLC chromatograms of cleavage products from single-reduced and
cyanylated isoform 4 (IF4) of RNase-A treated at room temperature with: Top, 1
M ammonia for 60 min; middle, 1 M ammonia for 10 min; bottom, 1 M
methylamine for 10 min. Cleavage products A, B, and C are defined in Figure
43 and in Figure 45. It is also shown the retention time for the intact IF4.

113

Using methylamine, yields of cleavage products showed a substantial
improvement over results using the conventional methodology (e.g. observe the
area corresponding to the cleavage fragment B). However, there is still a
presence of chromatographic peaks at 42.5 min in Figure 47 corresponding to a
mixture of non-specific reaction products near the peak representing cleavage
product A. The use of ammonia for only 10 minutes (middle panel of Figure 47)
appears to be somewhat better than 60 minutes (top panel of Figure 47) even
though products of apparent side reactions still obscure peaks for some of the
expected cleavage products. Extended reaction times (beyond 20 minutes) with
methylamine resulted in substantial loss of cleavage products A and B, and also
some C according to HPLC chromatograms (not shown) compared to that in the

lower panel of Figure 47.

Table 3. Calculated m/z values for protonated cleavage products of IF4 produced using
ammonia or methylamine as nucleophile.

 

 

 

 

 

Cleavage product
fragment* Residues W2 W2
-NH2 -NHCH3
A 1-57 6351 .1 6365.1
B itz-58-1 09 5766.4 5780.4
C itz-110-124 1659.8 1659.8

 

 

 

 

 

 

MALDI-MS spectra of the cleavage reaction mixture were obtained for
samples treated with methylamine for 10 minutes or with ammonia for 60 minutes
as shown in Figure 48. As expected, the strength of the mass spectra for certain
cleavage products is consistent with the intensity of the corresponding HPLC

peaks in Figure 47. This is especially evident for the middle (B) and N-terminal

114

(A) cleavage products (see Table 3 for m/z values) produced using methylamine
as the nucleophile. From these results, we concluded that a 10-minute reaction
time instead of the previously accepted 60 minutes constituted an analytical
advantage. For the other three isoforms (IF1, IF2, and IF3), similar experiments
were carried out and, in general, MALDI-MS signals for the cleavage products

produced by methylamine were comparable to those produced by using

 

 

 

 

 

 

 

ammonia.
C
100 CM
M
... so
°\° so
>
’2 7° 3 A
60
s .. ll
3 40
g 30
A
3
n: 3° 3
10
o
NH‘OH
CH3NH2

 

 

 

 

Figure 48. MALDI mass spectra of the cleavage reaction mixture from
cyanylated IF4 of RNase-A, using ammonia (60 min) or methylamine (10 min)
as the nucleophile. A, B, C are the cleavage products according to Figure 43
and Table 3.

Ethylamine, propylamine, and butylamine were also evaluated for possible

improvement in the yield of cleavage products. Because of the hydrophobic

115

region often found near the cysteine residues, it was thought that the longer
hydrophobic alkyl chains of the amine molecules might promote better interaction
of the nucleophile with the cyanylated cysteinyl isoforms of RNase-A. However,
results from experiments with these longer alkyl amines did not show an
advantage over the use of ammonia or methylamine at room temperature for 10

minutes.

3.2 Optimization of experimental conditions for the cleavage reaction

using ammonia and methylamine as nucleophiles

3.2a Evaluation of concentration, pH and the use of microwaves at

room temperature

Given the experiments described previously, it was necessary to consider
if at room temperature there was the possibility for improving the cleavage
reaction in order to obtain better yields of cleavage products while minimizing
side-reaction products. Figure 49 summarizes results of a series of experiments
that were carried out using IF2 as substrate with ammonia or methylamine as the
nucleophiles (IF2 was used instead of IF4 to take advantage of available purified
isoforms). Thus, because cleavage methodologies have reported using 0.1 M
borate buffer at pH 9.0 for up to 18 hours of reaction time [6, 9, 17], a comparison
between these experimental conditions and similar conditions using methyl

amine or ammonia were carried out. For this comparison, the concentration of

116

the nucleophile basic form was 1 M at a pH close to the pKa value for the
nucleophile in an effort to avoid side reactions at higher pH values (e g.
deamidation and beta elimination reactions are more probable). The cleavage
reaction was carried out for 12 hours and then chromatograms were obtained
from each cleavage reaction mixture. From the chromatograms (not shown), it
was evident that the reaction yields were not good using any of the three
experimental conditions. Moreover, with the use of ammonia and borate buffer,
some intact and only partially degraded IF2 was still observed in the

chromatograms, meaning that the cleavage reaction did not proceed completely.

 

 

 

 

 

 

 

 

 

 

 

NH,
MT“: —1 Near-o-
”I?”
. , 10' oo’n
WEI 1 3°’ 1
E pH-e (——10"M
10'1" w.9j __1g:M
pl-l-lo ,_1 M
s-—1o~?llll 20m 1 ,_104M
~<——10~‘M 10' 10"»
K—10“M 20, r
K—10'5M 30' mg...”
fl 37°C l
Nedeavage 60°C NH
1 Poorcleevage ' 3 l
mm um... 10'3M 1M
”W” l l
’1MNH3 pH-8.3 Microwave 10"T_1oec
0.1MBorateBuffer pit-9.0 30%power WW
*3.5MMeNH2 pH=10 T~10°C Pawer:30%
40%
l a.
12hr: New [10%
l a, Nobettsrolesvege
mm- m ......

 

 

 

 

 

 

 

 

Figure 49. Summary of representative experiments at room temperature using lF2. Squares
point out the conditions for the control experiments.

117

Experiments at lower pH were carried out at different pH values varying
temperature and reaction times also in order to check if at values where pH <
pKa nucleophiles may promote the cleavage reaction with a good reaction yield.
However, only poor cleavage reaction was observed. Similar results were
obtained with the use of low nucleophile concentrations. From results of such
studies, it was concluded that there was not an advantage to increase the

cleavage product yields under those tested experimental conditions.

Microwave-assisted organic synthesis has been applied in a wide variety
of reactions [25, 26]; in the protein field it has been applied to shorten acid
hydrolysis time for proteins [27, 28]. Also, there are reports of enzymatic
digestion assisted by microwaves as an effective and useful method [29-32].
Because of these reports, it was thought that the use of microwaves might be
advantageous for the chemical cleavage reaction. Based on results obtained in
this laboratory for other studies using microwaves in protein analysis, a series of
experiments was carried out using a laboratory-adapted microwave oven.
However, there was no significant improvement in the yield of cleavage products;

thus, further use of microwaves was discontinued.

118

3.2b Effect of low temperature and reaction time on the cleavage

reaction

Performance of the cleavage reaction at 2 °C results in a significant
diminution in the products of side-reactions as shown in the comparison of
chromatograms in Figure 50. This observation is consistent with results from
earlier studies of the cleavage reaction of some peptides or proteins [18, 33]. In
our study of the effect of reduced temperature on the CN-induced cleavage
reaction of a more complex protein, IF4 was treated with 1 M methylamine or 1 M
ammonia at 2 °C in a cold room. A noticeable improvement in the yield of the
CN-induced cleavage product was realized at 2 °C when methylamine was the
nucleophile for a 10-min reaction (see Figure 50). In comparison, the area under
the curve for the middle cleavage product B at 2 °C is about 25% greater than
that for the equivalent reaction carried out at room temperature. Similarly, the
cleaner and significantly different HPLC peak for the cleavage product A in the
lower panel of Figure 50 suggests reduced competition from side reactions at
lower temperature. We also investigated the influence of longer reaction time for

samples treated with methylamine at low temperature.

119

 

  
 

 

 

 

 

 

 

 

C
0.10-
‘ CH3NH2 10'
:5; 0.05. room temperature
If ' a —
Fa 0.00' i i i Ifi i
.e 20 40 60 80 100 120
8 0.15.
.o
<
.. BC
0'10 CH3NH2 2°c 10'
0.05-
0-00' I'I'I'T'I‘I'lfi'l'l
20 30 40 50 60 70 80 90 100110120
Time(min)

 

 

 

Figure 50. Comparison of HPLC chromatograms of cleavage reaction mixtures of IF4
using 1 M methylamine at room temperature or at 2 °C for 10 minutes. The peaks
labeled A, B, and C represent the cleavage products as defined in Figure 43.

The results (data not shown) indicated no appreciable increase in the

yields of the CN-induced cleavage products. Hence, 10 minutes at 2°C seems to

be optimal in maximizing the yield of cleavage products, while minimizing side-

reactions.

Samples treated with 1 M ammonia at low temperature for 10 minutes
showed only a moderate improvement in the yield of cleavage products over
those obtained using previously described conditions [9, 10]. Additional

experiments with longer reaction times in 1 M NH3 showed only a slight

120

improvement in yield of the CN-induced cleavage fragments, but less than the

yields using 1 M methylamine (data not shown).

3.2c Effect of nucleophile concentration on the cleavage reaction

The next factor evaluated was the concentration of the nucleophiles (at pH
~ 11.6). Samples of IF4 were treated with different concentrations of ammonia or
methylamine, and the cleavage reaction mixtures were analyzed using p-HPLC.
The cleavage product yield increased linearly with [NH3] up to about 5 M and with
[CH3NHz] up to 2 M. These concentrations were considered optimum because,
in the case of ammonia beyond 5 M, there is not a noticeable increment (<10%)
of the cleavage product yields that could justify an increase in the nucleophile
concentration; for methylamine, at concentrations higher than 2 M, the yield of
some cleavage products decreased, suggesting the involvement of side

reactions (Figure 51).

121

 

 

100 I
so I
60
40
20

 

eel
>Om

 

 

I}

 

‘I

100
80
60
40
20

Relative yield (7.)

CH3NH2

t V I V

1.0 1.5 2:0 2.5 3:0
Concentration (M)

 

 

 

Figure 51. Relative yield of cleavage products using methylamine for IF4 from RNase-A
as function of the concentration at 2°C. Labels B, C, and A correspond to the cleavage
fragments according Table 3 and Figure 43.

3.2d Repeatability of cleavage reaction yields

Because it is important to achieve statistical control of a measurement
[34], a series of analyses was conducted to evaluate the precision of the
cleavage reaction yield. Samples at three isoform concentration levels were
selected for the experiments: ~0.022 nmol/pL (denoted 1x), ~0.103 nmol/pL (5x),
and ~0.340 nmol/pL (15x); three reaction conditions for each level were studied:
5 M ammonia (5000 nmol/pL) at 2 °C for 10 minutes, 2 M methylamine (2000

nmol/pL) at 2 °C for 10 minutes, and the conventional conditions previously

122

described in section 2.6 and elsewhere [9, 10]. Triplicate analyses were carried

out for each level.

The cleavage reaction mixtures were analyzed by HPLC with UV
absorption for determination of the yield of each cleavage product. Because UV
absorptivity is proportional to the number of peptide bonds in a protein or peptide
[23, 35], calibration curves were made using the peak area from chromatograms
for known concentrations of peptides containing a specified number of amino
acid residues. From the peak areas in chromatograms for the separation of the
cleavage reaction mixture, cleavage product amounts and yields were calculated
for each experiment taking into account that ideally 1 nmol of IF4 would produce

1 nmol of each cleavage fragments.

Statistical data were obtained for isoform levels 1x, 5x, and 15x using the
JMP 5.1© statistical software [36]. The use of this software allows graphical

visualization to compare multiple data groups (Figure 52).

123

 

 

sample data

/ .
4} overlap marks (% 95% confidence level
A) v ¥v

7
group MOSH

 

 

 

 

 

 

A

 

 

not significantly different

:0: ‘0:

B)

significantly different

‘5’

Comparison circles

 

 

 

 

 

 

Figure 52. A) Mean diamonds graphs. The middle line in the diamond represents the
mean for a group of data. The vertical endpoints form the 95% confidence interval for the
mean. Overlap marks are useful for comparison of means, but it is better to use of
comparison circle plots. 8) Two hypothetical cases where means of two sample data
groups are not significantly and are significantly different. Comparison circle plots for each
group are evident each case.

These statistical results indicate that similar treatments produce similar
amounts of cleavage products proportional to the initial amount of the isoform;
e.g., the amount of cleavage products at level 15x was roughly 15 times the
amount at level 1x under identical experimental conditions. In general, there was
no significant difference (p = 0.05) in the proportional amount of cleavage
products at different levels. It is evident that the molar ratio of
nucleophile/isoform is so high that a change in the amount of available isoform

for the reaction does not affect the yield of cleavage fragments significantly. The

124

coefficients of variance (CV) at levels 15x and 5x are, in general, no higher than
10% (see Table 4 and Figure 53). For level 1x, the CV values are relatively high
(>10%) mainly because the chromatographic peaks are smaller and the

chromatographic integration is more affected by the background noise.

Comparison of mean values for the amount of the cleavage products
obtained shows that the cleavage reaction yields for fragments A and B vary
significantly with reaction conditions (see Table 4 ); the yield of these cleavage
fragments is much higher under the optimized methodology. It is not surprising
that the yield of cleavage fragment C did not change significantly with reaction
conditions because according to our previous studies, this fragment is produced
with great facility. For example, the yield of fragment C is the same at 10
minutes under the optimized cleavage conditions as it was after using the

conventional conditions.

125

 

 

 

 

 

 

 

0.45
m 1
E 0.” "t" ©
3 9. ..____LL_.
El 0.35: °
§ 0.30.
> -- . -
8 . O
T: 0.25I * z
3
5 020*
*2.
c 0.151
.9 - .
‘La’ 0.104 TL ©
M10_1x M10_5x N10_1x N10_5x mo_1x mojx Each Par
Studert'st0.05
Experimertd oondtions

 

 

 

Figure 53. Distribution analysis and mean comparison for yield of cleavage product B. M10 =
Experiments with methylamine, 10 minutes at 2 °C. N10 = Experiments with ammonia, 10
minutes at 2 °C. N60 = Experiments with ammonia, 60 minutes at room temperature.

Table 4. Representative data for cleavage product yields for IF4 obtained at level 15x under
different experimental conditions. In parenthesis, the coefficient of variation (CV) is expressed
as a percentage, for experiments under the same treatment.

 

 

 

 

 

A yield in percent B yield in percent C yield in percent
15x (CV) (CV) (CV)
lMeNHz, 2 °c, 10' 41 (10.8%) 30 (4.8%) 89 (4.9%)
INH3, 2 °c, 10' 27 (4.4%L 21 (3.7%) 101 (2.14%)
[NH3, rt, 60’ 2 (4.9%) 6 (9.6%) 95 (5.0%)

 

 

 

rt = room temperature

126

 

3.3 Evaluation of the optimized experimental conditions

3.3a Comparison of MALDI-mass spectra of cleavage reaction

mixtures prepared under modified conditions at low temperature

Our main goal is to determine the disulfide linkage in a cystinyl protein
through the correct assignment and identification of the CN-induced cleavage
products by using mass spectrometry. Better yields of cleavage products lead to
better quality mass spectra, and by consequence, more precise and
unambiguous assignment of cleavage fragments. In order to evaluate mass
spectra of cleavage reaction mixtures prepared under the newly optimized
experimental conditions, single-reduced and cyanylated isoforms of RNase-A
(IF1 to IF4) were treated with methylamine, ammonia, or other candidate
nucleophiles including: ethylamine, propylamine, and butylamine. For a
controlled comparison, a sample of IF4 was also treated under the conventional
methodology (1 M ammonia, 60’ reaction time). Samples of cleavage reaction
mixtures were spotted on a plate for the MALDI-MS experiment according to the

method described in the experimental section.

Figure 54 shows a comparison of MALDI-mass spectra for the cleavage
reaction mixture of IF4 using CH3NH2 or NH3 under different experimental
conditions. It is evident that peaks labeled A and 8, representing the N-terminal

and middle fragments, respectively, become more prominent with 5 M NH4OH

127

and 2 M CH3NH2 compared to those observed in the spectrum of the reaction
mixture prepared under the traditional experimental conditions (60 min at room

temperature in 1 M of NH4OH).

Expected and characteristic is the shift in the mass of the cleavage
products A and B (as defined in Figure 43) according to the nucleophile used.
These two cleavage products terminate in the form of an amide that incorporates
the nucleophile (Nu). For example, if the nucleophile is ammonia, Nu is —NH2; in
the case of methylamine, Nu is -NH-CH3, etc. Thus, cleavage products A and B
generated with a mixture of ammonia and methylamine as nucleophiles will be
represented by a pair of mass spectral peaks separated by 14 m/z units. On the
other hand, the itz-C-terminal product (C) will always show the same molecular

mass regardless of the nucleophile used.

128

 

 

 

 

 

 

 

 

 

 

 

 

C
100 C M
M
A 90
2‘: an
:E‘ 70
3 so
2 50
g 40
3 30
C,
or. 20
10
o A B 3
1MNH4OH,I11,50'
5MNH‘OH,2°C,10' L- 1 12000
8000
6000
2MMeNH2,2°C,10' 4000 /
2000 in"
C'm‘g" Residues m/z III/z
product
Fragment -NH1 -NHCH3
A 157 6351.1 6365.1
B itz-58409 5766.4 5730.4
C itz-110424 1659.8 1659.8

 

 

 

 

 

 

 

 

 

Figure 54. MALDI mass spectra of the reaction mixture resulting from the cleavage of
cyanylated IF4 from RNase-A by the indicated nucleophiles. Peaks labeled A, B, and C
represent segments of the sequence as defined in Figure 43.

Analytical advantage can be gained in the use of a binary mixture of
homologous nucleophiles for cyanylation-induced cleavage of partially reduced
isoforms of cystinyl proteins. The appearance of pairs of mass spectral peaks
offset by 14 (or 28, 42, etc) mass units facilitates recognition of diagnostic CN-
induced cleavage products, and increases confidence in results when samples

present weak signals or interfering chemical noise during analysis by MS. The

129

simultaneous use of two homologous nucleophiles is also applicable to the

identification of cleavage products from double-reduced isoforms, which are also

useful in the elucidation of disulfide linkages [20].

Figure 55 shows a comparison of the MALDI mass spectra for the
cleavage reaction mixture of IF4 treated with different primary amines. Mass
spectral peaks representing cleavage products from reactions with methylamine

show a higher intensity compared to those from analysis of reaction mixtures

processed with the other amines.

 

 

 

 

 

 

 

 

 

100 1/
90 j C
90 '1
70 1
60 ".
50 ‘.
40 1 3A
30 ‘. N.
20 ‘. A
10 1 14000
0 ‘ u‘ III” A 103002000
2M MeNHz m /
2M EtNHZ . I“ . 40006000 \~
2M PropNH2 m
2M ButNH2

 

 

Figure 55. MALDI-MS spectra of the cleavage product mixture resulting from the
processing of IF4 from RNase-A with primary amines at low temperature (2 °C)

and a concentration of 2 M. From back to the front: methylamine, ethylamine,
propylamine, or butylamine.

130

It has been demonstrated [9] that amino acids with bulky side chain
groups, such as Pro and Tyr, on the N-terrninal side of the cyanylated cysteine
inhibit the cleavage process. The alkyl chain of the alkyl amine increases the
nucleophilicity, but steric problems can be present with larger hydrocarbon
chains. According to our observations, methylamine produces a better cleavage
product yield because it is a good nucleophile, but it has a relatively small methyl
group. In the case of the other primary amines, even though they could also be
considered good nucleophiles [37, 38], the length of the hydrocarbon chain
seems to impair their effectiveness during the nucleophilic attack. However,
there is no evidence for any specific side reaction. Therefore, methylamine
appears to be a good compromise between functioning as a strong nucleophile

and avoiding steric hindrance.

3.3b Application of the modified methodology to the analysis of a-

Iactalbumin

In order to evaluate the modified methodology with a different protein, a-
Iactalbumin (m/z for MH"average = 14176) was used as another model protein; it
contains 123 amino acid residues among which eight cysteines are linked by four
disulfide bonds: Cys 6-Cys 120, Cys 28-Cys 111, Cys 61-Cys 77, and Cys 73-
Cys 91. Partial reduction and cyanylation of tit-Iactalbumin were carried out
following a procedure similar to that described for RNase-A. Single-reduced and

cyanylated isoforms of tat-Iactalbumin were separated and collected using reverse

131

phase-HPLC. Each isoform was dried, reconstituted in 6 M guanidine, and used
to prepare samples for treatment with different nucleophiles: ammonia,
methylamine, or ethylamine under the optimized experimental conditions. MALDI
mass spectra of reaction mixtures under optimized conditions were compared to

those previously reported using the conventional methodology [9, 10].

Although MALDI-MS is not a quantitative technique, relative peak
intensities for the cleavage products might be used as a rough estimate of the
cleavage product yields. Mass spectral peaks were more discernible during
analysis of cleavage reaction mixtures prepared with ammonia under optimized
reaction conditions (5 M NH4OH, 2 °C, 10 minutes) than in those prepared with
the conventional methodology; thus, we can infer that yields of the cleavage

products were better under the optimized reaction conditions with ammonia.

The pattern of cleavage products changed when using methyl- or
ethylamine as the nucleophile in that cleavage fragment 1-60 was the most
abundant, whereas cleavage fragment itz-77-123 was the most abundant when
using ammonia (see Table 5). These observations suggest that it is easier for
the primary amines to cleave the cyanylated cysteine located at position 77 than
the one located at position 61. This phenomenon can be rationalized by
examining the amino acid sequence for a-lactalbumin in which the amino acid
residue on the N-terrninal side of cysteine 61 is Trp, the bulky side chain of which

may present steric hindrance to the cyclization cleavage reaction on the N-

132

terminal side of the cyanylated cysteine residue. This phenomenon may be
analogous to that reported previously [9] for Pro or Tyr. It also seems reasonable
that the sidechain of Trp might interfere with access of the nucleophile to the

cyanylated cysteine.

It was explained earlier that cleavage product B of a given single-reduced
isoform, obtained by reaction with methylamine, is 14 Da heavier than that
obtained with ammonia. This feature made it possible to recognize the cleavage
product itz-73-90 from one of the isoforms of til-Iactalbumin even though the
relative intensity in the MALDI mass spectrum (data not shown) for this fragment

was low (<1%).

133

Table 5. Relative intensities from MALDI mass spectra for protonated cleavage products
produced using ammonia, methylamine, or ethylamine as cleavage reagents for single-
reducedlcyanylated tat-Iactalbumin isoforms (IF1L, |F2L, lF3L, IF4L). Calculated m/z values are
for protonated cleavage products produced by nucleophilic attack by ammonia.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IF1 L
Open Cys 6 - Cys 120 Nucleophile
*NH30H, NH3OH CH3NH2 CH3CH2NH3
Fragment W2
1 M,rt,60’ 5M,2°C,10’ 2M,2°C,10’ 5M,2°C,10’
1 — 5 618.7 ND ND ND ND
itz-6 -119 13, 135.7 ND 2.91 1.61 2.18
itz-120 - 123 517.6 100 100 100 100
IF2L
Open Cys 61 - Cys 77 Nucleophile
‘NH30H, NH30H CHaNH; CH3CH2NH2
Fragment m/z
1 M, rt, 60’ 5M,2°C,10’ 2M,2°C,10' 5M,2°C,10’
1 - 60 6918.7 ~ 8 100 76.0 65.1
itz-61 - 76 1800.8 ~ 8 26.2 45.8 21.3
itz-77 - 123 5552.5 100 69.7 100 100
lF3L
Open Cys 28 — Cys 111 Nucleophile
*NH30H, NH30H CH3NH2 CH,CH2NH;
Fragment ml:
1 M, rt, 60’ 5M,2°C,10’ 2M,2°C,10’ 5M,2°C, 10’
1-27 3125.6 100 100 100 51.0
itz-28 - 110 9525.6 ~ 2 2.01 1.99 2.10
itz-111 - 123 1620.8 ~ 98 98.5 84.4 100
IF4L
Open Cys 73 - Cys 91 Nucleophile
*NI-I3OH, NH;OH CH3NH2 CH3CH2NH2
Fragment m/z
1 M, rt, 60’ 5M,2°C,10’ 2 M,2°C,10’ 5M,2°C,10’
1 — 72 8258.1 ~ 5.00 7.86 12.9 5.18
itz-73 — 90 2098.3 ~ 4.00 7.08 4.10 1.00
itz-91 — 123 3915.6 100 100 100 100

 

 

 

 

 

ND = Not detected, rt = room temperature, * Data from reference [10]

134

 

4.- CONCLUSIONS

Methylamine (2 M) was found to be superior to ammonia as the cleavage
nucleophile in promoting the CN-induced cleavage reaction during disulfide mass
mapping. However, increasing the concentration of ammonia to 5 M (beyond the
previously used 1 M) in combination with low temperature (2° C instead of room
temperature) and short reaction time increased the cleavage product yields
substantially (in some cases, by an order of magnitude). Low temperature (2 °C)
and a short reaction time (10 min) minimize alkali-induced damage of the
cleavage products and other side reactions that occur during longer incubation
times in the original protocol (1 hour or more). Good repeatability (less than 10%
relative standard deviation) was observed in the yield of cleavage products from

~100-1000 pmol of starting material.

The use of two or more homologous nucleophiles can be visualized as an
additional tool to facilitate the identification of cleavage products from mass
spectra. Representation of certain cleavage products by pairs of mass spectral
peaks separated by 14 mass units helps the analyst to distinguish these
structurally diagnostic species, especially in cases of low intensity signals and/or

in the presence of potentially interfering chemical noise.

135

5.- REFERENCES

[1] K. K. Han, C. Richard, G. Biserte, Current developments in chemical cleavage
of proteins, Int. J. Biochem. 15 (1983) 875-884.

[2] B. J. Smith, Chemical cleavage of polypeptides, Methods MOI. Biol. 211
(2003) 63-82.

[3] T. F. Spande, B. Witkop, Y. Degani, A. Patchomik, Selective cleavage and
modification of peptides and proteins, Adv. Protein Chem. 24 (1970) 97-260.

[4] B. Witkop, Chemical cleavage of proteins. Selective fragmentations and
modifications reveal structure, Science 162 (1968) 31 8-326.

[5] S. Doonan, Peptides and proteins, Royal Society of Chemistry, Cambridge,
UK., 2002.

[6] G. R. Jacobson, M. H. Schaffer, G. R. Stark, T. C. Vanaman, Specific
chemical cleavage in high yield at the amino peptide bonds of cysteine and
cystine residues, J. Biol. Chem. 248 (1973) 6583-6591.

[7] S. Nakagawa, Y. Tamakashi, T. Hamana, M. Kawase, S. Taketomi, Y.
lshibashi, O. Nishimura, T. Fukuda, Chemical cleavage of recombinant fusion
proteins to yield peptide amides, J. Am. Chem. Soc. 116 (1994) 5513-5514.

[8] G. R. Stark, Cleavage at cysteine after cyanylation, Methods Enzymol. 47
(1977) 129-132.

[9] J. Wu, J. T. Watson, Optimization of the cleavage reaction for cyanylated
cysteinyl proteins for efficient and simplified mass mapping, Anal. Biochem. 258
(1998) 268-276.

[10] J. Wu, J. T. Watson, A novel methodology for assignment of disulfide bond
pairings in proteins, Protein Sci. 6 (1997) 391-398.

[11] J. Wu, J. T. Watson, Assignment of disulfide bonds in proteins by chemical
cleavage and peptide mapping by mass spectrometry, Methods in Molecular
Biology (Totowa, NJ, United States) 194 (2002) 1-22.

[12] J. Qi, J. Wu, G. A. Somkuti, J. T. Watson, Determination of the disulfide

structure of sillucin, a highly knotted, cysteine-rich peptide, by
cyanylation/cleavage mass mapping, Biochemistry 40 (2001) 4531-4538.

136

[13] J. Wu, Y. Yang, J. T. Watson, Trapping of intermediates during the refolding
of recombinant human epidermal growth factor (hEGF) by cyanylation, and
subsequent structural elucidation by mass spectrometry, Protein Science 7
(1998) 1017-1028.

[14] J. Qi, W. Wu, C. R. Borges, D. Hang, M. Rupp, E. Torng, J. T. Watson,
Automated data interpretation based on the concept of "negative signature mass"
for mass-mapping disulfide structures of cystinyl proteins, J. Am. Soc. Mass
Spectrom. 14 (2003) 1032-1038.

[15] W. Wu, J. T. Watson, F. J. Stevens, R. Yousefzai, L. Anderson, Chloroplast
fructose bisphosphatase contains alternate disulfide bonds, Abstracts, 35th Great
Lakes Regional Meeting of the American Chemical Society, Chicago, IL, United
States, May 31 -June 2 (2003) 110.

[16] M. Xu, A. Arulandu, D. K. Struck, S. Swanson, J. C. Sacchettini, R. Young,
Disulfide isomerization after membrane release of its SAR domain activates P1
lysozyme, Science 307 (2005) 1 13-1 17.

[17] Y. Degani, A. Patchomik, Cyanylation of sulfhyde groups by 2-nitrO-5-
thiocyanobenzoic acid. High-yield modification and cleavage of peptides at
cysteine residues, Biochemistry 13 (1974) 1-11.

[18] S. Nakagawa, Y. Tamakashi, Y. lshibashi, M. Kawase, S. Taketomi, O.
Nishimura, T. Fukuda, Production of hPTH(1-34) via a recombinant DNA
technique, Pept. Chem. 31 (1993) 5-8.

[19] S. Nakagawa, Y. Tamakashi, Y. lshibashi, M. Kawase, S. Taketomi, O.
Nishimura, T. Fukuda, Production of human PTH(1-34) via a recombinant DNA
technique, Biochem. Biophys. Res. Commun. 200 (1 994) 1 735-1 741 .

[20] W. Wu, W. Huang, J. Qi, Y. T. Chou, E. Torng, J. T. Watson, 'Signature
sets', minimal fragment sets for identifying protein disulfide structures with
cyanylation-based mass mapping methodology, J. Proteome Res. 3 (2004) 770-
777.

[21] ASTM D 1193-99e1, "Standard Specification for Reagent Water", ASTM
International.

[22] M. Cadene, B. T. Chait, A robust, detergent-friendly method for mass
spectrometric analysis of integral membrane proteins, Anal. Chem. 72 (2000)
5655-5658. -

[23] R. K. Scopes, Measurement of protein by spectrophotometry at 205 nm,
Anal. Biochem. 59 (1974) 277-282.

137

[24] J. R. Whitaker, P. E. Granum, An absolute method for protein determination
based on difference in absorbance at 235 and 280 nm, Anal. Biochem. 109
(1980) 156-159.

[25] M. Larhed, C. Moberg, A. Hallberg, Microwave-accelerated homogeneous
catalysis in organic chemistry, Acc. Chem. Res. 35 (2002) 717-727.

[26] N. S. Wilson, G. P. Roth, Recent trends in microwave-assisted synthesis,
Curr. Opin. Drug Discov. Dave]. 5 (2002) 620-629.

[27] S. T. Chen, S. H. Chiou, Y. H. Chu, K. T. Wang, Rapid hydrolysis of proteins
and peptides by means of microwave technology and its application to amino
acid analysis, Int. J. Pept. Protein Res. 30 (1987) 572-576.

[28] B. W. Li, Comparison of microwave oven and convection oven for acid
hydrolysis of dietary fiber polysaccharides, J. AOAC. Int. 81 (1998) 1277-1280.

[29] V. J. Nesatyy, A. Dacanay, J. F. Kelly, N. W. Ross, Microwave-assisted
protein staining: mass spectrometry compatible methods for rapid protein
visualisation, Rapid Commun. Mass Spectrom. 16 (2002) 272-280.

[30] B. N. Pramanik, U. A. Mirza, Y. H. lng, Y. H. Liu, P. L. Bartner, P. C. Weber,
A. K. Bose, Microwave-enhanced enzyme reaction for protein mapping by mass
spectrometry: a new approach to protein digestion in minutes, Protein Sci. 11
(2002) 2676-2687.

[31] H. Zhong, S. L. Marcus, L. Li, Microwave-assisted acid hydrolysis of proteins
combined with liquid chromatography MALDI MSIMS for protein identification, J.
Am. Soc. Mass Spectrom. 16 (2005) 471-481.

[32] S. S. Lin, C. H. Wu, M. C. Sun, C. M. Sun, Y. P. Ho, Microwave-assisted
enzyme-catalyzed reactions in various solvent systems, J. Am. Soc. Mass
Spectrom. 16 (2005) 581-588.

[33] E. Gross, J. L. Morell, The reaction of cyanogen bromide with S-
methylcysteine: fragmentation of the peptide 14-29 of bovine pancreatic
ribonuclease A, Biochem. Biophys. Res. Commun. 59 (1974) 1145-1150.

[34] G. T. Wemimont, W. Spendley, Use of Statistics to Develop and Evaluate
Analytical Methods, AOAC Intl, Arlington, VA., 1985.

[35] C. N. Pace, F. Vajdos, L. Fee, G. Grimsley, T. Gray, How to measure and
predict the molar absorption coefficient of a protein, Protein Sci. 4 (1995) 2411-
2423.

[36] JMP IN software for statistical analysis, Version 5.1, SAS Institute Inc., Cary,
NC., 2003.

138

[37] A. Miller, P. H. Solomon, Writing Reaction Mechanisms in Organic
Chemistry, Second Edition, Academic Press, Orlando, FL., 1999.

[38] C. G. Swain, C. B. Scott, Concerted displacement reactions. X. Quantitative
correlation of relative rates. Comparison of hydroxide ions with other nucleophilic
reagents toward alkyl halides, esters, epoxides and acyl halides, J. Am. Chem.
Soc. 75 (1953) 141-147.

139

CHAPTER 3. USE OF HOMOLOGOUS NUCLEOPHILES FOR CHEMICAL
CLEAVAGE AND LABELING IN CYANYLATION-BASED DISULFIDE MASS

MAPPING

1.- INTRODUCTION

1.1 Chemical Tagging Methodology for Quantitative Proteomics

Mass spectrometry has become the technology of choice for rapid
identification of proteins in a biological mixture [1]. Mass spectrometry has
several advantages over other analytical techniques; it has high sensitivity,
accuracy and capacity. Tandem mass spectrometry provides an adequate tool
to identify peptides and proteins through the two principal fragment ions that are
produced; the so-called N- and C-terminal b- and y-type ions, respectively. High
quality CID-MSIMS of enzymatic peptides show prominent b- and y-ion series.
Database search algorithms can be used in conjunction with MSIMS to readily
identify peptides and proteins. Computer-based identification of dozens to

hundreds of proteins now can be attained in a single MS analysis [2].

There are several ways to face the more challenging problem of de novo
sequencing (manual interpretation of peptide spectra for protein identification
without a database). One is to reduce the complexity of peptide spectra by using

stable-isotope labeling. Isotopic labeling can be performed by enzymatic

140

digestion of proteins in 16O/mO water, eliminating the need for peptide
derivatization. Specifically, protein digestion by trypsin in the presence of 50% of
H2180 incorporates the 18O atom selectively into the C-terminal carboxyl groups
of peptides. Subsequent fragmentation by CID-MSIMS distinguishes the y-ion
series by a characteristic 2-mass unit shift, which facilitates read-out of the

sequence[3L

Other approaches for peptide sequencing and quantitative analysis
applied in the genomics and proteomics fields involve protein-tag approaches.
The MCAT (Mass-Coded Abundace Tagging) method relies on the selective and
quantitative guanidation of the e-amino group of the C-terminal lysine residues of
tryptic peptides. At high pH, reaction with O-methylisourea selectively transforms
lysine into homoarginine, which is 42 Da heavier than lysine. For peptide
sequencing, the comparison of MSIMS spectra for each sister peptide pair
permits the informative y-ion peaks to be readily discerned, allowing for
systematic sequence determination (see Figure 56) [2]. Programs such as
SEOUEST [4] can be useful, readily recognizing MCAT-derivatized peptides. For
quantitative comparison of proteins present in samples derived from different cell
states, one of the samples is treated with O-methylurea; then, the two samples
are combined and analyzed by LC-MS. Full-scan MS spectra are recorded and
the relative abundances of sister peptide species are determined by comparing

the profiles of reconstructed single-ion chromatograms (see C in Figure 56).

141

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a) Guanldlnatlon of lysine
NH;
|
CH-NH
NH, I
l NH
CH2 I
l cu,
CH2 I
l CH2
OH; |
l cu2
CH2 OCH: I
| pH 10 CH2
-NH-CH-CH-C- " "Hz.c.N" —" I "’ CH3OH
ll -NH-CH-CH-C-
° l
b) Peptide sequencing
YVLVPIEHVAI‘
I I I I
l l l l
Trypsln ~° modification Combining ' H 1'
digestion in deselting
< Modification with
O-rnethyiisouru_ ESI
0c": 7 De New
I identification of
NHz-Cle-i Imlno acid sequence
c) Peptide quantitation
Trypsln digestion‘ No modlfication_
Combining A: \
& deseltln
E —' l
Modlflcation with
' OCH3 ' ion chromatogram
l
NHz-CsNi-l

 

Figure 56. Overview of MCAT peptide sequencing and quantitation, from reference [2].

MCAT is based on the previously developed methodology for a
quantitative analysis of complex protein mixtures developed by Gigy et al using
Isotope-Coded Affinty Tags (lCATs) and tandem mass spectrometry. In this
approach, the ICAT reagent consists of three functional elements: a chemically
specific reactive group (e. g., with specificity toward thiol groups), an isotopically
coded linker (to incorporate stable isotopes), and an affinity tag (e .g., biotin). An

ICAT reagent with specificity to thiol groups is shown in Figure 57.

142

 

 

O

A . x x .
M" Lirlwx x moi-specific

: Reedylvegup

Heavyzx
Uth

 

 

 

Figure 57. Structure of an ICAT reagent. The reagent exists in two forms, heavy (containing 8
deuteriums, X = D) and light (contains no deuteriums, X = H).

In the lCAT approach, proteins from cell state 1 are labeled using the light
thiol-specific ICAT reagent (producing species do). Proteins coming from a cell
state 2 are labeled with the heavy ICAT (producing species d3). These protein
mixtures are combined and proteolyzed to peptides. lCAT-labeled peptides are
isolated using avidin affinity chromatography based on the biotin-avidin
interaction; thus, only labeled peptides are isolated. Finally, the isolated peptides
are separated and analyzed by p—LC-MS/MS. A pair of lCAT-labeled peptides is
chemically identical and is easily visualized because they essentially coelute, but
there is an 8-Da mass difference measured in mass spectra for single charged
species. The relative quantification is determined by the ratio of peptide pairs ((10
vs d3), which represent peptides of the same protein in the two different states

(See Figure 58) [5].

143

 

 

 

Cell State 1 Cell State 2
(all cysteines labeled (all cysteines labeled
with light ICAD with heavy ICAT)
\ /
Combine

optionally fractionate,
and proteolyze

I

Affinity isolation of
lCAT-labeled peptides

Analyze by LC-MS

 

 

Quantitative relative protein levels by measuring peak ratios

 

 

O
100 §§§$§§§§
: a“ «‘9 a” Q8 «34"

1

 

 

 

 

III I

A Retention time

Relative Abundance

 

 

Identify proteins by sequence levels by measuring peak ratios

 

 

K ICAT label

-NH,-EACDPLR-COOH —> Peptide defines Protein A

nub

O

O
n

l l I l

 

Relative Abundance
O

IIIIII IIIII

 

 

Figure 58. ICAT strategy for quantifying differential protein expression,
based on thiol-specific lCAT reagents (adapted from reference [5]).

144

Wu et al proposed a methodology that uses brominated, structurally
homologous reagents to quantify cysteinyl peptides. The two reagents used in
this approach are 2-bromoacetamide (2-BA) and 2—bromopropinamide (2-BP).
Even though the use of alkylating reagents could lead to problems of under— or
over-modification, after optimization this strategy seems to be a good choice to

analyze the relative concentration ratio of disulfide isomers [6].

1.2 Application of labeled methodologies in determination of

disulfide bonds

As with the methods reviewed earlier, the use of methodologies where
peptides are labeled are not common for the determination of disulfide linkages.
However, the use of pepsin in solutions containing 50% of H2180 has been
proposed. After digestion, resultant disulfide-linked peptides have distinct
isotope profiles compared to the same peptides with only 160 in their terminal
carboxylates. Thus, it is possible to recognize disulfide-linked peptide in digests
and chromatographic fractions using these mass-specific markers, and to
rationalize mass changes upon reduction in terms of half-cystinyl sequences of
the protein of interest. The use of CID-MSIMS in this approach also is facilitated
by the presence of 18O in the peptide product of the enzymatic reaction [7, 8].
Even though some applications have been reported [9], this method is quite

complicated experimentally, and its use is not widespread.

145

The cyanylation (CN)-based mass mapping methodology [10] for
assignment of disulfide bond connectivities has proven useful even in proteins
containing adjacent cysteines. As was explained in chapter two, the essence of
the method is based on selective cyanylation of the sulfhydryl group of a cysteine
residue, and subsequent nucleophilic attack to promote CN-induced cleavage on
the N-terminal side of the modified cysteine. However, in some cases, the low
yield or poor ionization of some cleavage fragments, makes data interpretation
difficult because signals cannot be considered as “true signals" representing the
cleavage fragments. Thus, it is possible to lose some valuable information that
might be useful for a correct interpretation or for feeding algorithms that help with
the assignment of the disulfide bonds [6, 11, 12]. We optimized experimental
conditions using ammonia or methylamine to produce good yields of cleavage
fragments in the context of the CN-based disulfide mass mapping methodology
based on results from both UV-absorption HPLC chromatograms and MALDI-MS
(see Chapter 2). The parallel use of two nucleophiles (e.g., ammonia and methyl
amine) increases confidence in distinguishing cleavage fragments from chemical

byproducts or noise.

During the nucleophilic cleavage step in CN-based disulfide mass
mapping methodology, three cleavage fragments are produced: the N-terminal
peptide (A), the middle iminothiazolidine (itz)-peptide (B), and the itz-C-terrninal
peptide (C). In the case of the N-terminal fragment (A) and the middle itz-peptide

(B), the C-terminus of these peptides contains the nucleophile group (Nu). For

146

example, if the nucleophile is ammonia, the Nu group is -NH2; in the case of
methylamine, the Nu is -NH-CH3; and for ethylamine, —NH-CH2-CH3, etc. Thus,
when a binary mixture of two homologous nucleophiles is used, the A and B
fragments will show a mass difference of 14 Da in MALDI-MS. On the other
hand, the C-terminal fragment (C) will always show the same mass in spite the
nucleophile used. This expected pattern improves confidence in detecting CN-

induced cleavage products (see Figure 59).

 

' Nucleophiiic cleavage step of a single-reduced and cyanylated
protein (two cyanylated cysteines) gives three cleavage products:

' A is the N-terminal peptide
° B is the middle iminothiazolidine (itz)-peptide
° C is the C-terminal itz-peptide

s7 Itz 109 Itz

 

1———\ ‘ \ \— 124
N" 58 N" 11° Am/zA _ A: 14 units
A B C 14 o
“.1 Ala A14
- If the nucleophile is: u: I: II
III
- ammonia, the Nu group is -NH2 (A0, 80, C) W I I

“mm

- methylamine, the Nu group is -NH-CH3 (A14. 314. C) ...I {III
a. I l I

._ -J
unuseuosnuuoosueuuuuuee

 

 

 

 

 

1 s7 itk 109 [k 124
NH2 58 ‘NHZ 110
A0 Bo C

 

Figure 59. impact of different nucleophiles on the cleavage reaction. Species produced using
ammonia are designated with “o” and cleavage fragments based on methylamine are
designated with subscript “14".

Proposed in this chapter is the development of a method to facilitate

recognition of cleavage products using a binary mixture of homologous

147

nucleophiles during mass mapping of the disulfide structure of cystinyl proteins
by means of the Partial Reduction/ Cyanylation/ Cleavage methodology (PRCC).
Experimental results obtained after cleavage of the single-reduced/cyanylated
isoforms of RNase-A using ammonia and methylamine are presented to show
the feasibility of using a reagent containing both nucleophiles to produce pairs of

labeled peptides that facilitate recognition of cleavage products A and B.

2.- MATERIAL AND METHODS

2.1 Chemicals

Guanidine hydrochloride was obtained from Boehringer Mannheim
Biochemichals, (Indianapolis, lN); 1-cyano-4-dimethyiaminopyridinium
tetrafluoroborate (CDAP) and tris-(2-carboxyethyl)phosphine (TCEP)
hydrochloride were purchased from Sigma (St. Louis, MO). Water grade I [13] or
better was obtained from a Millipore system. Methylamine, was acquired as the
hydrochloride salt from Aldrich. Ammonia was obtained from EM Science as an
aqueous solution containing approximately 14.4 M ammonium hydroxide. Bovine
pancreatic ribonuclease A type Ill-A (RNase-A) was purchased from Sigma (St.

Louis, MO).

148

2.2 Mass spectrometry

MALDI mass spectra were acquired on a Voyager DE-STR time-of-flight
(TOF) mass spectrometer (Perkin-Elmer Biosystems inc.) equipped with a 337-
nm nitrogen laser. Measurements were made in linear mode with the
accelerating voltage set typically to 25,000 V with grid voltage at 95%, guide wire
at 0.05%, and extraction delay time at 150 nsec. Time-of-flight to mass
conversion was achieved by external and/or intemai calibration using standards
of bradykinin (MH’average = 1061.22 Da), bovine pancreatic insulin (MH‘average =
5734.56 Da), horse skeletal myoglobin (MH‘average = 16,952 Da), and/or
cytochrome c (MH‘awerage = 12,361.1 Da) obtained from Sigma (St. Louis, MO).
Samples were prepared by spotting 0.5 uL of analyte + 0.5 uL of matrix solution
onto a stainless steel sample plate with a-cyano-4-hydroxy cinnamic acid (from
Sigma Chemical Co., St. Louis, MO) as the matrix (10 mg/mL) in the modified
thin-layer method as described by Cadene and Chait [14] or by the conventional
dried-droplet method. For some experiments, samples were desalted using Zip-

Tips from Millipore (Billerica, MA).

149

2.3 Chromatography

HPLC was used for separation of mixtures using a Waters 2695
separation module coupled to a Waters 2487 Dual A Absorbance detector at 214
nm. The column for purification of isoforms was a C4 (Vydac # 214TP54, 5-pm

particle size, 300 A pore, 4.6 x 250 mm) and for p—HPLC separation of reaction

mixtures, the column was a 018 (Vydac # 218TP5115, 5-pm particle size, 300-A

pore, 1.0 x 150 mm). HPLC fractions were collected manually and, if necessary,

dried for further use.

2.4 Partial Reduction of RNase-A

The protein sample was placed in a vial where the partial reduction was
carried out by adding TCEP in 0.1 M citrate buffer at pH 3.0 for the cysteine
content and keeping a concentration of 4 nmol/pL (e.g., 40 nmol of TCEP was
reacted with 10 nmol of RNase-A in a total volume of 10 pL ) followed by

incubation at room temperature for 15 min.

2.5 Cyanylation and isolation of single-reduced isoforms of RNase-A

A 20-fold molar excess of 0.1 M CDAP (in 0.1 M Citrate buffer at pH = 3.0)
over the total cysteine content was added to the mixture of partially reduced

isoforms for the cyanylation step at room temperature for 15 min. Single-reduced

150

and cyanylated species were purified using HPLC, and then dried. A 6M
guanidine hydrochloride solution was used to reconstitute the sample in order to
obtain solutions between 0.3 and 1 nmol/pL. Accurate concentration was
determined, if necessary, by injecting an aliquot of this solution into a u—HPLC
system, and comparing the response to that of standards. After partial reduction
and cyanylation, single-reduced/cyanylated isoforms of RNase-A were isolated
using reverse phase-HPLC. Because the cyanylated isoforms elute after the
intact protein (lP), we refer to them according to their elution order (IF1, lF2, IF3,

lF4).

2.6 Cleavage of single-reduced and cyanylated RNase-A isoforms

The nucleophile solution was made using water grade I [13] or better.
Preformed single-reduced cyanylated isoforms of RNase-A (~ 1 nmol in 1 pL of
6M guanidine) were treated with a given nucleophile or with a mixture of
nucleophiles under optimized experimental conditions (2 °C, 10 min) of

concentration, temperature, and reaction time.

2.7 Quenching of the cleavage reaction

After the cleavage reaction was allowed to proceed for the specified time,

it was quenched by adding citric acid to establish pH = 3, followed by adding

151

0.1 M TCEP solution and incubating at 37 °C to completely reduce the remaining

disulfide bonds.

3.- RESULTS AND DISCUSSION

3.1 Application of “external” labeling to locate CN-induced cleavage

fragments

The simplest approach for the use of two nucleophiles to identify CN-
induced cleavage fragments during the cyanylation-mass mapping methodology
can be named as “extemal” labeling. it consists of the treatment of aliquots of a
sample with different nucleophiles, for example, ammonia and methylamine in
separate experiments. After the cleavage reaction takes place, it is quenched,
and the protein is totally reduced prior to acquiring the mass spectra. Thus, in
order to identify the cleavage fragments and consequently infer the disulfide
linkage, the two individual spectra are examined for mass spectral peaks
representing a specific mass shift, depending on the mass difference between
the nucleophiles used. This methodology can be slightly modified by doing what
follows: after the cleavage reaction and quenching, samples can be combined in
order to obtain a single mass spectrum where the presence of pairs of labeled
peptides is easily visualized. This is because there is a 14-Da mass difference in
mass spectra for single-charged species; an appropriate algorithm could detect

pairs of these labeled peptides easily.

152

Another variation to this approach is the in silico comparison of mass
spectra of samples individually treated with the different nucleophiles. The
computer would process and compare mass spectra that have been obtained
individually. This approach is useful because pairs of labeled peaks can be
visualized in a single spectrum even though they come from physically different
experimental samples. One example of this approach is given in Figure 60,
where MALDI mass spectra for one of the isoforms for a-lactalbumin were
combined in silica in order to detect cleavage fragments. In this case, the single-
cyanylated isoform after reduction of the disulfide bond between cysteine 73 and
cysteine 91 was split into three aliquots, and they were treated with ammonia,
methylamine, and ethylamine independently. The MALDI spectra were
combined in silica, and cleavage fragments were identified. Figure 60 shows an
expansion of the region between m/z 8200 and 8400 where mass spectral peaks

corresponding to the cleavage fragment A for this isoform are labeled.

153

 

3333

   

a
16 i
15 f
3. 14 z"
'5 13 5
5 12 :- 1
E 11 j f
10 I "
3 e ”I i.
7 I
O: 6 (g
5
4
3 A
2 f / ... -~
1
o

 

I ‘ I ' I ' I ' I ' I ' I ' I ' I ' I ' I
8200 8220 8240 8200 8280 8300 8320 8340 8300 8380 8400

WI

 

 

 

Figure 60. Expanded section of an overlay of three mass spectra processed in silico
for one isoform of a-lactalbumin. Ao represents the cleavage fragment produced
using ammonia as nucleophile, A14 when methylamine was the nucleophiles and A23
in the case of ethylamine.

We observe that the relative intensity of these peaks is very low (less than
10%), which would give us a high uncertainty in determining if these peaks are
real signals from cleavage fragments if only one spectrum in a single experiment
had been analyzed. However, in analyzing the combination of the three MALDI
mass spectra, we are confident about assigning these signals to the A cleavage
product coming from this particular isoform. in the spectrum, we observe that the
intensity of the peaks is not similar. Ideally, peaks should have the same
intensity, but this depends on the cleavage reaction and the MALDI-sample

preparation. That is why we next looked for a single reagent that could produce

154

a pair of labeled cleavage fragments of similar peak intensity in a MALDI mass

spectrum.

3.2 Optimization of the conditions for a combined ammonia-

methylamine reagent for “intemai” labeling

We define “internal" labeling for the CN-induced cleavage fragments as
the use of a mixture of two nucleophiles as the cleavage reagent in order to
produce a mixture of labeled cleavage fragments A and B in a single mass
spectrum. The cleavage reaction takes place in one single sample and
recognition of cleavage products A and B (see Figure 59) is especially facile
when the concentrations of the homologous nucleophiles are adjusted to allow

detection of equally intense peaks (relative intensity ratio a rir = 1) as is shown in

 

 

 

 

 

Figure61.
A
100- 0A14
80-
;- Am/z A14’Ao= 14 units
g 60-
IE.
5 ‘ '
1.i ‘0‘ fif= I “as 1
1: ‘ A0
20d
.l
o ' t fu’; ' I V I 'Mfi‘fifli " I V I
5600 5650 5700 5750 5800 5850 5900 5950 8000
ml:

 

 

 

Figure 61. Ideal relative intensity ratio (rir) close to unity is represented for a MALDI
signal for a pair of labeled cleavage fragments produced by treatment of the sample
with a mixture of ammonia and methylamine.

155

in order to select the best nucleophile mixture based on the relative
intensity ratio (rir) for the labeled cleavage products identified in MALDI mass
spectra, singly reduced and cyanylated isoforms of RNase-A were chosen as
model compounds. Taking into account the experimental conditions of
temperature and reaction time (2°C, 10 min) previously evaluated for ammonia
and methylamine, mixtures of these nucleophiles were evaluated for the

production of cleavage products.

An aliquot of a stock solution of a single-reduced/cyanylated isoform was
treated with a solution containing different methylamine/ammonia concentration

ratios (see Table 6) under the optimized experimental conditions.

Table 6. Nucleophile
concentrations for the
reagent mixtures tested.

 

 

 

 

 

 

 

 

 

 

 

 

[NI-13] [Mom-1,]
mollL mollL

a 5 0

b 5 0.14

c 5 0.21

d 5 0.28

e 5 0.35

f 5 0.42

g 5 0.49

 

Cleavage products A and B were identified in the MALDI mass spectra
(see values in Table 7) and relative intensity ratios (rir) for the labeled cleavage

products A and B were calculated from the obtained MALDI-MS spectra.

156

Table 7. Calculated m/z values for the protonated cleavage
products using methylamine or ammonia as nucleophiles.

 

A0 A14 Bo B14 c
lF1I7082.0 7096.0 788.9 802.9 5906.6
llF2 2704.9 2718.9 6546.5 6560.5 4526.1
lF3I44‘l1.9 4426.0 6062.8 6076.8 3302.8
.lF4 6351.26365257665 5780.5 1659.8

 

 

 

 

 

 

Figure 62 shows the rir values for the labeled cleavage products of the
different RNase A isoforms. Dashed lines show the ideal rir value (= 1). We
observe that the rir for the labeled cleavage products reaches the value of 1 at
slightly different nucleophile mixtures (depending on the isoform treated)
because of the different accessibility of the cyanylated cysteines during the

cleavage reaction.

 

 

 

 

 

 

 

 

 

 

 

 

 

‘- 4

:1- ............. i .-.-r.-.‘_.-‘._.-‘— ........

01—4 7—
" qt. ................. ....-.v-._>._._b .........
I 0‘ % p
E,‘ :_ .................. . -.-.‘._.-..._._.. .... ‘ .....
§ 04—. ’ - lF1_A_7095
i 4: V V y e IF2_A_2705
g 1 ............ -'._.+.-.¥ .................. - A lF2_B_6546
. 0 7 ‘ v IF3_A_4413
g - A e IF3_B_6062
5 1.. ............. ‘._._‘_._.z......L .............. ‘ IF4_A_6353
.3 0 ‘ .5 '— > IF4_B__5767
. .1
n- «I
o 1 O 0
a 4 ............. .._.+.-.9 ...................

c:_ ............. ;.-.—r.—.‘._._..._._I_._.2.—.-

0 7 V I ' I I ' I I j

00 0.1 0.2 0.3 04 05
[NHJ = 5M [MeNHzL M

 

 

 

Figure 62. Relative intensity ratios for labeled cleavage products A and B for RNase A
isoforms.

157

The graphed results indicate that a concentration ratio of 0.3M/5M for
methylamine/ ammonia is a good compromise to obtain a rir of unity, which can
can be corroborated by observing in detail the MALDI mass spectra under those

experimental conditions.

Figure 63 and Figure 64 show the labeled A and B cleavage products for
each of the RNase-A isoforms. From mass spectra, it is evident that the use of
primary amines as cleavage reagents facilitates determination of the cysteine

linkage in cystinyl proteins beacuse cleavage fragments are easily identified.

158

 

100-
80-

00-

40- A0

Relative Intensity

 

 

 

 

5500 ' 0&0 7'0'00 ' 7500
ml:
100- C
. BO
80-
‘ \ A14 B14
80- A0

I.

d
20-

Relative intensity

 

    

 

 

 

 

 

‘ JIL I
o.r.‘I_.I‘*7—_-‘L_* f“ 1“ ' ‘ _ I“ I ,
3000 4000 5000 6000 W00

m/z

 

 

 

Figure 63. Top: Mass spectrum of labeled cleavage products for IF1. Bottom: Mass
spectrum of labeled cleavage products for lF2.

159

 

3
a
1
0
in

4,
3 I
I

 

 

 

 

 

 

 

\ I
\ I,Bo
II
80- HIM/B14
g .
a ...,m/I I»...
8 604
5 .
C
g 40-
C
m 1
20-1
0 JL ' bk“ I 4[
3000 4000 V 5000 600‘
m/z
43
a
C
O
E .
3
5 ‘~
. \
1:

 

 

 

 

 

ML 14 44. bulb-3‘41 I‘
o ' I V I v I ' I V“ I V l ' I

2000 3000 4000 5000 6000 7000 8000
ml:

 

 

 

 

Figure 64. Top: Mass spectrum of labeled cleavage products for IF3. Bottom: Mass
spectrum of labeled cleavage products for |F4.

160

4.- CONCLUSION

Mixtures of methylamine/ammonia as nucleophiles produce two different
species for cleavage products A and B during the cleavage reaction of a

cyanylated isoform.

Yields of the species produced during the cleavage reaction containing the
nucleophile group —NH2 or —NH—CH3 depend mainly on the nucleophile
concentration in the mixture even though there is an influence of the protein

structure.

Methylamine is a better nucleophile than ammonia.

A concentration ratio for methylamine/ammonia of 0.3M/5M represents a

good compromise to obtain a mixture of cleavage products A and B with MALDI

mass spectra showing an intensity ratio equal to unity for species containing

nucleophile groups —NH2 and -NH—CH3.

161

5.- REFERENCES

[1] A. Pandey, M. Mann, Proteomics to study genes and genomes, Nature 405
(2000) 837-846.

[2] G. Cagney, A. Emili, De novo peptide sequencing and quantitative profiling of
complex protein mixtures using mass-coded abundance tagging, Nat. Biotechnol.
20 (2002) 163-170.

[3] A. Shevchenko, I. Chemushevich, W. Ens, K. G. Standing, B. Thomson, M.
Wilm, M. Mann, Rapid 'de novo' peptide sequencing by a combination of
nanoelectrospray, isotopic labeling and a quadrupole/time-of-fiight mass
spectrometer, Rapid Commun. Mass Spectrom. 11 (1997) 1015-1024.

[4] J. K. Eng, A. L. McCormack, J. R. Yates, ill, An approach to correlate tandem
mass spectral data of peptides with amino acid sequences in a protein database,
J. Am. Soc. Mass Spectrom. 5 (1994) 976-989.

[5] S. P. Gygi, B. Rist, S. A. Gerber, F. Turecek, M. H. Gelb, R. Aebersold,
Quantitative analysis of complex protein mixtures using isotope-coded affinity
tags, Nat. Biotechnol. 17 (1999) 994-999.

[6] W. Wu. Applications and computational considerations of the cyanylation-
based disulfide mapping methodology for cystinyl proteins, PhD Thesis, Michigan
State University, East Lansing, MI, December 2003.

[7] J. J. Gorman, T. P. Wallis, J. J. Pitt, Protein disulfide bond determination by
mass spectrometry, Mass Spectrom. Rev. 21 (2002) 183-216.

[8] T. P. Wallis, J. J. Pitt, J. J. Gonnan, Identification of disulfide-linked peptides
by isotope profiles produced by peptic digestion of proteins in 50% (18)O water,
Protein Sci. 10 (2001 ) 2251-2271.

[9] T. P. Wallis, C. Y. Huang, S. B. Nimkar, P. R. Young, J. J. Gorrnan,
Determination of the disulfide bond arrangement of dengue virus NS1 protein, J.
Biol. Chem. 279 (2004) 20729-20741.

[10] J. Wu, J. T. Watson, A novel methodology for assignment of disulfide bond
pairings in proteins, Protein Sci. 6 (1997) 391-398.

[11] J. Qi, W. Wu, C. R. Borges, D. Hang, M. Rupp, E. Torng, J. T. Watson,
Automated data interpretation based on the concept of "negative signature mass"
for mass-mapping disulfide structures of cystinyl proteins, J. Am. Soc. Mass
Spectrom. 14 (2003) 1032-1038.

162

[12] W. Wu, W. Huang, J. Qi, Y. T. Chou, E. Torng, J. T. Watson, 'Signature
sets', minimal fragment sets for identifying protein disulfide structures with
cyanylation-based mass mapping methodology, J. Proteome Res. 3 (2004) 770-
777.

[13] ASTM D 1193-99e1, ”Standard Specification for Reagent Water", ASTM
lntemational.

[14] M. Cadene, B. T. Chait, A robust, detergent-friendly method for mass
spectrometric analysis of integral membrane proteins, Anal. Chem. 72 (2000)
5655-5658.

163

CHAPTER 4. DEVELOPMENT OF AN ON-LINE SYSTEM FOR THE
CYANYLATION-BASED DISULFIDE MASS MAPPING

1.- INTRODUCTION

1.1 On-line methods In analytical chemistry and proteomics

During many years, before and during the genomics and proteomics era,
an increasing number of publications demonstrated the potential of column-
switching techniques for the determination of drugs in biological and
environmental samples using high performance liquid chromatography (HPLC).
The use of these systems greatly facilitated the sample preparation step,
providing better precision and accuracy in quantitative analysis and the possibility
of time reduction in analysis [1-7]. in the simplest setup, a small column (trap) is
located between the injection port and the analytical-HPLC column by means of
a switching valve as illustrated in Figure 65. Two positions are available for this
system. For the first position, the sample is loaded into the trap and the
analytical column is conditioned and equilibrated at the same time. In the trap,
the analytes are pre-concentrated, cleaned up and, if necessary, derivatized.
Different washing steps with different solvents are carried out, if necessary, in
order to remove impurities or undesirable analytes. However, one should be
careful during this process to avoid eluting the analytes of interest during those

steps. After analytes are ready to be eluted, the system is changed to position 2,

164

and the solvent gradient is started in order to desorb analytes from the trap,
carrying out the separation by means of the analytical column (see Figure 65).
Traps are filled with the appropriate stationary phase (resin) according to the
analysis that is carried out. The most commonly used resins are reverse phase
and ion exchange particles; however, some recent research suggested that the

use of immunosorbents is a good choice for analysis of small molecules [8].

 

Position 1: Load and Cleaning up

 

=43:
.9

fl

—

 

 

  

 

 

 

 

 
   

Analytical

 

 

 

 

... .. column Detector
.9 2‘. ' . \.
injection 3 4 Waste
valve
Pumps Switching valve

Position 2: Separation

 

.
=5."-
I...
'

 

 

 
 

 

 

   
 

Analytical
column Detector

 

 

 

 

 

5
==:::: . ,6
I9. 0 ‘
' ‘ injection 3 4
valve
Pumps Switching valve

 

 

 

Figure 65. Simple on-Iine system for cleaning up and concentrating of samples. The system has
two positions; read the text for details.

165

Automated HPLC systems were also used before the advent of biological
mass spectrometry; anion-exchange columns were combined in tandem with
reversed-phase columns for peptide mapping of very large proteins and the
analysis of extremely complex peptide mixtures [9, 10]. Numerous reports exist
concerning the use of tandem columns in liquid chromatography to address
specific separation problems using column switching arrays or multidimensional
chromatographic systems. In two-dimensional (ZD) HPLC (2D-HPLC), the first
column is filled with a stationary phase and it is operated under specific gradient
conditions and separation mechanism, as the first dimension separation. Effluent
from this first column is delivered into a second column for the second separation
using a different mechanism and a different gradient program from the first
separation (second dimension separation) [11-16]. The differences among
several 2D-HPLC are based on the type of stationary phases used (separation
mechanism) and/or the way columns are connected by means of switching and
injection valves. An on-line two-dimensional HPLC-system is shown in Figure
66. This on-line system includes a preparation step with a silica-based restricted
access material (RAM) with ion exchange functionalities. The separation is
carried out in two dimensions by means of an ion exchange (IEX) column (first
dimension) and four reverse phase (RP) columns (second dimension separation)
that are used alternately as the fractions are eluted from the first separation. In

this system, final fractions are collected for subsequent analysis.

166

 

Injector

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Pradon collection A Fraction collection 8
for MALDI MS analysis for MALDI MS analysis

 

 

 

Figure 66. Schematic representation of an on-line system with sample preparation and two-
dimensional HPLC. From reference [14].

167

Another interesting and quite novel strategy in liquid chromatography (LC)
is called “thermally tuned tandem column (T3C)”. T3C can be considered to be
an optimal combination of adjusting selectivity by varying stationary phase type

and temperature [17-19].

in the genomics and proteomics fields, switching and multi-dimensional
separation alternatives have been used to couple separations of enzymatic
digestions with mass spectrometry [12, 13, 20]. The necessity of working with
small sample amounts have provoked the development of scaled down
separation systems in order to detect peptide and protein quantities in the range
of pico-, femto- and atto-moles. The use of membranes to immobilize proteins
has also been developed for on-line systems applied in proteins [21-24]. The
use of on-line digestions for proteins coupled on-line to mass spectrometry
analysis is becoming more common, thereby improving the sensitivity of LC-

hydrogen exchange-mass spectrometry experiments [24-28].

1.2 On-line method for the Partial Reduction-Cyanylation-CN induced

Cleavage (PRCC) Mass Mapping of Cysteinyl Proteins

Disulfide linkages play an essential role in protein tertiary structures.
Changes in the disulfide bond structure of a protein may result in important
changes of activity and function. The Partial Reduction-Cyanylation induced

Chemical Cleavage for mass mapping of cysteinyl proteins is a methodology that

168

is used as a batch procedure, called the ‘in-vial’ approach, in which sequential
reactions are carried out in a small vial [29]. During these steps, it is necessary
to remove reagents and salts that have been added for one reaction before
reagents can be added for the subsequent reaction. This chapter describes two
different systems that were analyzed as possible analytical instrumentation
arrangements for the PRCC method. In the first approach, only one small
column, filled with a resin, was used for binding the protein and to cany out all
chemical steps of the PRCC methodology. In the second approach, the on-line
PRCC hardware consists of three components connected in series: T1-R-T2; a
reactor (R) and two traps (T1, T2) containing polymeric resin. In this PRCC on-
line system, the protein is loaded and subsequently chemically degraded in a
controlled manner to give structurally diagnostic cleavage products, which are
collected and analyzed using mass spectrometry (MS) in order to establish the
cysteine linkage. The sample is loaded onto T1, where it is partially reduced and
cyanylated. Then, cyanylated species are transferred to R where cleavage
occurs in a continuous flow step. Species from R are captured in T2, and after
total reduction, they are eluted for analysis by MS. This second instrumental
configuration is the one recommended in the conclusions after a complete
evaluation of the performance system, where the overall analysis time is reduced
and the sample manipulation is minimized to help minimize sample loss. During
development of the on-line PRCC system, the integrity of each physical or
chemical step was tested with standard quantities of model compounds, primarily

using ribonuclease A (RNase-A) and bradykinin (BDK). A comparison of the

169

general steps that are necessary for in-viai and on-Iine approaches is shown in

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 67.
- ln-vial approach: - On-Ilne approach:
a TCEP
I Partial reduction I v I Load Sample I
’ $ ’ 4 CDAP ’
' r—J— ( TCEP
I Cyanylation I v Partial reduction CDAP
Cyanylation ann or
I HPLC I a d d d Cleavage ( CH3NH2
WWW? Ttale ti TCEP
0 e “C on
Dry [, @ 37°C
i I: * *
I Cl I NH40H 0" Elution of
. eavage V" CH3NH1 cleavage
products
| Dry l i
‘L TCEP I MALDI-MS |
Total '59 @ 37°C
Reduction
I MALDI-MS I

 

 

 

 

Figure 67. Comparison of the in-vial and on-line steps for the Partial Reduction Cyanylation and
Cleavage methodology.

2.- MATERIAL AND METHODS

2.1 Chemicals and Enzymes

Guanidine hydrochloride was obtained from Boehringer Mannheim

Biochemichals, (Indianapolis, IN); 1-cyano-4-dimethylaminopyridinium

170

tetrafluoroborate (CDAP) and tris-(2-carboxyethyl)phosphine (TCEP)

hydrochloride were purchased from Sigma Chemical Co. (St. Louis, MO).
Water grade I [30] or better was obtained from a Millipore system.

Methylamine was acquired as hydrochloride salt from Aldrich. Ammonia
was obtained from EM Science as an aqueous solution containing approximately
14.4 M ammonium hydroxide. Bovine pancreatic ribonuclease A type ||l-A
(RNase-A, 8 cysteines, 4 disulfide bridges, 124 amino acid residues) and
bradykinin (BDK, NO cysteines, 9 amino acid residues) were purchased from

Sigma (St. Louis, M0).
2.2 Columns and reactor for the cleavage reaction

Empty microbore-HPLC columns (1.0 mm ID x 2 cm; V ~ 15.7 uL) made
of PEEK©Istainless steel and narrowbore short-HPLC columns (2.0 mm ID x 2
cm; V ~ 62.8 pL), made of stainless steel from Upchurch (Oak Harbor, WA),
were used to prepare lab-made traps (T) with different types of stationary

phases. An empty analytical column was used as reactor (R) for the cleavage

reaction (4.6 mm ID x 25 cm; V ~ 4.15 cma).

171

2.3 Stationary phases

PRP-3 polymeric reversed phase (pore diameter = 300 A) was purchased
from Hamilton (Reno, NV). Poros R2 was obtained from PerSeptive Biosystems
(Framingham, MA); it is a polymeric stationary phase developed for perfusion
chromatography. C4 and C18 silica-based stationary phases were used from
available resins in the laboratory. A polymeric cartridge from Michrom was used

also.
2.4 Preparation of traps with stationary phases

Empty traps were filled using a setup similar to the one shown in Figure
68. A slurry (usually 50:50 v/v methanol/water) containing the selected resin was

deposited in the funnel and the vacuum carried the resin down the column (trap).

Funnel ’
I

 

Trap

Connector ,
l

' Rubber stopper

Vacuum <— ;

Vacuum flask

 

 

 

 

Figure 68. Lab-made setup to fill traps
with stationary phase.

172

2.5 Mass spectrometry

MALDI mass spectra were acquired on a Voyager DE-STR time-of-flight
(TOF) mass spectrometer (Perkin-Elmer Biosystems Inc.) equipped with a 337-
nm nitrogen laser. Measurements were made in linear mode with the
accelerating voltage set typically to 25,000 V with grid voltage at 95%, guide wire
at 0.05%, and extraction delay time at 150 nsec. Time-of-flight to mass
conversion was achieved by external and/or internal calibration using standards
of bradykinin (MH",,,,.arage = 1061.22 Da), bovine pancreatic insulin (MH‘average =
5734.56 Da), horse skeletal myoglobin (MH"average = 16,952 Da), and/or
cytochrome c (MH‘avemge = 12,361.1 Da) obtained from Sigma Chemical Co. (St.
Louis, MO). Samples were prepared by spotting 0.5 pL of analyte + 0.5 pL of
matrix solution onto a stainless steel sample plate with a-cyano-4-hydroxy
cinnamic acid (from Sigma Chemical Co., St. Louis, MO) as the matrix (10
mg/mL) in the modified thin-layer method as described by Cadene and Chait [31]

or by the conventional dried-droplet method [32, 33].

173

2.6 Partial Reduction, Cyanylation and Isolation of Single-Reduced

lsoforms of RNase-A

During evaluation of the online system for the PRCC methodology, it was
necessary to work with single-reduced or single-reduced/cyanylated isoforms.
Their preparation was according to the description in Chapter 2. Accurate
concentration was determined, if necessary, by injecting an aliquot of standard
solutions into u—HPLC or narrowbore-HPLC systems and comparing the peak
areas in the corresponding chromatograms versus areas in a calibration curve.
Because the cyanylated isoforms for RNase-A elute after the intact protein (lP),

we refer to them according to their elution order (IF1, lF2, IF3, IF4).

2.7 Chromatography

HPLC was used for separation of mixtures using a Waters 2695
separation module coupled to a Waters 2487 Dual 7L Absorbance detector at 214
nm. The column for purification of isoforms contained a C4 stationary phase
(Vydac # 214TP54, 5-um particle size, 300 A pore, 4.6 x 250 mm), the column
for narrowbore-HPLC separation of reaction mixtures contained a C18 sationary

phase (Vydac # 218TP3215, 3-pm particle size, BOO-A pore, 2.1 x 150 mm) and

the column for p-HPLC separation of reaction mixtures contained a slightly

different C18 stationary phase (Vydac # 218TP5115, 5-pm particle size, SOD-A

174

pore, 1.0 x 150 mm). HPLC fractions were collected manually and, if necessary,

dried for further use.

3.- RESULTS AND DISCUSSION

3.1 One-column approach for the on-line PRCC methodology

3.1a Proposed hardware for the one-column on-line system

This approach (see Figure 69) consisted of the use of only one trap (trap

1, T1) filled with a reverse stationary phase; it is connected on-line to a pumping
system and a detector. The protein is injected into the system by means of the
injection valve, loaded into T1, and bound to the stationary phase; reagents are
added consecutively to carry out the chemical derivatization. Washing steps for
removing of the chemical reagents are monitored through the UV-Vis detector.
After total reduction (the last step), T1 can be connected to a HPLC system for
partial separation of the cleavage products. Chromatographic fractions can be
collected for a later analysis by MALDI-MS or can be analyzed by coupling a MS
system online. In summary, the steps to carry out the PRCC methodology using
this setup are:

1) Load sample on T1

2) Add TCEP for partial reduction

3) Washing step

175

4) Add CDAP for cyanylation

5) Washing step

6) Add Cleavage reagent

7) Washing step

8) Add TCEP for total reduction

9) Elute and collect cleavage products from trap using a HPLC
coupled to T1.

10) Analysis using Mass Spectrometry (MALDI/T OF)

 

 

    

    
 
   

A) _
Injection m
valve UV'VIS ‘4 Sample Acquisition

Pumps ’ Detector . Collection Data System
Acquisition
Data System
,, fi ,
% TRAP 1
detector Sample
Pumps collection

 

 

 

Figure 69. Single-trap approach for an on-line method for PRCC methodology. The
instrumental setup in A) is used for the PRCC methodology (steps 1-8), after this, T1 is
connected to a HPLC system (as in B) for elution and collection of cleavage products (step 9).

176

3.1b Evaluation of the one-column on-iine approach

The first issue to be solved was the selection of the packing material for
the trap 1; silica-based stationary phases dominate the field of separation
science because of the high level of reproducibility, selectivity and sensitivity they
can provide. However, these types of stationary phases are only hydrolytically
stable over the pH range ~ 285 [27, 34, 35]. Thus, even though the use of C4
or C18 silica-based stationary phases had been shown to be suitable for the
partial reduction and cyanylation of a cystinyl protein bound to the resin, it is
necessary to select yet another stationary phase, because during the cleavage
step, the pH is higher than 8.0. The inherent stability of PRP-3 columns (a cross-
linked poly(styrene-divinyl benzene) polymeric stationary phase as illustrated in
Figure 70) allows their use for any protein purification method in extreme pH-
conditions from pH 1 to 13 (there is no silica to dissolve at pH higher than 8 or
siloxane bonds in the C18 groups to hydrolyze at pH lower than 2) [36]. The
PRP-3 support is pressure-stable up to 5,000 psi, the crosslinking prevents
shrinkage or swelling when the mobile phase is changed. This kind of resin has
been used successfully for recovering relatively non-polar analytes in on-line
systems used in environmental chemistry [3, 37]. This PRP-3 resin is very
helpful for separation of proteins that are unstable or have poor solubility at a pH
below 8.0. While most proteins are best chromatographed at pH 2.0, it is good to
know that they can be separated at elevated pH, if the need arises. Thus, PRP-3

columns allow processing of proteins at low and high pH extremes.

177

 

 

 

 

 

Figure 70. Chemical structure of
PRP-3: poly(styrene-divinyl)benzene
resin

According to the manufacturer, the 300—A pores on the PRP-3 support

help ensure good recovery (>90%) of the protein, and the highly inert
poly(styrene-divinylbenzene) packing enhances protein recovery because there

are no silanol groups on the support to cause irreversible protein adsorption [38].

Initial screening experiments showed that it was feasible to carry out
partial reduction and cyanylation of the cystinyl protein (RNase-A) as it was
bound to the PRP-3 resin, thus, it was decided also to evaluate the chemical
cleavage reaction in the presence of the pRP-3 resin. Pre-formed single-
reduced/cyanylated species of RNase-A (IF1... IF4) were loaded into T1;
cleavage reagent (in this case, 1 M ammonia) was added to the trap and held
there for the designated reaction time(s). After this, the excess cleavage reagent
was washed off, reducing agent was added for total reduction, and finally the
cleavage products were eluted from T1 and collected to be analyzed by MALDI-
TOF MS. Results from screening experiments for the cleavage reaction showed

that for lF1. lF2, and IF3, the cleavage reaction took place and cleavage

178

fragments were detected easily. The MALDI-MS spectrum for IF3 cleavage
products as produced on-line under the described experimental conditions is

shown in Figure 71. Mass spectral peaks for cleavage products are easily

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

discernible.
100- c “=3
1 Cys 40-95
90- Cleavage
1 product figment m/z
80- 1-39 4413.9 A
1 11240-94 6063.6 3
g. 704 Itz-95-124 3302.7 c
99 1 p (1-94) 10399.5
g 60- pun-40124) 9288.3
E q
z 50-
.2 1
E 404
O 4
“5 30-
20-
10-
0 . . . .
3000 3500 4000 4500 5000 5500 6000 6500
m/z

 

 

 

Figure 71. MALDI-MS spectrum for lF3 of RNase-A after being loaded into a small column
packed (Trap1) with PRP-3 resin.

However, in the case of IF4, disappointing results were obtained. In
Figure 72, the MALDI-MS spectrum is shown for on-line and in-vial approaches.
It turned out that in using the on-column approach, two of the cleavage products
could not be detected. Furthermore, none of the intact isoform could be detected

efihen

179

 

1658.70

 

 

 

 

 

  

100-
1
16603 ”I
100- ' 301
90.! g. 70:
1 E 60-
80‘ g 1
, ; 50-
z 4
g! 70‘ E 401
Q ‘ O 4
°‘ 30 \ /
g 60: I 576912 *635564
C 20. . .
3 50‘ ,0; t 1
.2 I i * * I
E 40‘ 0 $19“. .Lcr . . we
3; 30* 2000 3000 4000 5000 6000
q
mix
a *
20‘ 3
104 I
d

 

1
2000 3000 4000 5000 6000

m/z

 

 

 

Figure 72. MALDI-MS spectrum for lF4 of RNase-A obtained using in-vial or on-column
approaches. During the in-vial approach, two of the cleavage fragments (pointed out with arrows)
are not observed. Peaks corresponding to i) cleavage fragments of minor cyanylated isoforms
that coeluted with lF4 or B-elimination products are designated with (*) and ii) those some

unknown side-reaction products are designated with t.

Table 8 shows a list of a series of experiments that were carried out in
order to evaluate: 1) whether the cleavage products might be irreversibly
adsorbed to the stationary phase in T1 and 2) other experimental conditions for
the on-line cleavage reaction. For the first question, preformed cleavage-
products coming from lF4 were loaded into trap 1. T1 was washed with water
(0.1% TFA), and cleavage products retained in the trap were eluted using an
ACN/water mixture. Eluate was analyzed by MALDI-MS. From MALDI-MS

spectra, it was possible to identify all the cleavage products previously loaded

180

into T1. Thus, according to these results, there was not an irreversible binding of

the IF4-cleavage products to the stationary phase.

Table 8. Checking list of experimental conditions tested for cleavage of lF4 on-column.

 

a) Evaluate Interactions Resin-Cleavage Products

‘/ Load and elution of pre-formed cleavage products
(Residence time: 1 hr and overnight)

b) Cleavage Reaction Variations

X 1 M NH, in 6 M Guanidine x 1 hr @ room temperature
X 1 M NH, In 6 M Guanidine x 1 hr @ 37°C

1 M NH, in 6 M Guanidine x 1 hr @ 50°C

1 M NH, in 1 hr @ room temperature

3.7 M NH, In 6 M Guanidine x1 hr @ room temperature
Concentrated NH, x 1 hr @ room temperature

3.7 M NH, in 6 M Guanidine x 1 hr (using sonicator)

XXXXXX

1 M NH, in 6 M Guanidine x 1 hr @ room temperature
(Flow rate aprox. 0.02 lemln)

X

1 M NH, in 6 M Guanidine in presence of a surfactant

 

 

 

For the other experiments, no improvement was seen in the production of
cleavage products from IF4. Analyzing the structure of RNase-A, using a Kyte &
Doolittle plot for hydrophobicity of the protein [39], it is evident that cysteines that
form the disulfide bond for IF4 are located in regions of high hydrophobicity in the
protein (see Figure 73). Hydrophobic interactions retain the protein on the resin,
and probably affect the way that the protein lies on the stationary phase; these

interactions also might promote secondary reactions to compete with the

181

nucleophilic attack involved in the cleavage reaction and by consequence,
diminish the yield of the expected cleavage fragments, in which case they may

not detected in the mass spectra.

 

|F4110

MIL

n-u'cr'uo-unwg

 

[MI/\‘I m5. r\/\
WWW m

":22“ "=1 72

 

5° man 10°

 

 

 

Figure 73. Kyte & Doolittle hydrophobicity plot for designated single reduced isoforms of RNase-
A. The location of the cysteine contained in each isoform is indicated after the numeral identifying
the isoform.

Thus, after the analysis of results obtained for the possible cleavage of the
cystinyl protein adsorbed in a polymeric stationary showed this reaction was not
always carried out efficiently, and it was necessary to think about a different
approach for the PRCC methodology on-line. In this second approach the
protein is desorbed from the trap, cleaved in solution, and trapped in a second

trap. The next section describes and evaluates this alternative approach.

182

3.2 Two-column approach for on-line PRCC methodology
3.2a Proposed hardware for the two-column on-Iine system

This second approach consists of the use of two traps (T1, T2) for the on-
line methodology. Partial reduction and cyanylation are carried out in T1 (see
Figure 74). Cyanylated species are eluted from this trap and they are
transported from T1 to T2 in a stream of the cleavage reagent through a reactor
(R) while the cleavage reaction takes place. Formed cleavage products are
captured by means of T2 where, after washing to remove the cleavage reagent,

the remaining disulfide bonds are reduced by adding a reducing agent (TCEP in

(e i

this case).

 

 

 

 

 

 

 

 

 

 

 

 

Acquisition
=9=§:_ Cleavage reagent Data System
11 mp” ROM“ T2. uv-vrs U
Pumps detector Sample

collection

 

 

 

Figure 74. The two-trap and reactor approach for an on-line method for PRCC methodology.

Without taking into account the washing steps, the sequential steps

necessary to complete the PRCC methodology using this setup are:

183

1) Load sample on Trap 1

2) Add TCEP to Trap 1 for partial reduction

3) Add CDAP to Trap 1 for cyanylation

4) Cleave the cyanylated isoforms during transfer from Trap 1 to
Tmp2

5) Add TCEP for total reduction in Trap 2

6) Elute Cleavage Products from Trap 2

7) Analyze by MALDI-MS

In order to accomplish these steps, it was necessary to evaluate different
parameters for the components involved in the setup: trap 1, the reactor, and trap

2. Parameters that were evaluated are summarized in Table 9.

Table 9. Evaluated parameters to test the feasibility of the two-column approach for developing
an on-line system for the PRCC methodology.

 

TRAP 1 REACTOR TRAP 2
- RESIN SELECTION - CLEAVAGE REACTION ° CAPTURE OF CLEAVAGE
° LOAD CAPACITY EFFICIENCY EFFICIENCY PRODUCTS
0 FEASIBILITY OF CHEMICAL ° TRANSFER EFFICIENCY
REACTIONS: PARTIAL FROM TRAP 1 TO TRAP 2
REDUCTION AND
CYANYLATION
° DESORPTION EFFICIENCY

 

 

 

184

3.2b Trap 2 evaluation: Capture of cleavage products

Trap 2 was evaluated first because if T2 was not able to efficiently retain
cleavage products under the cleavage experimental conditions, it would not have
been possible to use the proposed setup in Figure 74. For T2, we used a
narrowbore short-HPLC column filled with Poros R2 stationary phase. The Poros
R2 stationary phase is a revolutionary product invented by Perseptive
Biosystems for perfusion chromatography. This material consists of cross-linked
poly(styrene—divinylbenzene) flow-through particles with a patented pore size
distribution. This resin possesses large throughpores (6000-8000 A) that allow
analyte molecules to “perfuse” rapidly through the interior of the particles, as well

as through very short “diffusive" pores (500-1500 A) [27, 40].

 

Possible analyte pathways

Throughpore
\5 «if

  

Diffusive pore

 

 

 

Figure 75. Diagram of a perfusion particle.

185

An additional property of Poros R2 is that it offers the lowest backpressure
among the many types of resin in the marketing. Thus, knowing that Poros R2 is
also a polymeric resin, it was decided to use this type of stationary phase for T2.
RNase-A and BDK were chosen as model proteins. BDK was chosen because,
after the chromatographic analysis of all cleavage products coming from the
cyanylated RNase-A isoforms, it was noticed that BDK elutes approximately at
the same time as the least retained cleavage products from RNase-A. Flow rate
and percentage of acetonitrile that allow an effective trapping of cleavage
products in trap 2 were evaluated. T2 was connected according to the diagram
shown in Figure 76. Parameters were evaluated according to the methodology

described next:

1) The cleavage reaction uses ammonia at a high pH, so a mixture of
acetonitrile/ammonia solution was adjusted to the desired proportion of reagents
and it was pumped through T2 following the “load” pathway that is shown in

Figure 76.

186

 

I ELUTION

. . —>
Ifa—fl—fi-fi
' HPLC column

Trap 2 Uv-Vis
LOAD detector

Uv-Vis
detector

 

 

 

Figure 76. Scheme of the instrumental setup used to evaluate the capture of
cleavage products by Trap 2.

2) A volume of 100 mL of sample mixture (1 nmol of RNase-A and 5 nmol
of BDK in a solution of 6 M guanidine) was injected through the injection valve
while the solvent mixture was continuously pumped (continuous flow).
Adsorption of the proteins was followed by means of an UV detector. On

average, the loading time was 15 minutes.
3) Trap 2 was washed with a solution of H20 to remove the excess of the
ACN/NH3, and it was connected on-line to an analytical column through the

“elution” pathway shown in Figure 76.

4) A gradient program of acetonitrile/water was carried out to elute RNase-

A and BDK adsorbed onto Trap 2.

187

5) From the chromatogram, areas under the curve of the peaks
representing the model proteins were calculated and compared with standard

solutions in order to determine the amount of proteins adsorbed in Trap 2.

Thus, following the steps described above, two different sets of

experiments were carried out:

1) Evaluation of the maximum content of ACN: RNase-A and BDK were
adsorbed into Trap 2 using a flow rate of 100 uL/min using a solvent mixture of 0,
5, and 10% of ACN in 1.4 M ammonia. From results obtained here, it was
decided that trapping of the model protein from a solution containing more than
5% ACN was inefficient, and by consequence it could be inferred that, for
cleavage products, a concentration of ACN higher than 5% would result in a poor

adsorption of the protein by T2.

2) Evaluation of the flow rate: Based on results obtained in 1), a solvent
mixture containing 5% of ACN in 1.4 M ammonia was used to test three different
flow rates: 100, 200, and 500 pL/min. Statistical results showed that about 90%

of RNase and 100% of BDK were trapped even at the highest tested flow rate.

A complementary experiment was carried out using a real cleavage

product. The least-retained cleavage product for all isoforms of RNase-A is the

A-fragment 1-25 coming from the cleavage of IF2. This product was purified and

188

loaded into T2 under the experimental conditions that were described above. The
recovery for this cleavage product was calculated at 95.2% (See Figure 77).
From obtained results for the parameters evaluated for T2, we conclude that in
order to obtain a good recovery in T2, it was necessary to use no more than 5%

of ACN with a possible a flow rate up to 500 pL/min.

 

 

 

 

 

0 5 _ From lF2: Cleavage product 1.25
' Recovery: 95.2%
0.4 - -- Standard
‘ "'0 Sample
0.3 -i
a I
<
0.2 -I
0.1 -
0.0 _ I _ : _ i _ ' _ 1
Tim e (m In)

 

 

 

Figure 77. Chromatogram of cleavage product 1-25 from IF2 for RNase-A. The fact that
the peaks for the 'standard' and the 'sample' cleavage product are almost overlapped
indicates that recovery of the sample is close to 100% (in this example, 95.2%).

3.2e Trap 1 evaluation: Resin selection, load capacity efficiency,

feasibility of partial reduction and cyanylation chemical reactions.

In this approach, trap 1 is never in contact with an extremely alkaline or
acidic solution, which is why it is not essential that the resin in T1 be pH-
resistant. Thus, several possible resins for partial reduction were tested under
the same experimental conditions. RNase-A (aprox. 1 nmol) was loaded into

trap 1, and a 0.02 M TCEP in 6 M guanidine/0.1 M citrate buffer pH = 3 solution

189

was injected to the trap. The solution was kept inside the column for 15 minutes
and then trap 1 was connected immediately on-line to a narrowbore HPLC-
column for a chromatographic separation of the partially-reduced species formed
in T1. Chromatograms obtained with different resins were compared to those
relating to the in-vial approach in order to select the most adequate resin for trap

1, as is shown in Figure 78.

 

Intact protein (RNase-A)

”I l ":1 "=2 IF3 IF4

' l 1 1......

 

 

0.3- POI'O. R1

 

 

 

     
 

PorosRZ

 

 

We. (AU)

 

 

 

 

 

 

 

 

0.2 km
Polymer (from Michrom)
0.1 #JJ
«4
0.0 CL
20'2'2'2'4'2'6'2'6'3'0'3'2'3'4'

Tlme(mln)

 

 

 

Figure 78. Comparison of chromatograms of a standard sample of partially reduced RNase-A
as trapped by different stationary phase types. Intact RNase-A and its isoforms (lF1.. IF4)
are labeled.

Because several different phases showed approximately the same
efficiency for the partial reduction reaction, it was decided to use Poros R2 resin
to fill the microbore column T1 in order to take advantage of its polymeric
character and the low backpressure that this support offers, which would have

the benefit of a lower overall pressure in the on—Iine system.

190

In order to evaluate the load capacity efficiency for T1 filled with Poros R2,
a similar instrumental setup to the one shown in Figure 76 (with T1 instead of T2)
was used. RNase-A (~ 1 nmol) and BDK (~ 5 nmol) were loaded under acidic
conditions (pH = 3) at a flow rate of 50 pL/min (using the “load" pathway). In a
second step, trap 1 was connected to the narrowbore column (“elution” pathway);
retained proteins were eluted and chromatograms were obtained. The load
capacity efficiency was evaluated from chromatograms by comparison with the
ones of standard solutions of BDK and RNase-A. As expected from the loading

capacity studies for trap 2, more than 95% of the proteins were retained in trap 1.

In order to optimize the partial reduction of a protein bound to the
polymeric stationary phase, about 1 nmol of RNase-A was loaded and bound to
trap 1, then TCEP was added to the trap at different concentrations. The
reaction was allowed to proceed for specific reaction times; then, chromatograms
of the reaction mixture were obtained at each partial reduction condition and the
profiles were analyzed. Figure 79 shows representative chromatograms of this
analysis in a comparison between the in-vial and the on-Iine approach. A
concentration of about 20 nmol/pL for a reaction time of 15 minutes seems to be
a good starting point for the partial reduction of a protein. It should be pointed
out that using either the in-vial approach or the on-line approach, the
experimental conditions for partial reduction could be modified according to the
protein structure, which determines the accessibility of the reducing reagent to

the disulfide bonds.

191

 

intact protein (RNase-A)

 
   

"=1 IF2 ":3 m

  
 
 
 
 
  

 

 

 

 

 

 

 

 

 

 

0 . 4 .. ln-vial (4 nmoIIpL). 15 min
2: On-column (16 nmoinLI. 15 min
3 0 -2 d On-column (4 nmol/(LL), 20 min
5 ‘ A On-column (4 nmol/(LL), 10 min
0 -° ‘ ' On-column (4 nmol/(LL), 5 min
2 0 f 2's ' 3'0 ' 3'5

 

Time (m In)

 

 

 

Figure 79. Comparison of HPLC chromatograms of partial reduction reaction mixture of 1 nmol
of RNase-A on microbore column filled with Poros R2.

Cyanylation was evaluated in a quite simple manner. Each of the single
reduced isoforms of the RNase-A were loaded, in independent experiments, into
T1. CDAP solution was added to the trap for the cyanylation reaction to take
place for 15 minutes. After that, T1 was connected on-line with the narrowbore-
HPLC column and chromatograms were obtained. In Figure 80, three different
chromatograms are shown. One corresponds to the single reduced IF4 (bottom
panel) without cyanylation treatment, the other two chromatograms correspond to
the same isoform under two different cyanylation treatments. From these results,
it was determined that a solution of 0.2 M CDAP prepared in 6 M guanidine in 0.1

M citrate buffer allows a complete and quantitative cyanylation reaction.

192

 

 
 
   

MA:- 0.1 M CDAP 15'

lF4 standard

30 35 40 45 50 55
Time (m In)

 

 

0.3

 

 

Absorbance (AU)

99?
O-iN

  

 

 

 

Figure 80. Chromatograms for cyanylation on-line with the two-column system. About 1
nmol of RNase-A was treated with CDAP at different concentrations during 15 minutes.
From comparing the chromatograms, it is possible to observe that using a concentration of
0.2 M of CDAP drives the cyanylation to completion.

3.2d Trap 1 evaluation: Desorption efficiency

For successful PRCC methodology (as stated in 3.2a), after the partial
reduction and cyanylation is carried out in trap 1, the products formed must be
eluted quantitatively from T1. According to the instrumental setup shown in
Figure 74 species eluted from T1 go directly to the reactor (R) where they are
cleaved; the resulting cleavage products are captured in trap 2. For efficient
cleavage, species desorbed from T1 must spend a certain time (residence time)
in the reactor in order for the cleavage reaction to reach completion. For this
experiment, the cleavage reagent solution (that is pumped into the mixer from a
independent reagent bottle) was set to a continuous flow rate of 400 pL/min to

allow the cleavage reaction to take place in the reactor in about 10 minutes of

193

residence time for species desorbed from T1. In order efficiently trap the
cleavage products, the concentration of ACN in the eluent coming from the mixer
to trap 2 has to be lower than 5%, otherwise cleavage products will flow out of
T2. Thus, a solvent flow program was used that would desorb the proteins from
trap 1 while providing a continuous constant flow of the cleavage reagent. The
flow program for desorption from T1 was set as listed in Table 10, while different
percentages of ACN were tested to evaluate the desorption efficiency of proteins

from T1.

Table 10. Flow program to
desorb isoforms from trap 1.

T (min) F (yL/min)
50

 

 

0 — 0.5
0.5 — 1.0 40
1.0 — 30.0 20
30.0 0.0

 

For desorption efficiency experiments, RNase-A (~ 1 nmol) and BDK (~ 5
nmol) were loaded ontro trap 1 under acidic conditions (pH = 3). T1 was
connected to the system as shown in Figure 74 (PRCC-system), and
independent experiments using different percentages of ACN were carried out.
After every experiment, T1 was removed from the PRCC-system and connected
to a HPLC system (see Figure 81) to obtain the corresponding chromatogram. It
was expected that after the desorption step, the total absence of
chromatographic peaks for BDK and RNase-A would indicate that the desorption

step had been successful.

194

 

 

 

 

3 RNase-A

g 0.75-

8
A) E 0.50 BDK

(a

f 025.

0.00
51'01'52'02'53055404'5505'5
Tlme(mln)

8)

HPLC column

Pumps

 

No presence of BDK or RNase-A

/ if there is a successful desorption

Chromatogram showing BDK and RNase-A
if there is not a successful desorption

 
  
 
      

 

Acquisition
Data System

 
    

UV-Vls ,
detector Sample

collection

 

Figure 81. Conceptualized HPLC chromatogram (see A) showing that after desorption
experiments, when T1 is connected to a HPLC system (see B) there is an absence of residual

proteins in T1, which indicates a successful desorption step.

After several experiments (data not shown), it was determined that the use
of 90% or 100% ACN (in water containing 0.1% trifluoroacetic acid) was not

efficient for the desorption of the protein. However, the use of 80% ACN was

optimum for a quantitative elution of the protein isoforms from trap 1.

3.29 Reactor Evaluation: Protein Transfer and CN-Protein Cleavage

After the partial reduction and cyanylation in trap 1, cyanylated isoforms

are desorbed from it, cleaved in the reactor, and trapped by means of T2. Thus,

195

 

it was necessary to evaluate the efficiency of the transfer from T1 to T2. For this

purpose, using a system setup as shown in Figure 82, RNase-A (~1 nmol) and

BDK (~5 nmol) were loaded into trap 1. Using the flow program from Table 10

and a flow rate of 400 pL/min for the cleavage solution (1.4 M ammonia), proteins

were desorbed from trap 1 and transferred through the reactor in a continuous

flow to trap 2.

 

 

 

 

 

 

 

 

 

 

Cleavage reagent

 

 
 

===== *- , ' 2'
e .--— 2’5 LI—[J— e”

r
-l- -_- 4v

 
 

Trap 2 uv-vr.
detector

   

J
O O
a

‘

Trap 2

HPLC column

detector

 

 

Figure 82. Top: Setup for the PRCC methodology using a two-column on-line system.
Bottom: Schematic diagram indicating that trap 2 can be connected to a separation
system for analysis of the cleavage products formed in the on-line system.

After this, trap 2 was connected on-line with a separation system as

illustrated on the bottom scheme in Figure 82. Chromatograms were obtained

196

 

and recovery was calculated. In both cases, recovery was above 90% for

RNase-A (92 %) and BDK (95 %).

For testing the on-line system for the cleavage reaction, preformed lF3
was loaded into trap 1 and, using the experimental steps described previously,

IF3 was transferred from trap 1 to trap 2.

It was hypothesized that during this transfer step the cleavage reaction
would take place cleaving IF3 at the CN-cysteines. Thus, after the transfer from
T1 to T2, cleaved isoforms of lF3 would be the species trapped in T2. Trap 2
was washed with water and 0.1 M TCEP was added; the trap was immersed in a
water bath at 30 °C for 30 min for the total reduction reaction. Then, trap 2,
containing the released cleavage products, was connected to a separation
system and a gradient solvent program was started to obtain the chromatogram
shown in Figure 83. Chromatographic fractions were collected and cleavage
products were identified by means of MALDI-MS. In Figure 83, a comparison of
chromatograms for products of the in-vial and on-line approaches is shown. In
general, both chromatograms show similar patterns; thus, the on-line approach

demonstrated its feasibility for the PRCC methodology.

197

 

 

lF3
Cys 40-95
Cleavage
roduct fragment M
1-39 4413.9
112 -40-94 6063.6
ltz -95-124 3302.7
0 (1 -94) 10399.5
9 (itz-40424) 9288.3

 

 

 

 

00>

 

 

 

 

 

 

 

A In-vlal

.8-

 

Absorbance (AU)

.4-

.2--

 

OOOOO

 

.0 - .
20 3O 4O 50 60

Tlmo (min)

 

 

 

 

 

Figure 83. Comparison of chromatograms showing cleavage products for IF3 with in-vial and
on-line approaches under similar experimental conditions (experimental conditions were similar
to the described in section 3.20). Table inset shows the m/z values for cleavage products for
lF3.

3.2f Overall Evaluation of the on-line system: The on-line PRCC

methodology

For the overall evaluation of the system, the next steps were followed:

1) Load sample load into trap 1

2) Add 0.02 M TCEP in 6M guanidine-0.1M citrate buffer at pH = 3 for

partial reduction (15’)

3) Wash trap to remove TCEP

198

4) Add 0.2 M CDAP in GM guanidine-0.1M citrate buffer at pH = 3 for
cyanylation (15’)

5) Wash trap to remove CDAP

6) Cleavage in the reactor during transfer from Trap 1 to Trap 2 while
pumping 1 M NH4OH (or 1 M methylamine) through the mixer (40')

7) Wash Trap

8) Add 0.1M TCEP in GM guanidine-0.1M citrate buffer at pH 3, for total
reduction in Trap 2 (30 min at 37°C)

9) Connect Trap 2 to the HPLC-system and elute cleavage products using
a solvent gradient program of ACN (0.1% TFA)NVater (0.1%TFA).

10) Collect fractions from the HPLC system and acquire MALDI-MS

spectra.

RNase-A (~ 1 nmol) was loaded into the system (Trap 1) and all steps
described above were carried out. MALDI-MS spectra were acquired for the
partial separation of cleavage products eluted from Trap 2. Two independent
experiments were carried out using ammonia or methylamine as cleavage
reagents. At the top of Figure 84, a MALDI mass spectrum is shown from one of
the fractions collected after the whole PRCC methodology was applied on-line
using NH3 as the nucleophile. At the bottom of the Figure 84 MALDI-MS signals
are shown for cleavage products A and B for IF4 coming from experiments using

ammonia or methylamine as cleavage reagents.

199

 

  
 
  

 

 

 

 

 

 

 

 

 

 

 

 

1oo- 3302.554‘12'08
80 -
3
z. - 2646.69 9;
3 3
g 60-1659.22 ‘ 152532
E ’6:
o ‘ g
>
3 4o - 3
e
x I
20 - 5906.84
7062.62
Lh‘ré-
0 I v I v I - T ' l L ' I ' I
2000 4000 8000 8000 10000 12000 14000
m/z
A14
\
15 ‘ 25 6368.50
3 l
14 z. 20 . ‘I I
g ,0 5761.50 2 I
c
o 3 15 1 I
E I 5 I I
3 II 3 1o 4 II ‘
5 5 J - 5 (M ( , II
C. ' & ’le U " r (VI, r‘f (M
' ‘w‘r “ ’1‘. VII!” 1"» 5 ‘ MAI; \‘f'fld‘fl/‘N “W I‘V-f'“
“9.47.445. - a. I V WM
0
o r e
5600 5800 6000 6200 6400 6200 6400 6600
20 ml: m/z
‘ A0
B0 \
\5767 18 6352.59
g, .
V)
r.-
2
E
3 .
5
3 I
0 MM I V V U '7' ' V U '
5500 6000 6500
ml:

 

 

Figure 84. Cleavage products produced using an on-line system for PRCC with NH3 as cleavage
reagent. Top: Peaks representing cleavage products are identified with an arrow. Bottom:
Cleavage products identified for IF4 with different nucleophiles, ammonia (Ao, Bo) or
methylamine (A14, B14). Experiments with different nucleophiles were carried out independently.

200

From the MALDI mass spectra, cleavage products were identified as usual

using the in-vial approach.

In Table 11 is presented a summary of the cleavage products that could
be assigned from the MALDI spectra obtained from different collected fractions
using ammonia as the cleavage reagent. It is apparent that not all expected
cleavage products coming from singly reduced/cyanylated isoforms were
identified in these experiments; however some other cleavage products coming
from double and triply/cyanylated isoforms could be distinguished and assigned.
Data obtained from these experiments with the ‘on-line’ approach were used as
input for the two algorithms previously developed in this laboratory by former
group members [41, 42]. Both algorithms turned out the correct linkage for the
Thus, the ‘on-line’ approach can be considered as a viable

model protein.

alternative for the application of the PRCC methodology in characterizing cystinyl

 

 

proteins.
Table 11. Summary Calcite“
of the cleavage products "’1 VII"!
assigned for an ‘on—line’ '" “W 'W'ID'I
experiment using ammonia. I" 1/ Y 1‘“ 700390
Fragments indicated with * I“ (Y 72"“ 590550
lF1 N x 65-71 78660
were produced by double or .....qu ............. .
. {Y 125 270667
triply reduced and cyanylated IF, {Y 84424 452600
isoforms of RNase-A. 172 N x 2“, 6547.30
IF3 J Y 139 441390
lF3 J Y 4094 600000
If: ,{Y 96424 330270
IF4 J Y 110124 160900
IF4 J Y 157 635310
lF4 J Y 50409 5767.00
* J v 2640 160696
* J v 4056 1751.06
* J v 58-84 2901.22
' J Y 72110 429060
* J Y 7294 2647.00

201

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4.- CONCLUSIONS

During use of the ‘on-line’ approach, there are no sample transfers or
drying steps involved; thus, sample manipulation is minimal and sample losses
are minimized. Because sample losses are minimized, a lower amount of protein

is necessary as a starting material.

The ‘on-line' approach takes only half the time required for the regular in-

vial approach (see Table 12).

Table 12. Comparison of sample processing time required for in-vial and on-line
approaches.

 

 

Time evaluation
- Time associated with the complete determination of the location of disulfide
bonds:
ln-Vial On-llne
1A) Load Sample 1' 10'
1) Partial reduction 15' 15’
2) Cyanylation 15’ 15’
3) Chromatography 30-60’ -
4) Solvent removal 120’ -
5) Cleavage 60' 10’
6) Solvent removal 20' -
7) Total reduction 30’ 30’
8) Optional: Chromatography 30-60’ 60'
Estimated total time: 6 - 6 hrs 3 - 4 hrs

 

 

Trap 2, containing the cleavage products, can be adapted easily to a

separation system that could be coupled to a MS system, such a LC-MS.

202

5.- REFERENCES

[1] N. F. Visser, H. Lingeman, H. lrth, On-line SPE-RP-LC for the determination
of insulin derivatives in biological matrices, J. Pharm. Biomed. Anal. 32 (2003)
295-309.

[2] L. E. Vera-Avila, E. A. Cazares-lbanez, R. Covarrubias-Herrera, E. Camacho-
Frias, On-line methodology for ultratrace determination of polynuclear aromatic
hydrocarbons in water, Rev. Int. Cont. Amb. 18 (2002) 5-16.

[3] L. E. Vera-Avila, J. L. Meraz-Lira, P. Padilla-Cortes, On line methodology for
the determination of trace levels of the herbicides 2.4-D and 2.4-DB in water,
Revista Intemacional de Contaminacion Ambiental 13 (1997) 63-71.

[4] S. Salado C, L. E. Vera-Avila, Online solid-phase extraction and high-
performance liquid chromatographic determination of chlorthalidone in urine, J.
Chromatogr., B 690 (1997) 195-202.

[5] E. Papadopoulou-Mourkidou, J. Patsias, E. Papadakis, A. Koukourikou, Use
of an automated on-line SPE-HPLC method to monitor caffeine and selected
aniline and phenol compounds in aquatic systems of Macedonia-Thrace, Greece,
Fresenius J. Anal. Chem. 371 (2001) 491-6.

[6] P. Campins-Falco, R. Herraez-Hemandez, A. Sevillano—Cabeza, Column-
switching techniques for high-performance liquid chromatography of drugs in
biological samples, J. Chromatogr. 619 (1993) 177-190.

[7] P. Campins-Falco, R. Herraez-Hemandez, A. Sevillano-Cabeza, Column-
switching techniques for screening of diuretics and probenecid in urine samples,
Anal. Chem. 66 (1994) 244-248.

[8] L. E. Vera-Avila, L. Rangel-Ordonez, R. Covarrubias-Herrera, Evaluation and
characterization of a commercial immunosorbent cartridge for the solid-phase
extraction of phenylureas from aqueous matrices, J. Chromatogr. Sci. 41 (2003)
480-488.

[9] N. Takahashi, Y. Takahashi, N. lshioka, B. S. Blumberg, F. W. Putnam,
Application of an automated tandem high-performance liquid chromatographic
system to peptide mapping of genetic variants of human serum albumin, J.
Chromatogr. 359 (1986) 181-191.

[10] N. Takahashi, N. lshioka, Y. Takahashi, F. W. Putnam, Automated tandem
high-performance liquid chromatographic system for separation of extremely
complex peptide mixtures, J. Chromatogr. 326 (1985) 407-418.

203

[11] M. M. Bushey, J. W. Jorgenson, Automated instrumentation for
comprehensive two-dimensional high-performance liquid chromatography of
proteins, Anal. Chem. 62 (1990) 161-167.

[12] G. J. Opiteck, K. C. Lewis, J. W. Jorgenson, R. J. Anderegg. Comprehensive
on-line LC/LC/MS of proteins, Anal. Chem. 69 (1997) 1518-1524.

[13] G. J. Opiteck, J. W. Jorgenson, Two-dimensional SEC/RPLC coupled to
mass spectrometry for the analysis of peptides, Anal. Chem. 69 (1997) 2283-
2291.

[14] K. Wagner, T. Miliotis, G. Marko-Varga, R. Bischoff, K. K. Unger, An
automated on-line multidimensional HPLC system for protein and peptide
mapping with integrated sample preparation, Anal. Chem. 74 (2002) 809-820.

[15] R. E. Majors, Multidimensional and comprehensive liquid chromatography,
LC-GC North America 23 (2005) 1074-1076, 1078, 1080-1082, 1084.

[16] J. C. Giddings, Concepts and comparisons in multidimensional separation, J.
High Resolut. Chromatogr. 10 (1987) 319-323.

[17] Y. Mao, P. W. Carr, Adjusting selectivity in liquid chromatography by use of
the thermally tuned tandem column concept, Anal. Chem. 72 (2000) 110-118.

[18] Y. Mao, P. W. Carr, Application of the thermally tuned tandem column
concept to the separation of several families of environmental toxicants, Anal.
Chem. 72 (2000) 2788-2796.

[19] Y. Mao, P. W. Carr, Separation of selected basic pharmaceuticals by
reversed-phase and ion-exchange chromatography using thermally tuned
tandem columns, Anal. Chem. 73 (2001) 4478-4485.

[20] E. Edwards, J. thomas-Oates, Hyphenating liquid phase separation
techniques with mass spectrometry: on-line or off-line, Analyst 130 (2005) 13-17.

[21] B. K. Choi, Y. M. Cho, S. H. Bae, C. C. Zoubaulis, Y. K. Paik, Single-step
perfusion chromatography with a throughput potential for enhanced peptide

detection by matrix-assisted laser desorption/ ionization-mass spectrometry,
Proteomics 3 (2003) 1955-1961.

[22] Y. Shen, N. Tolic, C. Masselon, L. Pasa-Tolic, D. G. Camp, 2nd, K. K.
Hixson, R. Zhao, G. A. Anderson, R. D. Smith, Ultrasensitive proteomics using
high-efficiency on-line micro-SPE-nanoLC-nanoESl MS and MSIMS, Anal.
Chem. 76 (2004) 144-154.

204

[23] R. Xiang, C. Horvath, J. A. Wilkins, Elution-modified displacement
chromatography coupled with electrospray ionization-MS: on-line detection of
trace peptides at low-femtomole level in peptide digests, Anal. Chem. 75 (2003)
1819-1827.

[24] Y. Jiang, C. S. Lee, On-line coupling of micro-enzyme reactor with micro-
membrane chromatography for protein digestion, peptide separation, and protein
identification using electrospray ionization mass spectrometry, J. Chromatogr. A
924 (2001 ) 315-322.

[25] L. Wang, D. L. Smith, Downsizing improves sensitivity 100-fold for hydrogen
exchange-mass spectrometry, Anal. Biochem. 314 (2003) 46-53.

[26] S. Hara, V. Katta, H. 8. Lu, Peptide map procedure using immobilized
protease cartridges in tandem for disulfide linkage identification of neu
differentiation factor epidermal growth factor domain, J. Chromatogr. A 867
(2000) 151-160.

[27] G. Massolini, E. Calleri, Immobilized trypsin systems coupled on-line to
separation methods: recent developments and analytical applications, J. Sep.
Sci. 28 (2005) 7-21.

[28] G. Marie, L. Serani, O. Laprevote, An on-line protein digestion device for
rapid peptide mapping by electrospray mass spectrometry, Anal. Chem. 72
(2000) 5423-5430.

[29] J. Wu, J. T. Watson, A novel methodology for assignment of disulfide bond
pairings in proteins, Protein Sci. 6 (1997) 391-398.

[30] ASTM D 1193-99e1, ”Standard Specification for Reagent Water", ASTM
lntemational.

[31] M. Cadene, B. T. Chait, A robust, detergent-friendly method for mass
spectrometric analysis of integral membrane proteins, Anal. Chem. 72 (2000)
5655-5658.

[32] M. Karas, D. Bachmann, U. Bahr, F. Hillenkamp, Matrix-assisted ultraviolet
laser desorption of non-volatile compounds, Int. J. Mass Spectrom. Ion
Processes 78 (1987) 53-68.

[33] F. Hillenkamp, U. Bahr, M. Karas, B. Spengler, Mechanisms of laser ion
formation for mass spectrometric analysis, Scanning Microscopy, Supplement 1
(1987) 33-39.

[34] R. L. Cunico, K. M. Gooding, T. Wehr, Basic HPLC and CE of biomolecules,
Bay Bioanalytical Laboratory, Richmond, CA, 1998.

205

 

[35] L. R. Snyder, J. J. Kirkland, Introduction to modern liquid chromatography,
Wiley, New York, NY., 1974.

[36] E. Katz, Handbook of HPLC, Marcel Dekker, New York, 1998.

[37] L. E. Vera-Avila, A. M. Nunez-Gaytan, M. R. Covarrubias-Herrera, Trace
level chromatographic determination of priority pollutant chloro- and nitrophenols
in water with online sample preparation. Int. J. Environ. Anal. Chem. 71 (1998)
147-164.

[38] Hamilton. PRP-3, 300 A Reversed Phase HPLC Columns, (accessed 2003,
2005).

[39] J. Kyte, R. F. Doolittle, A simple method for displaying the hydropathic
character of a protein, J. Mol. Biol. 157 (1982) 105-132.

[40] N. B. Afeyan, N. F. Gordon, |. Mazsaroff, L. Varady, S. P. Fulton, Y. B. Yang,
F. E. Regnier, Flow-through particles for the high-performance liquid
chromatographic separation of biomolecules: perfusion chromatography, J.
Chromatogr. 519 (1990) 1-29.

[41] W. Wu, W. Huang, J. Qi, Y. T. Chou, E. Torng, J. T. Watson, 'Signature
sets', minimal fragment sets for identifying protein disulfide structures with
cyanylation-based mass mapping methodology, J. Proteome Res. 3 (2004) 770-
777.

[42] J. Qi, W. Wu, C. R. Borges, D. Hang, M. Rupp, E. Torng, J. T. Watson,
Automated data interpretation based on the concept of "negative signature mass"
for mass-mapping disulfide structures of cystinyl proteins, J. Am. Soc. Mass
Spectrom. 14 (2003) 1032-1038.

206

CHAPTER 5. CONCLUSIONS AND FUTURE WORK

1.- CONCLUSIONS

During the implementation of an ‘on-line’ system for the Partial
Reduction/Cyanylation CN-induced cleavage-based mass mapping methodology
for cystinyl proteins, it was realized that the cleavage reaction needed to be
optimized. Following a quantitative assessment of the cleavage reaction using
ammonia, primary amines were proposed as alternative nucleophiles.
Methylamine was found to be superior to ammonia as the cleavage nucleophile
in promoting the CN-induced cleavage reaction during sample preparation for
disulfide mass mapping. However, increasing the concentration of ammonia to 5
M (beyond the previously used 1 M) in combination with low temperature (2° C
instead of room temperature) and short reaction time increased the cleavage
product yields substantially [1]. Thus, it seems that low temperature (2 °C) and a
short reaction time (10 min) minimize alkali-induced damage of the cleavage
products and other side reactions that occur during longer incubation times in the
original protocol (1 hour or more). Good repeatability (less than 10% relative
standard deviation) was observed in the yield of cleavage products from ~100-

1000 pmol of starting material.

After the optimization of the experimental conditions for the cleavage

reactions, the use of two or more homologous nucleophiles were recognized as

207

 

an additional tool to facilitate the identification of cleavage products from mass
spectra. Mixtures of methylamine/ammonia as nucleophiles produce
homologous species of certain cleavage products that a are represented by pairs
of mass spectral peaks separated by 14 mass units, which helps the analyst
distinguish these structurally diagnostic species, especially in cases of low-

intensity signals and/or in the presence of potentially interfering chemical noise.

The results of evaluating an ‘on-line’ system for the chemical modification
of cystinyl proteins show great promise for the implementation of an automatic
system in the future where small amounts of starting material (few pmols) can be

analyzed without transfer steps in relatively short experimental time.

2.- FUTURE WORK

As it was stated in Chapter 1, disulfide bonds are one of the key aspects
related to folding among cystinyl proteins and, consequently, their determination
is important for the understanding of the chemical structure and function of these
kinds of proteins, which are emerging as an important class of bioactive proteins
[2, 3]. Currently, proteomics is focused on the identification of proteins based on
a database search involving mass spectrometry; only minor attention is paid to
the determination of the disulfide proteome. However, it is expected that
proteomic research will start to focus on the state of disulfide linkages in cystinyl

proteins; that is why the cyanylated-based mass mapping methodology may gain

208

more importance in a near future; in fact, one feature that currently distinguishes
this methodology is its successful application to cystinyl proteins containing

adjacent cysteines [4-6].

Thus, the future of cyanylation-based mass mapping methodology should
rely on the development of a completely automated system, where the starting
point is the loading of the cystinyl protein and the end point is the report stating

the disulfide linkage. In order to reach this goal, I propose the following projects:

1) Evaluation of the partial reduction experimental conditions: In this
thesis, the experimental conditions for the cleavage reaction were optimized.
However, the first step in the methodology is the partial reduction and cyanylation
of the cystinyl protein. It seems that cyanylation does not represent a major
problem, however, sometimes partial reduction of a protein can be difficult to
accomplish due to the protein structure being so knotted that access of the
reducing agent is limited. The use of a step-wise partial reduction/alkylation (or
cyanylation)-sequence [7-11] in which the protein would be reduced and modified
(alkylated or cyanylated) sequentially should be evaluated for possible
incorporation into an ‘on-line’ system. The protein should be alkylated after the
first partial reduction step, and cyanylated after a second reduction step; in this
way, two disulfide linkages could be determined. However, it would be
necessary to determine whether alkylated cysteines with certain chemical groups

are affected during the cleavage step.

209

Another choice for dealing with the partial reduction step could be the use
of electrochemistry; currently, there is not too much scientific information on this

topic, and it has to be explored.

2) Development of new software: After the development of the CN-based
mass mapping methodology [12], it was evident that experimental data had to be
analyzed more efficiently, thus, the Negative Signature Algorithm (NSA) [13, 14]
was developed to help in the data analysis for the determination of the disulfide
linkage; however, because it was the first version, this program lacked flexibility
because it considered that the methodology was carried out under almost ideal
conditions. Sometimes results from this algorithm can be misleading if they are
not analyzed cautiously, especially from mass spectra of low signal strength.
Fragment Sets Algorithm (FSA) [15] was the second software developed; the
nice feature of this software is that it can deal with imperfect experimental data
providing an output list with a ranking of the most probable disulfide structures for
the protein sample. For both algorithms, reliable results depend on identification

of the cleavage fragments in the mass spectra.

In this thesis, more dependable identification of cleavage products is
proposed by the use of two homologous nucleophiles for labeling of the cleavage
products. Thus, a new software needs to be developed where the input is raw
experimental data from a sample containing labeled cleavage products. This

algorithm must be able to identify cleavage products by means of the mass

210

 

difference between pairs of mass spectral peaks representing the labeled
(homologous) cleavage products. After this, the algorithm could be linked to the

NSA or to FSA for the determination of the disulfide linkage.

3) Integration of the CN-based methodology into a totally automated
system: As it was sketched above, one of the main goals would be the
development of an automatic system in which the cystinyl protein sample could
be loaded, chemically derivatized. and cleaved. Cleavage products analyzed by
mass spectrometry for automatic data interpretation by the use of a new
algorithm could give the correct disulfide linkage. The on-line system proposed

in this thesis can be the starting point to reach this goal.

211

3.- REFERENCES

[1] J. L. Gallegos-Perez, L. Rangel-Ordonez, S. R. Bowman, C. O. Ngowe, J. T.
Watson, Study of primary amines for nucleophilic cleavage of cyanylated cystinyl
proteins in disulfide mass mapping methodology. Anal. Biochem. 346 (2005)
311-319.

[2] H. Yano, S. Kuroda, B. B. Buchanan, Disulfide proteome in the analysis of
protein function and structure, Proteomics 2 (2002) 1090-1096.

[3] J. J. Gorman, T. P. Wallis, J. J. Pitt, Protein disulfide bond determination by
mass spectrometry, Mass Spectrom. Rev. 21 (2002) 183-216.

[4] J. Wu, J. T. Watson, Optimization of the cleavage reaction for cyanylated
cysteinyl proteins for efficient and simplified mass mapping, Anal. Biochem. 258
(1998) 268-276.

[5] T. P. Wallis, J. J. Pitt, J. J. Gorman, Identification of disulfide-linked peptides
by isotope profiles produced by peptic digestion of proteins in 50% (18)O water,
Protein Sci. 10 (2001) 2251-2271.

[6] T. P. Wallis, C. Y. Huang, S. B. Nimkar, P. R. Young, J. J. Gorman,
Determination of the disulfide bond arrangement of dengue virus NS1 protein, J.
Biol. Chem. 279 (2004) 20729-20741.

[7] S. Sechi, B. T. Chait, Modification of cysteine residues by alkylation. A tool in
peptide mapping and protein identification, Anal. Chem. 70 (1998) 5150-5158.

[8] V. Schnaible, S. Wefing, A. Bucker, S. Wolf-Kummeth, D. Hoffmann, Partial
reduction and two-step modification of proteins for identification of disulfide
bonds, Anal. Chem. 74 (2002) 2386-2393.

[9] J. E. Hale, J. P. Butler, V. Gelfanova, J. S. You, M. D. Knierman, A simplified
procedure for the reduction and alkylation of cysteine residues in proteins prior to
proteolytic digestion and mass spectral analysis, Anal. Biochem. 333 (2004) 174-
181.

[10] U. Goransson, D. J. Craik, Disulfide mapping of the cyclotide kalata B1.
Chemical proof of the cystic cystine knot motif, J. Biol. Chem. 278 (2003) 48188-
48196.

[11] U. Goransson, A. M. Broussalis, P. Claeson, Expression of Viola cyclotides
by liquid chromatography-mass spectrometry and tandem mass spectrometry
sequencing of intercysteine loops after introduction of charges and cleavage
sites by aminoethylation, Anal. Biochem. 318 (2003) 107-117.

212

 

 

[12] J. Wu, J. T. Watson, A novel methodology for assignment of disulfide bond
pairings in proteins, Protein Sci. 6 (1997) 391-398.

[13] J. Qi, W. Wu, C. R. Borges, D. Hang, M. Rupp, E. Torng, J. T. Watson,
Automated data interpretation based on the concept of "negative signature mass"
for mass-mapping disulfide structures of cystinyl proteins, J. Am. Soc. Mass
Spectrom. 14 (2003) 1032-1038.

[14] J. Qi, J. Wu, G. A. Somkuti, J. T. Watson, Determination of the disulfide
structure of sillucin, a highly knotted, cysteine-rich peptide, by
cyanylation/cleavage mass mapping, Biochemistry 40 (2001 ) 4531 —4538.

[15] W. Wu. Applications and computational considerations of the cyanylation-
based disulfide mapping methodology for cystinyl proteins, PhD Thesis, Michigan
State University, East Lansing, MI, December 2003.

213

 

llllllllllllllilll