The Chemiluminescences and degradations of ß-lactam antibiotics after oxidation by potassium superoxide

MISSING IMAGE

Material Information

Title:
The Chemiluminescences and degradations of ß-lactam antibiotics after oxidation by potassium superoxide
Alternate title:
Chemiluminescences and degradations of <beta>-lactam antibiotics after oxidation by potassium superoxide
Physical Description:
xvii, 207 leaves : ill. ; 29 cm.
Language:
English
Creator:
Sun, Jingshun, 1967-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Chemiluminescence   ( mesh )
Anti-Bacterial Agents   ( mesh )
Superoxides -- metabolism   ( mesh )
Penicillin G   ( mesh )
Potassium   ( mesh )
Department of Medicinal Chemistry thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Pharmacy -- Department of Medicinal Chemistry -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Bibliography: leaves 199-206
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Jingshun Sun.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029445382
oclc - 51005125
System ID:
AA00021355:00001

Full Text










THE CHEMILUMINESCENCES AND DEGRADATIONS OF P3 LACTAM
ANTIBIOTICS AFTER OXIDATION BY POTASSIUM SUPEROXIDE












By

JINGSHUN SUN


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1998
































Copyright 1998

by

Jingshun Sun

































This work is dedicated to my parents, and my wife, Shuang Yang.
















ACKNOWLEDGMENTS


I wish to offer my sincere thanks and gratitude to the chairman of my committee,

my mentor Dr. John H. Perrin, whose friendship, support and unending patience has

made my graduate career a very enjoyable experience.

I thank all of the graduate students in Medicinal Chemistry department for their

interest in my work, suggestions and friendship. Thanks go to Dr. James Winefordner,

Dr. Ian Tebbett, and Dr. Ray Bergeron for their enthusiasm and consistent support of this

project. Special thanks go to Dr. Kenneth Sloan and Dr. Stephen Schulman for their

helpful advice and friendship over the last several years.

I would like to thank my family for their love and support especially my parents;

they deserve much of the credit for making me the person I am today. Thanks to Jan

Kallman and Nancy Rosa who helped to make Florida my home. I would also like to

extend my thanks to all faculty, students and stuff at Chemistry department whose

assistance is an indispensable part of this achievement. Lastly, I would like to offer my

thanks to Shuang Yang for her love, patience and perseverance.
















TABLE OF CONTENTS


page


ACKNOWLEDGMENTS ...................................................................... iv

LIST OF TABLES ........................................................................................................... viii

LIST OF FIGURES ....................................................................................................... x

ABSTRACT .................................................................................................................... xvii

CHAPTERS

I INTRODUCTION ..................................................................................................... 1
Reactive Oxygen Species .......................................................................................... I
Singlet Oxygen ......................................................................................................... 1
Superoxide Anion .............................................................................................. 4
Oxygen Toxicity and Diseases ................................................................................. 6
Rheum atoid Arthritis ......................................................................................... 6
Pulm onary Emphysem a ..................................................................................... 7
Ischemic-Reperfusion Tissue Injury ................................................................... 8
Carcinogenesis ................................................................................................... 8
Aging ........................................................................................................................ 9
Neurodegenerative Disorders ............................................................................. 10
Chem otherapy .............................................................................................................. 11
Removal of Free Radicals by Superoxide Dismutase and Vitamin E .................... 11
Increasing Form ation of Free Radicals ............................................................ 12
Chemilum inescence and P-Lactam Antibiotics ..................................................... 14
13-Lactam Antibiotics ....................................................................................... 14
Chemiluminescence .......................................................................................... 15

2 OBJECTIVES .............................................................................................................. 32

3 CHEMILUMINESCENT METHODOLOGY OF STATIC MODE ...................... 36











M aterials and Apparatus .............................................................................................. 36
Experim ental M ethods ............................................................................................ 37
Preparations of Stock Solutions ............................................................................... 38
Preparation of Lum inol Solution ..................................................................... 38
Preparation of Hydrogen Peroxide Solution ................................................... 39
Preparation of Saturated Potassium Superoxide Solution ................................ 39
Preparation of Phosphate Buffer Solutions ...................................................... 39

4 THE CHEMICAL PROPERTIES OF POTASSIUM SUPROXIDE ..................... 47
Com parison Studies ................................................................................................. 47
Solvent Selection for Potassium Superoxide .......................................................... 48

5 CHEMILUMINESCENCES AND P-LACTAM ANTIBIOTICS ......................... 54
Introduction .................................................................................................................. 54
Probe of M axim um Em ission W avelength ............................................................ 55
Discussion .................................................................................................................... 55

6 M ECHANISTIC STU DIES ................................................................................... 65
The D euteration Experim ent ................................................................................... 65
NM R Consideration ............................................................................................... 69
NM R Spectra of Penicillin G ............................................................................ 70
NM R Spectra of Penicillin V ............................................................................ 72
Conclusion ....................................................................................................... 74
TLC Analysis ............................................................................................................... 74
Introduction ...................................................................................................... 74
Experim ental ..................................................................................................... 75
Results and D iscussions ................................................................................... 76
HPLC D etection ...................................................................................................... 78
Experim ental ..................................................................................................... 78
Results and D iscussion ..................................................................................... 80
Conclusion ........................................................................................................ 80
HPLC/ESI-MS and HPLC/APCI-MS Analysis of the Degradation of
Penicillin G Following the Oxidation by Potassium Superoxide ..................... 81
Introduction ...................................................................................................... 81
Experim ental ..................................................................................................... 84
Results and D iscussion ..................................................................................... 85

7 SUM M ARY ............................................................................................................... 120

GLOSSARY .................................................. : ................................................................. 123











APPENDICES

A ...................................................................................................................................... 124

B ...................................................................................................................................... 142

C ...................................................................................................................................... 173

D ...................................................................................................................................... 186

E ...................................................................................................................................... 194


REFERENCE .................................................................................................................. 199

BIOGRAPHICAL SKETCH ........................................................................................... 207

































vii











LIST OF TABLES


Table page

Table 1-1: Some deleterious effects of systems generating the superoxide radical ........... 28

Table 1-2: Neurodegenerative disorders associated with free radicals ......................... 29

Table 1-3: Family of 13-lactain antibiotics ................................................................... 30

Table 1-4: Analytical useful chemiluminescent emitters .............................................. 31

Table 3-1: Chemicals and reagents utilized in the static studies of chemiluminescence...41

Table 3-2: Chemical structures and sources of P-lactam antibiotics examined ............ 42

Table 3-2--continued .................................................................................................... 43

Table 3-2--contin ued ................................................................................................. 44

Table 3-3: The composition of phosphate buffer solutions ....................................... 45

Table 6-1: Effects of deuterium oxide on the intensities of chemiluminescence of
penicillin G ........................................................................................................... 92

Table 6-2: The responds of spraying the three reagents on sample PGKO2 and sample
K 0 2 ............................................................................................................................. 92

Table 6-3: The running time of TLC developments ..................................................... 92

Table 6-4: The retention times of the reference compounds and sample PGKO2 ..... 93

Table 6-5: The operating conditions of HPLC ............................................................ 94

Table 6-6: The products of recombination including dimerization after penicillin G
reacting with potassium superoxide .................................................................... 95

Table 6-7: The major products of hydrolysis of penicillin G after interacting with
potassium superoxide .......................................................................................... 96










Table 6-8: The major products of oxidation of penicillin G after interacting with
potassium superoxide .......................................................................................... 97

Table 6-9: The six 3-lactam antibiotics that emit no chemiluminescence ................... 98

Table 6-10: The seven 3-actam antibiotics that emit chemiluminescences ............... 99











LIST OF FIGURES


Figure oane

Figure 1-1: Bonding in the diatomic oxygen molecule ................................................ 19

Figure 1-2: Potential energy curves for the three low-lying electronic states of
m olecule oxygen ................................................................................................. 20

Figure 1-3: State correlation diagrams for the reactions of the three low-lying states
of molecule oxygen with a diene to produce endoperoxide in triplet (T) and
singlet (S) states ................................................................................................... 21

Figure 1-4: Possible mechanism for formation of oxygen free radical during ischemic
reperfusion .......................................................................................................... 22

Figure 1-5: Parent penicillin ....................................................................................... 23

Figure 1-6: Parent cephalosporins ............................................................................... 24

Figure 1-7: Oral beta-lactam antibiotics ...................................................................... 25

Figure 1-8: Nonclassical P-lactam antibiotics ............................................................ 26

Figure 1-9: Reaction scheme of luminol based chemiluminescence detection ....... 27

Figure 3-1: Schematic diagram of setup for chemiluminescent measurement in static
m ode ........................................................................................................................... 46

Figure 4-1: The chemiluminescence of luminol following the oxidation by (a) -1202
and (b) K0 2 .......................................................................................................... 51

Figure 4-2: The chemiluminescence of luminol reacting with K02 in CH3OH. (a) at
starting time, (b) two hours later, (c) three hours later ........................................ 52

Figure 4-3: The correlation between time and the chemiluminescent intensities of
luminol after repeating the injections of 50 jtl of 1:1 CH3CN-CH3OH solution
containing o.1 M K02 into 2.5 ml of 10' M luminol every 15 minutes .............. 53

Figure 5-1: The comparison of chemiluminescent intensities of thirteen 3-lactam
antibiotics following the oxidation by K02 ......................................................... 58











Figure 5-2: The calibration curve of penicillin G ....................................................... 59

Figure 5-3: The calibration curve of dicloxacilin ....................................................... 60

Figure 5-4: The photocounting chemiluminescent spectrum (intensity vs. wavelength)
of penicillin G reacting with superoxide ............................................................ 61

Figure 5-5: The photocounting chemiluminescent spectrum (intensity vs. wavelength)
of luminol reacting with hydrogen peroxide ........................................................ 62
Figure 5-6: The typical chemiluminescent spectrum (intensity vs. time) of 1P-lactam
antibioticals ........................................................................................................... 63

Figure 5-7: The chemiluminescent spectrum (intensity vs. time) of ampicillin ...... 64

Figure 6-1: The profile of chemiluminescences of penicillin G in different solutions
with varied deuterium oxide to water ratios ............................................................. 100

Figure 6-2:'H-NMR spectrum of potassium penicillin G in D20 .................................... 101

Figure 6-3:"3C-NMR spectrum of potassium penicillin G in D20 ................................... 102

Figure 6-4: APT spectrum of potassium penicillin G in D20 .......................................... 103

Figure 6-5:'H-NMR spectrum of the degradation products of potassium
penicillin G in D20 after the oxidation by solid KO2 .............................................. 104

Figure 6-6: '3C-NMR spectrum of the degradation products of potassium
penicillin G in D20 after the oxidation by solid KG2 .............................................. 105

Figure 6-7: 'H-NMR spectrum of potassium penicillin V in D20 .............. 106

Figure 6-8: "3C-NMR spectrum of potassium penicillin V in D20 .................................. 107

Figure 6-9: APT spectrum of potassium penicillin V in 1)20 .......................................... 108

Figure 6-10: 'H-NMR spectrum of the degradation products of potassium
penicillin V in D20 after the oxidation by K02 ....................................................... 109

Figure 6-11: 3C-NMR spectrum of the degradation products of potassium
penicillin V in D20 after the oxidation by KG2 ....................................................... 110











Figure 6-12: The schematic diagrams of TLC assays on sample PG and sample
P G K O 2 .....................................................................................................................111

Figure 6-13: the schematic diagrams of preparing the samples for mass
spectroscopy by the TLC separations ...................................................................... 112

Figure 6-14: The FAB-mass spectrum of the sample from spot 13 ................................ 113

Figure 6-15: The chromatograms of sample PG (a) and sample PGKO2 (b). The
buffer system: pH5.66 phosphate buffer containing 0.1% acetonitrile .................... 114

Figure 6-16: Schematic of major processes occurring in electrospray ............................ 115

Figure 6-17: The total chromatograms of sample PG and sample PGKO2 ..................... 116

Figure 6-18: The zoom-in chromatogram of the new oxidation products
generated by sam ple PGKO 2 .................................................................................... 117

Figure 6-19: The scheme of generation of sulfoxides after penicillin G reacts with
potassium superoxide ............................................................................................... 118

Figure 6-20: The proposed chemiluminescent mechanism of penicillin G after
reacting with potassium superoxide ......................................................................... 119

Figure A-1: The positive HPLC/ESI-MS/MS of benzylpenilloic acid ............................ 125

Figure A-2: The interpretation of the positive HPLC/ESI-MSJMS of benzylpenilloic
acid .......................................................................................................................... 126

Figure A-3: The negative HPLC/ESI-MS/MS of benzylpenilloic acid ........................... 127

Figure A-4: The interpretation of the negative HPLC/ESI-MS/MS of benzylpenilloic
acid ........................................................................................................................... 128

Figure A-5: The positive HPLC/ESI-MS/MS of penicillin G and MS/MS of m/z 335..129

Figure A-6: The negative of HPLC/ESI-MS/MS of penicillin G and MS/MS of
m /z 333 ..................................................................................................................... 130

Figure A-7: The interpretations of the positive and negative HPLC/ESI-MS/MS of
penicillin G ............................................................................................................... 131











Figure A-8: The positive HPLC/ESI-MS/MS of benzylpenillic acid .............................. 132

Figure A-9: The interpretation of the positive HPLC/ESI-MS/MS of beuzylpenillic
acid ........................................................................................................................... 133

Figure A-10: The positive HPLC/ESI-MS/MS of benzylpenicillenic acid ..................... 134

Figure A-11: The interpretation of the positive IPLC/ESI-MS/MS of
benzylpenicillenic acid ............................................................................................. 135

Figure A-12: The negative HPLC/ESI-MS/MS of benzylpenicillenic acid ................... 136

Figure A-13: The interpretation of the negative HIPLC/ESI-MS/MS of
benzylpenicillenic acid ............................................................................................. 137

Figure A-14: The positive HPLC/ESI-MS/MS of benzylpenicilloic acid ....................... 138

Figure A-15: The interpretation of the positive HIPLC/ESI-MS/MS of
benzylpenicilloic acid .............................................................................................. 139

Figure A- 16: The negative HPLC/ESI-MS/MS of benzylpenicilloic acid ...................... 140

Figure A-17: The interpretation of the negative HPLC/ESI-MS/MS of
benzylpenicilloic acid .............................................................................................. 141

Figure B-i: The positive HPLC/ESI-MS/MS of the sulfoxide of penicillamine
dim er ........................................................................................................................ 143

Figure B-2: The interpretation of the positive HPLC/ESI-MS/MS of the sulfoxide
of penicillam ine dim er ............................................................................................. 144

Figure B-3: The negative HPLC/ESI-MS/MS of the sulfoxide of penicillamine
dim er ........................................................................................................................ 145

Figure B-4: The interpretation of the negative HPLCIESI-MS/MS of the sulfoxide
of penicillam ine dim er ............................................................................................. 146

Figure B-5: The positive HPLCiESI-MS/MS of benzylpenilloic acid sulfoxide ............ 147

Figure B-6: The interpretation of the positive HPLC/ESI-MS/MS of benzylpenilloic
acid sulfoxide ........................................................................................................... 148











Figure B-7: The negative HPLC/ESI-MS/MS of benzylpenilloic acid sulfoxide ........... 149

Figure B-8: The interpretation of the negative HPLC/ESI-MS/MS of benzylpenilloic
acid sulfoxide ........................................................................................................... 150

Figure B-9: The positive HPLC/ESI-MS/MS of benzylpenillic acid sulfoxide .............. 151

Figure B-10: The interpretation of the positive HPLC/ESI-MS/MS of benzylpenillic
acid sulfoxide ........................................................................................................... 152

Figure B-11: The negative HPLC/ESI-MS/MS of benzylpenillic acid sulfoxide ........... 153

Figure B-12: The interpretation of the negative HPLC/ESI-MS/MS of benzylpenillic
acid sulfoxide ........................................................................................................... 154

Figure B-13: The positive HPLC/ESI-MS/MS of benzylpenicillenic acid sulfoxide ...... 155
Figure B-14: The interpretation of the positive HPLC/ESI-MS/MS of
benzylpenicillenic acid sulfoxide ............................................................................. 156

Figure B- 15: The negative HPLC/ESI-MS/MS of benzylpenicillenic acid sulfoxide..... 157

Figure B-16: The interpretation of negative HPLC/ESI-MS/MS of
benzylpenicillenic acid sulfoxide ............................................................................. 158

Figure B- 17: The positive HPLC/ESI-MS/MS of penicillin G sulfoxide ...................... 159

Figure B- 18: The interpretation of the positive HPLC/ESI-MS/MS of penicillin G
sulfoxide ................................................................................................................... 160

Figure B-19: The negative HPLC/ESI-MS/MS of penicillin G sulfoxide ....................... 161

Figure B-20: The interpretation of the negative lPLCIESI-MS/MS of penicillin G
sulfoxide ................................................................................................................... 162

Figure B-21: The positive HPLC/ESI-MS/MS of compound 366a-penicillin G
sulfone ...................................................................................................................... 163

Figure B-22: The interpretation of the positive HPLC/ESI-MS/MS of compound
366a-penicillin G sulfone ......................................................................................... 164











Figure B-23: The positive HPLC/ESI-MS/MS of compound 366b-isomer of
penicillin G sulfone .................................................................................................. 165

Figure B-24: The interpretation of the positive HPLC/ESI-MS/MS of compound
366b-isomer of penicillin G sulfone ........................................................................ 166

Figure B-25: The positive HPLC/ESI-MS/MS of compound 366c-isomer of
penicillin G sulfone .................................................................................................. 167

Figure B-26: The interpretation of the positive HPLC/ESI-MS/MS of compound
366c-isomer of penicillin G sulfone ........................................................................ 168

Figure B-27: The negative HPLC/ESI-MS/MS of compound 366c-isomer of
penicillin G sulfone .................................................................................................. 169
Figure B-28: The interpretation of the negative HPLC/ESI-MS/MS of compound
366c-isomer of penicillin G sulfone ........................................................................ 170

Figure B-29: The positive HPLC/ESI-MS/MS of benzylpenicilloic acid sulfoxide ....... 171

Figure B-30: The interpretation of the positive HPLC/ESI-MS/MS of
benzylpenicilloic acid sulfoxide .............................................................................. 172

Figure C-1: The positive HPLC/ESI-MS/MS of penicillamine dimer ........................... 174

Figure C-2: The interpretation of the positive HPLC/ESI-MS/MS of penicillamine
dim er ........................................................................................................................ 175

Figure C-3: The positive HPLC/ESI-MS/MS of compound 440 .................................... 176

Figure C-4: The interpretation of the positive HPLC/ESI-MS/MS of compound 440.... 177

Figure C-5: The positive HPLC/ESI-MS/MS of compound 632 .................................... 178

Figure C-6: The interpretation of the positive HPLC/ESI-MS/MS of compound 632 .... 179

Figure C-7: The negative HPLC/ESI-MS/MS of compound 632 ................................... 180

Figure C-8: The interpretation of the negative HPLC/ESI-MS/MS of compound 632... 181

Figure C-9: The positive HPLC/ESI-MS/MS of compound 642 .................................... 182












Figure C- 10: The interpretation of the positive HPLC/ESI-MS/MS of compound
642 ............................................................................................................................ 183

Figure C-11: The negative HPLC/ESI-MS/MS of compound 642 .................................. 184

Figure C-12: The interpretation of the negative HPLC/ESI-MS/MS of compound
642 ............................................................................................................................ 185

Figure D-1: The negative HPLC/APCI-MS of benzoic acid ........................................... 187

Figure D-2: The interpretation of the negative HPLC/APCI-MS of benzoic acid ........ 188

Figure D-3: The positive and negative HPLC/ESI-MS/MS of compound 336 ............... 189

Figure D-4: The interpretation of the positive HPLC/ESI-MS/MS of compound 336 .... 190

Figure D-5: The interpretation of the negative HPLC/ESI-MS/MS of compound 336... 191

Figure D-6: The positive HPLCfESI-MS/MS of compound 511 ................................... 192

Figure D-7: The interpretation of the positive HPLC/ESI-MS/MS of compound 511 .... 193

Figure E-1: The negative HPLCiESI-MS/MS of compound 269 .................................... 195

Figure E-2: The interpretation of the negative HPLC/ESI-MS/MS of compound 269 .... 196

Figure E-3: The positive HPLC/ESI-MS/MS of compound 353 ..................................... 197

Figure E-4: The interpretation of the positive HPLC/ESI-MS/MS of compound 353.... 198
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
THE CHEMILUMINESCENCES AND DEGRADATIONS OF l3 LACTAM
ANTIBIOTICS AFTER OXIDATION BY POTASSIUM SUPEROXIDE

By

Jingshun Sun

August 1998


Chairman: John H. Perrin, PhD

Major Department: Medicinal Chemistry

Penicillin G is investigated as a model to study how thirteen 3 lactam antibiotics

emit chemiluminescences with different intensities after oxidation with potassium

superoxide. The degradation products of aqueous Penicillin G, which reacts with

potassium superoxide, are analyzed via direct infusion by HPLC electrospray ionization

(ESI) and atmospheric pressure chemical ionization (APCI) mass spectrometry. A

number of products derived from the hydrolysis, oxidation, polymerization and

chemiluminescent reaction are identified, and reaction schemes are proposed.

A positive correlation is observed between the chemiluminescences and the

autoxidations initiated by the singlet oxygen molecule-02('Ag). Deuteration experiments


xvii










reveal that the intensity of the chemiluminescence from Penicillin G in deuterium oxide is

approximately ten times as strong as that in neutral water. It is consistent with the longer

lifetime of singlet oxygen molecule in deuterium oxide. The results presented here

reinforce that singlet oxygen molecules play a crucial role in structural breakage of

protein or DNA as well as in the specific disease entities.


xviii
















CHAPTER 1
INTRODUCTION


Reactive Oxygen Species


A free radical can be defined, as any atom or molecule possessing one or more

unpaired electrons. It can be anionic, cationic, or neutral. In biological and other related

fields, the major free radical species of interest have been oxygen free radicals. The term

oxygen free radicals includes the superoxide anion radical (O2), the hydroperoxyl radical

(HO2), the hydroxyl radical (OH), and the peroxide radical (ROO-, R = Lipid). Actually

these free radicals belong to a group of oxygen molecules called reactive oxygen species

because they have stronger oxidizing ability than oxygen itself. Besides the oxygen free

radicals these reactive oxygen species consist of hydrogen peroxide (H202), hypochlorous

acid (HOCI), lipid peroxide (LOOH), and singlet oxygen ('02).

Singlet Oxygen

When the configuration of the oxygen molecule orbital is described as

KK(2ag)2(2Gj2(3Gg)2(lu)4(l7t1g)4, the ground state oxygen itself is a biradical, as shown in

Figure 1-1. This triplet oxygen has two unpaired electrons, each located in a different n*

antibonding orbital with the same spin quantum number. In accordance with Pauli's

principle that a pair of electrons in an atomic of molecular orbital would have antiparallel











spins, this imposes a restriction on the oxidation by the ground state oxygens because the

new electrons removed concomitantly from nonradical molecules must be of parallel spin

so as to fit into the vacant spaces in the 7t* orbitals. The transition metals found at the

active site of oxidase and oxygenase enzymes have shown the capability of accepting or

donating one electron at a time to increase the reactivity of ground state oxygen and to

overcome the spin restriction. Another way of enhancing the reactivity of ground state

oxygen is to change the spin direction of one of the two electrons in the 7r* orbital. Two

higher energy singlet states of oxygen molecule were found by Childe and Mecke [1] (the

'Zg+ in 1931) and by Herzberg [2] (the 'A. in 1934). These two forms of the oxygen

molecule had an excess energy fraction of a chemical bond, 37.5 and 22.5 Kcal mo1-,

respectively, as shown in Figure 1-2.

Singlet O2('F'g) is readily relaxed to 02(Ag) before it has time to react with

anything. By definition, delta singlet oxygen is not a free radical. Like oxygen free

radicals, however, if singlet oxygen is released in biologic systems, it is capable of

rapidly oxidizing many molecules. In vitro chemical studies of singlet oxygen show

singlet oxygen has the same damaging effects as does the superoxide anion toward

various biological structures including nucleic acids, proteins, and lipids [3].

The two principle reactions characteristic of singlet oxygen are cycloaddition and

the 'ene' reaction [reaction (1) and (2), respectively]:












o0 o + R-c CR C1
c-C" ___=/\
/k \ /\C C/
R R







R



The formation of endoperoxides shown in reaction (1) requires a conjugated

double bond. The formation of hydroperoxides as shown in reaction (2) requires an allylic

hydrogen. When the double bonds are surrounded by electron-donating groups, the

activation energy for both reactions can be reduced to nearly zero and the rate constant

then approaches its maximum value of about 108 1 mol' s-' [4]. The formation of

dioxetanes in singlet oxygen reactions appears to have a much higher activation energy

and is observed chiefly when other pathways are blocked [5,6]. This is consistent with the

predictions of the Woodward-Hoffman selection rules for concerted cycloaddition

reactions [7]. Such orbital symmetry considerations and more quantitative calculations

indicate that the 'Ag+ component is the only one which correlates directly with the ground-

state products in reaction (1) [7]. These correlations are illustrated by the potential energy

curves shown in Figure 1-3.











Superoxide Anion

Given a single electron the ground state oxygen molecule becomes a superoxide

anion (Figure 1-1). It is formed in almost all aerobic cells [8,9,10]. Superoxide chemistry

differs greatly regarding whether reactions are carried out in aqueous solution or in

organic solvents. In nonpolar environments superoxide is a powerful base, nucleophile,

and reducing agent [11,12]. Superoxide can also act as an oxidizing agent, but this ability

is only seen with compounds that can donate protons. In aqueous solution, superoxide

undergoes the so-called dismutation reaction to form hydrogen peroxide and oxygen. The

overall reaction can be written as



202 + 2H+ 1 H2O2 + 02 (3)



Under physiological conditions, even though uncatalyzed, this reaction can be very fast,

i.e. 10 M' S' [13]. The reactivity of superoxide is greatly reduced in water, and

dismutation reaction being favored in this media.

The early superoxide theory of oxygen toxicity had been established on the

accumulation of evidence showing that superoxide dismutation (SOD) enzymes, which

remove superoxide by accelerating the dismutation reaction, are of great importance in

allowing organisms to survive in the presence of oxygen and to tolerate the increased

oxygen concentrations [8-10, 14, 15]. Since SOD enzymes are specific for superoxide as

the substrate, it follows that superoxide must be a toxic species. Indeed, many damaging











effects summarizing in Table 1-1 are associated with superoxide generating systems [16].

However, the damage effects listed in Table 1-1 were produced by superoxide in aqueous

solution, moreover, superoxide has proven to be relatively unreactive toward most

biological components [17]. Therefore, it seems unlikely that superoxide alone can create

such damage.

Historically, it is generally accepted that the superoxide anion radical is not a

particularly reactive species but is potentially toxic. It can be transformed into the highly

dangerous hydroxyl radical following the metal catalyzed Haber-Weiss reaction [8].


transition metal
H202 + O2" catalyst 02 + 0H +-OH (4)



The hydroxyl radical will react with a wide range of biological molecules in its vicinity.

Inspection of Table 1-1 shows that the metal-dependent Haber-Weiss reaction

does not explain all, as there are several damage examples of prevented by SOD but not

by catalase. In many other cases, for example the organism Streptococcus Sanguis, whose

growth appears to be independent of the availability of iron [18] as it contains no heme

compounds and lacks catalase and peroxidase, is damaged by exposure to a superoxide-

generating system, but this damage is not prevented by scavengers of 'OH. Actually, there

is no direct evidence of hydroxyl radical in any superoxide-hydrogen peroxide reacting

enzymatic system, and the Haber-Weiss proposal is totally based on the inhibitory effects

of SOD enzymes and that of hydroxyl radical scavengers. On the other hand the presence











of singlet oxygen molecule 02 ('Ag) generated in the Haber Weiss reactions is proved

by the characteristic 1268 rn chemiluminescence emission spectrum of single

molecule oxygen [19]. In view of the fact that singlet 02 (1Ag) has a sufficiently long

lifetime to diffuse in cells and selectively damage some cell constituents, singlet 02 (Ag)

may be the prime oxidant.

Ironically, it is has been known since 1977 that singlet 02 ('Ag) can be produced

either in the non enzymatic dismutation reaction of superoxide or in the metal catalyzed

Haber Weiss reaction [20]. Singlet 02 ('Ag) has never drawn as much attention as the

hydroxyl radical does due to its low concentration and the difficulty of detection.

Oxygen Toxicity and Diseases


Abnormal production of reactive oxygen species have been associated with a

number of diseases including rheumatoid arthritis, pulmonary emphysema caused by

cigarette smoking, ischemic reperfusion tissue injury, carcinogenesis, aging, and

neurodegeneration disorders such as Alzheimer's Disease, Parkinson's Disease, etc.

Conclusive evidence suggests that all cellular components appear to be sensitive to the

reactive oxygen species damage, lipids, proteins and nucleic acids being the most

susceptible to this injury.

Rheumatoid Arthritis

Rheumatoid arthritis has many characteristics of a free-radical-produced disease.

In rheumatoid arthritis, the synovium of the joins is swollen with an inflammatory











infiltrate, and the joint cartilage becomes eroded. Production of synovial fluid, which

lubricates the joint, is increased, but its viscosity is decreased because of the breakdown

of the polymer hyaluronic acid, which acts as a lubricant. This breakdown may be caused

by oxygen free radicals produced by neutrophils that accumulate in large numbers in the

affected joints of patients with rheumatoid arthritis [21]. Increased levels of the products

of lipid peroxidation reactions are found both in the synovial fluid and in the plasma of

patients with active rheumatoid arthritis [22], this indicates the involvement of hydroxyl

radical although the exact cause is unknown.

Pulmonary Emphysema

The correlation between pulmonary emphysema and a,-protease inhibitor

deficiency has led to the hypothesis that the lung connective tissue damage is

characteristically associated with smoking, and this lung tissue injury results from an

impaired ability of protease inhibitors to protect lung elastin from damage caused by

leukocyte protease. In human, a,-protease inhibitor is the major serum antiprotease, and

following bronchoalveolar lavage of normal persons, it is responsible for more than 90%

of such antielastase activity [23]. Electron spin resonance spectroscopy has shown that

cigarette smoke contains a variety of oxygen- and organic- based free radicals [24] and

that these free radicals completely prevent ct-protease inhibitor activity [25]. A recent

interesting in vitro observation demonstrates that catalase and the antioxidants

glutathione and ascorbic acid completely prevent the removal of elastase inhibitory

capacity of a,-protease inhibitor by cigarette smoke.











Ischemic-Reperfusion Tissue Injury

Ischemic-reperfusion tissue injury can occur in several tissues in addition to the

heart [26, 27]-for example, the small intestine, gastric mucosa, kidney, liver, and skin.

The enzyme xanthine oxidase is widely distributed among these tissues. The intestine,

lung, and liver have particularly high levels. The intestinal mucosa is very sensitive to

ischemic-reperfusion injury. One hour of regional intestinal ischemia produces a

considerable increase in capillary permeability, an effect that is reduced by superoxide

dismutase [28]. Biochemical changes during the ischemic period are thought to be the

basis for a burst of production of free radicals on reintroduction of molecular oxygenation

at reperfusion-Figure 1-4 [29]. As shown in Figure 1-4, during ischemia, adenosine

triphosphate (ATP) is metabolized to the substrate hypoxanthine, and the enzyme

xanthine dehydrogenase is converted to xanthine oxidase by a protease activated by

increased free calcium. On reperfusion, hypoxanthine reacts with molecule oxygen in the

presence of xanthine oxidase to form superoxide anion radicals. In the presence of iron

salts, superoxide anion radicals can form hydroxyl radicals, which can bring about a

number of damages.

Carcinogenesis

Carcinogenesis is thought to occur in two stages. In the initiation stage, a

physical, chemical, or biological agent directly causes an irreversible alternation in the

molecular structure of DNA of the cell. This alternation is followed by a promotion stage,

in which the expression of the genes that regulate cell differentiation and growth is











altered. Oxygen free radicals play a role mostly in the promotion phase of carcinogenesis

[30]. Hyperbaric oxygen, superoxide anion radical, and certain organic peroxides are

tumor promoters but may also be weak complete carcinogens [31]. In contrast, many

antioxidants are antipromoters and anticarcinogens [32]. However, the literature is not

extensive enough to fully describe how the molecular structure of DNA is broken down

by oxygen free radicals or other unknown oxygen species. Further investigations are

needed.



The universality of aging implies that its cause is basically the same in all species.

A free radical hypothesis of aging has been proposed. It suggests that the free radical

produced during normal metabolism of the cell over time damages DNA and other

macromolecules and leads to degenerative diseases, malignant lesions, and eventual death

of the animal [33, 34]. Oxidative DNA damage is rapidly and effectively repaired. The

human body is continually repairing oxidized DNA. An estimated several thousand

oxidative DNA damage sites are present in the human cell every day, most of which are

repaired [35]. A small fraction of unrepaired lesions could cause permanent changes in

DNA and might be a major contributor to aging and cancer. The hypothesis that oxygen

radicals play a role in aging is supported by the observation that, in general, long-lived

species produce fewer endogenous free radicals because of their lower metabolic rate

[36]. Long-lived animals also have more superoxide dismutase than do their short-lived

counterparts, and animal species with the longest life-spans have the highest levels of











superoxide dismutase [37]. A consequence of the free radical hypothesis of aging is the

concept that free radical scavenging agents might be used to prevent aging. Several

antioxidants, including vitamin E [38] and butylated hydroxytoluene (BHT) [39] have

been tested in animals and have yielded equivocal results. Interestingly, a study of a self-

selected group of high dose vitamin E users who were 65 years or older showed an

increased mortality associated with consumption of more than 1,000 IU of vitamin E a

day [40]. The correlation between life expectancy, life-span, and dietary antioxidant

intake in humans remains to be demonstrated.

Neurodegenerative Disorders

The central nervous system (CNS) is particularly vulnerable to free radical

damage because of the following anatomical, physiological and biochemical reasons [41]:

1. Relative to its size, there is an increased rate of oxidative metabolic activity.

2. There are relatively low levels of antioxidants (e.g., glutathione) and

protective enzyme activity (e.g., glutathione peroxidase, catalase, superoxide

dismutase).

3. Abundant readily oxidizable membrane polyunsaturated fatty acids are

present.

4. Endogeneous generation of reactive oxygen species via several specific

neurochemical reactions are possible.

5. The CNS contains non-replacing neuronal cells, once damaged they may be

dysfunctional for life.











6. The CNS neural network is readily disrupted.

The evidence for the role of free radicals in CNS disorders is generally indirect

due to their extremely reactivity, as well as due to the general inaccessibility of the brain.

In vivo, direct biochemical monitoring is impossible, however, numerous experiments

still indicate the association between free radicals and several major CNS disorders as

shown in Table 1-2.

Chemotherapy

The therapies based on modulation of the formation of free radicals include:

Removal of Free Radicals by Superoxide Dismutase and Vitamin E

Precedents have been made to the possible protection by superoxide dismutase

against some kinds of free radical damage in animals. Orgotein, a bovine Cu-Zn

superoxide dismutase, has been used as an anti-inflammatory protein drug in veterinary

practice [60]. This agent is reported to have beneficial effects in treating rheumatoid

arthritis, Duchenne's muscular dystrophy, and radiation-induced cystitis [61, 62, 63], but

further double-blind placebo controlled studies are needed to confirm these findings [64].

Human recombinant superoxide dismutase has recently been produced, and trials are

planned to evaluate its therapeutic efficacy in myocardial infraction, kidney

transplantation, and bronchopulmonary dysplasia in premature infants [65].

Vitamin E had been shown by studies with isolated cells and animals to protect

against damage caused by free radicals [66]. Under normal conditions, radical scavenging











by vitamin E is just one of many mechanisms involved in minimizing free radical-related

tissue damage. Many of the numerous clinical studies involving large doses of vitamin E

have not been adequately controlled, and the therapeutic benefits of vitamin E remain

controversial [67].

Increasing Formation of Free Radicals

The main emphasis in treating human disease caused by free radicals is to

decrease the formation of free radicals or limit their reaction at critical sites within the

body. However, the formation of free radicals might be considered as a useful therapeutic

strategy when the desired therapeutic effect is to damage or kill certain cells, but

selectivity for the desired site is a priority. The various ions plus short-lived and reactive

free radicals formed from the interaction of ionizing radiation and water have shown

cell-killing effect [68]. Two major free radicals, hydroxyl radical and the equated electron

(e-,), result from the radiolysis of water. The hydrated electron reacts by neucleophilic

addition to produce radical anions, it further reacts with molecular oxygen to produce

organic peroxide radicals:

R + e- R- (5)



R"+ 02 R02 (6)

A surprisingly large number of anticancer drugs in use today, almost half of those

approved drugs in the United States, can form free radicals. This fact, considered together

with the observation that tumor cells may be deficient in the same enzymes, i.e.











superoxide dismutase and catalase, that normally protect cells from free radical damage,

has led to suggestions that free radicals might be involved in the antitumor activity of

some of these drugs [69]. The best evidence of a free radical involvement in the cytotoxic

effect of an anticancer drug is from the glycopeptide antibiotic bleomycin [70].

Bleomycin binds to DNA and in the presence of ferrous iron and oxygen cleaves DNA. A

free radical species, possibly the hydroxyl radical is formed in near vicinity to the DNA

and results in its degradation [71].

During the past three decades, the classic transition metal catalyzed Haber-Weiss

reaction has met with almost universal acceptance and forms the foundation of this very

active research field. Although the interpretation of evidence presented on this subject has

been difficult because of the problems in identifying the free radicals, the toxicity of

hydroxyl radicals generated by the Haber-Weiss reaction have been blamed for almost

every disease caused by reactive oxygen species, and the chemotherapy is exclusively

based on modulating the formation of free radicals. That the oxygen free radicals might

be involved in human disease is not surprising in view of their existence in many

biological systems, which have seemingly evolved elaborate protective mechanisms

which normally effectively prevent damage by these radicals. In certain circumstances

singlet oxygen has been shown to create similar damages to the hydroxyl radical. Few but

strong evidences already support the argument that in some diseases it is singlet oxygen,

not the hydroxyl radical which acts as a prime oxidant. This controversy needs to be











clarified in order to better understand the oxygen toxicity and to invent new therapeutic

strategy as well as new antioxidants.

Chemiluminescence and O-Lactam Antibiotics


D-Lactam Antibiotics

The P-Lactam structure of penicillin was first proposed by Abraham and Chain in

1943, 14 years after Fleming made his first observations on the antibacterial action of

penicillin on a plate seeded with staphylococci. It was opposed by those committed to an

alternative thiazolidine-oxazolone structure, however, the P-lactam structure was finally

established beyond doubt in 1945 when an X-ray crystallographic analysis by Dorothy

Hodgkin and Barbara Low came to a successful conclusion[72]. Penicillin was

pharmacologically ahead of its time. In the early 1940s, as penicillin research was just

getting underway in earnest, the sulfa drugs were a revolutionary new concept in

chemotherapy. The discovery of sulfa drugs, which would selectively react with bacteria,

was a major innovation in chemotherapy [73]. By 1948 the outstanding value in medicine

of benzylpenicillin had been firmly demonstrated, more and more scientists were

encouraged to develop new P-lactam antibiotics. Chemical modifications of P-lactam ring

and the introduction of specific side chains have led to a new series of compounds with

characteristic biological activities. Due to endless efforts, a large family of P-lactam

antibiotics is established and still grows rapidly. Representative structures of those drugs

are shown in Table 1-3. Many 3-lactam antibiotics can be classified into four classes











according to their clinical usefulness and pharmacokinetic properties [74]: (I) parental

penicillins (Figure 1-5), (II) parental cephalosporins (Figure 1-6), (1II) oral penicillins and

cephalosporin (Figure 1-7), (IV) nonclassical P-lactam antibiotics (Figure 1-8). The

analytical means for particular P-lactam antibiotics are different because of their diverse

chemical, physical, and biological properties. In the case of the analysis of penicillin G,

many methods are adopted [75] including titration, colorimetry and UV

spectrophotometry, florescence, HPLC, thin layer chromatography, gas-liquid

chromatography, electrophoresis, polarography, enzyme electrodes, isotopic assay,

biological-based microbiological assay, immunoassay, and enzyme-aided assay

(hydrolysis). In 1991, Schulman and Perrin revealed that cephalothin has the ability to

prolong and intensify the chemiluminescence derived from the cobalt (II)-luminol-

hydrogen peroxide system. It can be used as the basis for the determination of

cephalothin in the range of 0.4-400 pgg m1[76]. Post column chemiluminescent detection

after liquid chromatography has also been used to prevent mixing problems and so

increase the reproductivity and sensitivity.

Chemiluminescence

Chemiluminescence can be defined as the emission of light as a result of the

generation of electronically excited states formed in a chemical reaction. The range of

wavelengths of light emitted is suprisingly large-from the near ultravilet to the infra-red.

Chemiluminescence is "cold light." The energy is released in the form of light rather than











heat. It has wider variety of color than the bioluminescence of the firefly but is far less

efficient. Analysis by chemiluminescence has simple instrumentation, has excellent

sensitivity, has very low limits of detection, and has great selectivity and wide dynamic

ranges. The efficiency of a luminescent reaction is defined as the number of photons

emitted per reacting molecule. The quantum yield is usually written as:
'cL = 4) x I XD (7)


where DR is the yield of product, 'DO the number of molecules entering the

excited state and OF is the fluorescence quantum yield. High yields of excitation

without strong visible emission are possible if OF is very low, as is the case for simple

dioxetanes. Phosphorescence is difficult to observe under the conditions of most of the

reactions, but in principle OD.. can replace OF. Sensitized chemiluminescence results

when transfer of energy takes place between chemiluluminescent reactants and

fluorescent acceptors. The efficiency of this energy transfer (ET) must be taken into

account:

DCL = R X DEsx ET x OF, (8)

where OF, is the fluorescence efficiency of the acceptor, and (Dr is the efficiency of

energy transfer. Addition of fluorescent acceptors can enhance light emission in the case

of dioxetanes. The analytically useful chemiluminescent emitters are listed in Table 1-4.

If a target analyte can be determined via HPLC chemiluminescence then it

probably has one of three characteristics:











1. it either emits chemiluminescence when mixed with a specific reagent;

2. it catalyzes chemiluminescence between other reagents;

3. it is suppresses chemiluminescence between other reagents.

Since not many pharmaceuticals can give out chemiluminescence, most analysis of

HPLC chemiluminescence are based on 2) and 3) reactions, in which luminol (5-amino-2,

3-dihydro- 1,4-phthalazinedione), reacts with oxidants like hydrogen peroxide in the

presence of a base and a metal catalyst to produce an excited state product (3-

aminophthalate, 3-APA) which gives off light at approximately 425 nm. When a

compound eluting from the LC column enhances or suppresses this background light, the

amount of light increased or decreased will represent the amount of analyte. The reaction

system is described in Figure 1-9. Recently Garcia Campana, Baeyens, and Zhao

reviewed chemiluminescence detection in capillary electrophoresis [92]. The combination

of capillary electrophoresis, which is versatile and robust, and chemiluminescence-based

reactions, which are extremely sensitive, is promising for numerous applications in fields

such as environmental analysis and medicine. However, it is still not used as a routine

technique. Several problems need to be solved, for instance, chemiluminescence lacks

selectivity, the emissions of chemiluminescences depend on environmental factors; and

the optimization of the volume and geometry of the detection cell and connecting

hardware.





18




In our experiments the investigation of the behavior of the reactive oxygen

species has always been the priority. Eleven P-lactam antibiotics have been reacted with

potassium superoxide and the oxidation as well as hydrolysis of penicillin G have been

examined. The deuteration experiment using penicillin G reinforces the concept that

singlet oxygen is generated by superoxide dismutation and that singlet oxygen play an

important role in tissue injury and related diseases.









0

(D0
O
(D



Singlet 02


0
O
00
00
0
0



Superoxide


0
()OD
00
00
T
0
0
0

Ground-
state 02
(3Fg02)


O 0
OD ( (D) (D)

T 0
0 0
0 0
0 0
(ID 0D
P Se
Peroxide Singlet 02


022-


(I g02)


Figure l-1: Bonding in the diatomic oxygen molecule.


orbital
0*2p
R*2p
7r2p
c2p
cr2s
cr*1s
alS


('A02)












































Internuclear Distance {A}


Figure 1-2: Potential energy curves for the three low-lying electronic states of molecular
oxygen.















































Figure 1-3: State correlation diagrams for the reactions of the three low-lying states of
molecule oxygen with a diene to produce endoperoxide in triplet (T) and singlet (S)
states.



















































Figure 1-4: Possible mechanism for formation of oxygen free radical during ischemic
reperfusion.











Parental penicillins


Piperacillin
Apalcillin
Azlocillin
Mezlocillin
Frazillin
Pirbencillin
BL-P1654
PC-455
CP-32537
CP-38118
PL-385
EMD-32412


Figure 1-5: Parental penicillins.









Parental cephalosporins


[ Moxalactam]


Figure 1-6: Parental cephalosporins.












Oral beta-lactam antibiotics


Penicillins


I
Penicillin V
Phenethicillin
Propicillin


I
Oxacillin
Cloxacillin
Dicloxacillin
Flucloxacillin
Nafcillin


I
Ampicillin
Amoxicillin
Epicillin


Hetacillin
Cyclacillin


Pivampicillin
Bacampieillin
Talampicillin


Cephalosporins


I
Carindacillin
Carfecillin


I
[Privmecillnm


Cephaloglycin


Cephalexin
Cepharadine
Cefadroxil
Cefroxadine
Cefaclor
FR-10612
RMI-19592
SCE-100
Cefatrizine


Privcephalexinj


Figure 1-7: Oral beta-lactam antibiotics.


i
























Nonclassical beta-lactam antibiotics


[Nocardicin A


Clavulanic acid
CP-45899
CP-47904


Thienamycin
Epithienamycin A
N-acetylthienamycin
MK-0787
Carpetimycin
MM-4550
PS-4-7


Figure 1-8: Nonclassical fi-lactam antibiotics.


In Figure 5, 6, 7, 8 the compounds with similar structures are drawn in the same block.

































Figure 1-9: Reaction scheme of luminol based chemiluminescence detection.







Table 1-1: Some deleterious effects of systems generating the superoxide radical._
Source of Superoxide System studied Damage Comements
Heart-muscle submitochondrial Activity of NADH-CoQ reductase Damage prevented by SOD. Catalase not
particles complex Activity lost protective.
Illuminated FMN Bacteria Loss of viability Protection by SOD
Xanthine + xanthine oxidase Huamn synovial fluid Degradation; loss of viscosity Both SOD and catalase protect
Xanthine + xanthine oxidase Bateriophage R17 Inactivation SOD protects partially
Illuminated FMN Ribonuclease Losss of activity SOD protects partially
Illuminated FMN Calf myoblast cells Growth abnormality, some cell SOD protects partially
death. SODprotectspartially
Hypoxanthine-xanthine oxidase Rat brain membrane (Na+, K )- Inactivation SOD protects partially
ATPase
Acetaldehyde + xanthine oxidase Erythrocyte membranes Lysis SOD protects

Acetaldehyde + xanthine oxidase Arachidonic acid Oxidation Both SOD and catalase protect

Hypoxanthine-xanthine oxidase DNA Degradation, single-strand breaks, SOD, 'OH scavengers and catalase
attack on sugare moiety. protect. Iron salts needed.
Autoxidation of Inhibition of Na+-dependent amino SOD protects, but not catalse.
dihydroxyfumarate acid uptake
Cheek pouch of living ham ster In r a e p rm bi ty o bl d
Hypoxanthine-xanthine oxidase perusedd with superoxide- Increased permeability of blood SOD protects
( eneratinf syperi vessels, leakage of contents.

Autoxidation of dialuric acid Escherichia coli Loss of viability Both SOD and catalase protect

Acetaldehyde + xanthine oxidase Staphylococcus aureus Loss of viability SOD, OH scavengers and catalase
protect. Iron salts needed.
Xanthine + xanthine oxidase Rat lung in vivo (instilled into Acute lung injury, oedema SOD protects but not catalase
lungs) Actuginuy_____Ortetuotcl _
Xanthine + xanthine oxidase Rat heart omithine decarboxylase Inactivation SOD protects, also mannitol (*OH
scavenger)
Xanthine + xanthine oxidase Rat heart mitochondria Lowered P/O ratios and Lower SOD and catalase protect
_________________________ ~respiratory control________________
Xanthine + xanthine oxidase Rat Heart or liver mitochondria Inhibition of net Ca2+ uptake SOD protects, also mannitol (OH
t_+_hxDgetaoairclDcadaut S p scavenger)
Xantbine + xanthine oxidase Dog heart sarcoplasmic reticulum, Decreased Ca24 take Some protection by SOD and mannitol








Table 1-2: Neurodegenerative disorders associated with free radicals.


Disorders
(reference) Syndrome Causes Observation Comments
Amyotrophic
lateral sclerosis ess in ar a s sig mutations in the gene encoding untreatable, pathogenesis unknown,
(42, 43) the form Of CuZnSOD' typically affects adults in midlife

cortical atrophy, neurofibrillary decreased antioxidant levels as
Alzheimer's progressive memory loss, tangles, neuritic (senile) well as glutathione synthetase
disease (44, 45, disorientation, striking mute, plaques, amyloid angiopathy, activity, vitamin E inhibition, b affects over 2 million Americans, rarely
46,47) immobile dementia granulovacuolar degenaration, amyloid fragments generate occurs before the age of 50 years
Hirano bodies free radicals
Down's syndrome mental retardation, accelerated an extra chromosome 21 increased expression of SOD high predisposition to develop
(48) aging (trisomy 21) activity Alzheimer's disease
Ischemia and cognitive impairment ischemic vascular dementia increased lipid peroxidation 3% by age of 65 to 74, 50% over 85
reperfusion (49) stoke
Mitochondrial seizures, myoclonus, optic lack protection of histones,
DNA disorders neuropathy, fatigueability, mitochondrial DNA mutation extensive oxidative very diverse group of uncommen
(50,51) ataxia, vascular headache phosphorylation conditions
Multiple sclerosis increased serum lipid peroxide affects about 250,000 Americans, twice
(52,53) impaired neuronal function demyelination levels, protein degradation as common in women as in men, no
within myelin effective treatment
increased lipid peroxidation,
Parkinson's stooped posture, slowness of abnormal expression of prevalence for both black and white
disease (54, 55, voluntary movement, dopamine depletion, neuronal antioxidant enzymes, increased populations after 50 years old in North
56, 57, 58, 59) expressionless faces, tremor loss in substantia nigra iron content, increased American, but varies greatly in different
dopamine turnover, induce of countries
Parkinson's disease by toxins


acopper-zinc superoxide dismutase, which catalzes the conversion of superoxide.












Table 1-3: Family of 3-lactam antibiotics.


Structure

Parental / oo
R H
Pa a H Penicillin G, R= H---
Penicillin 3 2


R : H o , H 3
0



H




CH-
Parenta C, Noaricn
Parenal F~ ~.L Cephalothin, RI=
Cepholosporin OO
OH OCOCH3



OH

H

HN
4H

Ho
Hp Nocardicin A0

Nonclassical

0
H c1 ,o H NH2

CHO
OH HO H3

Clavulanic Acid Sulfones Thienamycin












Table 1-4: Analytical useful chemiluminescent emitters.


Chemiluminescent X.x (wavelength Region) Reference
Emitters
3-Aminophthalate (from 425 nm 77
Luminol)
CH3Se 750-825 nm 78
CN 383-388 nm 79
HCF 475-750 nm 80
HCHO 350-500 nm 81
HF 670-700 nm 80
HSO 360-380 nm 82
IF 450-800 nm 83
N-methylacridone (from 420-500 nm 84
lucigene)
Na 589 nm 85
NO2 1200 nm 86
OH 306 nm 82
Oxyluciferin (from 562 nm 87
luciferin)
Ru(bpy)3 600 nm 88
S2 275-425 nm 89
SF2 550-875 nm 90
S02 260-480 nm 91
















CHAPTER 2
OBJECTIVES


Investigation of the chemiluminescence following the oxidation of 3-lactam

antibiotics by potassium superoxide will facilitate a better understanding of the

involvement of O2 in numerous human diseases caused by oxygen toxicity.

Recently more and more people began to realize that the generation of OH by

transition metal-catalyzed Haber-Weiss reaction is not enough to account for all the

damaging effects in the biological components. It is believed that the Haber-Weiss

reaction does not occur in biological tissues under normal conditions, but its occurrence

during pathological states, e.g. ischaemia, is possible. Even so the Haber-Weiss reaction

faces the competition from the Fenton type reaction to produce -OH. In the meantime, the

existence of '02 is indicated in the superoxide-generation systems. Mao [93] et al. utilized

2, 2, 6, 6-tetramethyl-4-piperdone as a spin trap of electron spin resonance (ESR)

spectroscopy to study the generation of '02. They observed the '02 spin adduct signal in

the incubation of xanthine, xanthine oxidase and H202, and the depletion of each of the

incubation components led to a sharp decrease in IO2 generation. '02 scavengers like

sodium azide inhibited '02 generation while the -OH scavenger, ethanol, only slightly

decrease the signal intensity. Catalase inhibited '02 generation and H202 enhanced it.











They believe superoxide is capable of generating '02 upon reaction with 11202, i.e. the

Haber-Weiss reaction. Interestingly they also found that the decomposition of KG2

generated 102, which is consistent with the experimental results obtained by Khan et al. in

1970 [94], 1976 [95], 1981 [96], and 1987 [97]. However, Nilsson and Keams [98] in

1974 challenged this conclusion with a deuteration study. They adopted the same reacting

system as did Khan and injected D20 rather than H20 into dry dimethyl sulfoxide

(DMSO) containing KG2, expecting increased intensity of chemiluminescence from the

relaxation of 02('A) to 302 because of the large differences in lifetimes in the two

solvents: 2 g.sec in H20 vs. 20 gsec in D20. However, no enhancement of

chemiluminescent emission was observed, and no oxidation of the '02-acceptor

tetramethylene (TME) could be detected in this system. Their data suggested that the

superoxide anion radical can not be a direct precursor of '02. Since no later evidence has

been presented, and the deuteration experiment is a very reliable approach to prove the

existence of 'O2, the Nilsson and Kearns' study still casts doubt over the possibility of the

generation of '02 from superoxide.

Indeed, despite a tremendous amount of work, this remains one of the most

controversial, confusing, and frustrating areas of research, and the indirect detection of

OH or other reactive oxygen species based on the scavenger's effects or on the spin

trap's signals is one of the major reasons. This is because OH and '02 are very reactive

species, and thus are able to react with many compounds. Therefore, certain compounds

can be spin traps for both of them, and some scavengers of OH may also be effective for











0.2. For example, histidine can eliminate '02 in addition to the molecular -OH radical

[99].

In the current study based on some similar molecular structures, fP-lactam

antibiotics are used as models to study how the superoxide-generating system damages

DNA molecular and protein structures. The following approaches are made to improve

the understanding of oxygen toxicity.

1). Evaluate K02's capability as an oxidizer for chemiluminescence. The same

amount of KG2 and H202 will be used to react with luminol respectively, and under

identical experimental conditions to check which gives the stronger chemiluminescence.

2). Study the chemical properties of KG2 in the liquid phase. A solvent must be

found to solubilize KG2 and allow it to keep its activity as long as possible, because the

biological nature of superoxide anion has to be mimicked and handling solid KG2 is very

dangerous.

3). Investigate whether KG2 can oxidize 3-lactam antibiotics with the resultant

emissions of chemiluminescence. This will answer the question that how and why oxygen

toxicity can damage cell components.

4). Design and carry out new D20 experiments to clarify the historical

controversy.

5). Study the mechanism of chemiluminescence of 3-lactam antibiotics. A

prediction of chemiluminescences from other compounds is desired.





35




6). Identify the degradation products and pathways to further confirm the

chemiluminescent mechanism and oxygen toxicity.

7). Try to answer the question: which plays the major role in structural breakage

of protein and DNA, hydroxyl radical or singlet oxygen?
















CHAPTER 3
CHEMILUMINESCENT METHODOLOGY OF STATIC MODE


Materials and Apparatus


The chemiluminescent studies were conducted with a FL-750 Spectrofluorescence

Detector (Mcpherson, Acton, MA) in the static mode. The data was collected either by a

Servogor 120 flatbed recorder (Norma Goerz Instruments, Elk Grove Village, IL) or by

an IBM personal computer XT with Spectra Calc (Galactic Industries Corporation,

Salem, NH) as the software and DT 2811 Analog plus Digital Input / Output board (Data

Translation Inc., Malbora, MA) as the interface. The sample was weighed accurately to

0.1 mg by a XE-100A Electronic Balance (Denver Instrument Company, Arvada,

Colorado). A Fisher Model 10 Accumet pH meter (Fisher Scientific, Pittsburgh, PA)

measured the pH values with the sensitivity of 0.01 pH unit (or 1 mv). A Branson

ultrasonic cleaner (Branson, Shelton, Conn.) and a Coming PC-351 Hot Plate Stirrer

(Coming Inc., Coming, NY) was used to facilitate the dissolution of solutes with low

solubilities. The saturated solution of potassium superoxide in 18-crown-6-ether-

acetonitrile was centrifuged with a Beckman CPR Centrifuge (Beckman Instruments Inc.,

Fullerton, CA). The sources and structures of chemicals utilized in these experiments are

listed in Table 3-1.











The cell compartment used for the chemiluminescent detection was modified in

order to prevent light leakage from the outside of the cell. The needle of a syringe was

connected to a curved metal tube, which was immersed in the center of the cuvette. A

black rubber stopper with a small hole allowing penetration of the metal tube covered the

cuvette holder. Since it can not travel through the curved metal tube, the light does not

enter the cuvette on withdrawal of the injection needle. Also a foam box coated with dull

black paint inside and outside served as a "dark room" to cover the whole cell

department. The setup is described in Figure 3-1. This light proof system was tested by

injecting pure water into the cuvette filled with a 10' M solution of alkaline luminol.

Even when the detector was set at its most sensitive status with the High Voltage (HV) of

Photomutiplier (PMT) at 1000 volts (Gain = 10.0), Sensitivity at 0.003, and Time

Constant at 0.25, no signal was observed. After injecting 3% H202 into the same solution,

the photons collected by PMT was off scale. This proves that in the current system only

the light emission inside the cuvette is detected.

Experimental Methods


After the detector and the recorder have been warmed up for several minutes, 2.5

ml of the solution to be detected, was pipetted into the cuvette, and the cuvette was then

placed into the cell holder. When the cell holder was capped by the rubber stopper and the

"dark box" was sealed, the system was ready for injection. At least three injection were

made to get an average of chemiluminescent peak heights or integrated areas with











reasonable reproducibility; usually the Relative Standard Deviation (RSD) was less than

10 %. The cuvette was first washed three times with tap water, then three times with

deionized water, finally rinsed with acetone, and dried with pressured air. The quartz

fluorescent cuvette was frequently used in these experiments. The UV quartz euvette can

also be used, but the transparent walls must face the PMT window, and no emission

filters must be installed so that the PMT collects the total photon emission from the

chemiluminescent reactions. Prior to every new experiment the detecting system was

validated by a luminol-based chemiluminescent system. A 20 jtl, 3% of H202 was

injected into a 2.5 ml, 10-3M of alkaline luminol solution, and a strong, broad peak was

detected. The average peak height was 238.5 mv. The experiment parameters were: Gain

= 7.10 (710 volts), Sensitivity = 0.1, Time Constant = 0.25, recorder Detecting Scale =

500 my, recorder Paper Rate = 1 cm/min. The high concentration of reactants and stable

signal made this validation very reliable.

Preparations of Stock Solutions


All chemicals and reagents were used as provided without further purification.

Preparation of Luminol Solution

The 0.001 M luminol solution was prepared by the following procedures: 88.58

mg of luminol was weighed and then dissolved in a mixture of about 150 ml deionized

water and 6.25 ml of 5 N sodium hydroxide solution with stirring. This solution was

transferred into a 250 ml volumetric flask and the volume was adjusted to 250 ml. After











shaking the flask gently to achieve a homogeneous mixture, the stock solution was stored

in refrigerator. The high pH of this solution benefited the further dilutions as no

additional base was needed to keep luminol in alkaline media.

Preparation of Hydrogen Peroxide Solution

The 3 % H202 solution was prepared by pipetting 10 ml of 30 % H202 solution

into a 100 ml volumetric flask then adjusted to volume with deionized water.

Preparation of Saturated Potassium Superoxide Solution

The KG2 was weighed quickly and put into dry acetonitrile, then 18-crown-6-ether

was added in excess to solubilize the KG2 powder. After gentle stirring, the solution was

stored overnight in the refrigerator. The acetonitrile solution, containing excess KG2 and

crown ether, was centrifuged at 2500 rpm for 10 minutes at room temperature to leave a

saturated solution. Acetonitrile was chosen as the solvent because its polarity enhanced

the solubility of ionic species, and also because it is miscible with the aqueous solutions.

Preparation of Phosphate Buffer Solutions

Based on the Henderson-Hasselbalch equation, Table 3-2 was developed to

prepare phosphate buffer solutions. To make a buffer with pH range from 3.7 to 9.2,

certain amounts of potassium phosphate monobasic and sodium phosphate dibasic were

weighed corresponding to the clost pH value in Table 3-2, and dissolved in about 450 ml

deionized water. While the solution was stirred by a hot plate stirrer until the dissolution

was completed, the pH change was monitored by a pH meter and small amounts of dilute











solutions of sodium hydroxide or hydrochloric acid were added to adjust the pH value to

the desired one. Then, the solution was transferred into a 500 ml volumetric flask and

made to volume. After gently shaking the solution, the pH value of a small portion of it

was measured to confirm the final pH.

All the 3-lactam antibiotic solutions were made just before the detection

experiments to minimize the effects of hydrolysis, and all the stock solutions were shaken

before being utilized.






41






Table 3-1: Chemicals and reagents utilized in static studies of chemiluminescence.

Class Chemical or Reagent j Source Grade

Milli-Q50, Ultra-pure water
Water system. Millipore' Deionized
Deuterium Oxide Aldrichb 99.9 atom % D
Methanol Fisherd HPLC grade
Solvent Reagent Alcohol Fisherd HPLC grade
n-Butanol Fisherd Certified A. C. S.
Acetone Fisherd Pesticide grade
Methylene Chloride Fisherd Certified A. C. S.
18-Crown-6-Ether Aldrichc 99%

Acid and 85 % Phosphoric Acid Fisher" Certified A. C. S.
Base Hydrochloric Acid (2N) Fisherd Certified
Sodium Hydroxide (5N) Fisherd Certified
Potassium Phosphate Monobasic Fisherd Primary standard
Sodium Phosphate Dibasic Fisher" Certified A. C. S.
Sodium Phosphate Tribasic Mallinckrodte Analytical Reagent
Buffer Buffer Solution Concentrate pH4.00 Fisher" Certified
Buffer Solution Concentrate pH6.00 Fisherd Certified
Buffer Solution Concentrate pH7.00 Fisher" Certified
Buffer Solution pH 10.00 Fisherd Certified

Oxidant 30 % Hydrogen Peroxide Fisherd Certified A. C. S.
_Potassium Suproxide Aldrichb N/A (Powder)


Milliporea: Millipore Corp., Bedford, MA.
Aldrichb: Aldrich Chemical Company, Inc., Milwaukee, WI 53233.
Aldrichc: Aldrich Chemical Company, Milwaukee, W153201.
Fisher: Fisher Scientific, Fair Lawn, NJ
Mallinckrodt: Mallinckrodt, Inc., Paris, Kenturkey.






42






Table 3-2: Chemical structures and sources of P-lactam antibiotics examined.


Beta-Lactam Stucture Source
Antibiotic




Azetidinone OT Aldrich!


0


H3
Sulbactamn H3C N Flukab


OH




Clavulanic Acid Beechar

OH


0
HHNO

Cephalothin 0- N Lilly

CHOH

H



Cefotaxime H3C ',O N.sgma


0 OH H3c NH













Table 3-2--continued


Beta-Lactam Structure Source
Antibiotic




H"o
Penicillin G Fluka

OH




H3
Penicillin V Sigma







Ampicillin Interchem


OH

H2f

/. Sigmac
Amoxicillin

OH





Piperacillin "N H Sim

0














Table 3-2---continued


Beta-Lactam Structure Source
Antibiotic



Dicloxacillin 0
H, Sigma!

C"3







H3C






Hetacillin 0 Beechamc
H3

ON1


Aldrich': Aldrich Chemical Co., Inc. Milwaukee, WI.
Flukab: Fluka Chemical Corp. Milwaukee, WI
Beechamc: Smithkline Beecham Pharmaceuticals, Philadelphia, PA.
Lilly": Eli Lilly and Company, Indiapolis, Ind.
Sigma2: Sigma Chemical Co., St. Louis, MO.
Interchem: Interchem Corporation, Paramus, NJ.










Table 3-3: The composition of phosphate buffer solutions.


85 % H3PO4, ml KH2PO4, g Na2HPO4, g Na3PO4, g Measured pH
0.99 6.794 2.5
0.61 6.7983 2.75
0.34 6.8023 3.05
0.12 6.8095 3.43
0.05 6.8103 3.7
6.8133 4.43
6.7312 0.0453 4.73
6.463 0.107 5
6.0688 0.25 5.42
5.232 0.5488 5.84
4.6136 1.0166 6.17
2.7122 1.415 6.56
1.1775 1.9653 7.06
0.5317 2.1866 7.48
0.1553 2.3198 8.04
0.0256 2.3677 8.77
2.363 9.21
2.3614 0.0771 9.59
2.2292 1.868 10.2
2.0447 0.4343 10.64
3.1617 11.79
"The final volume of all buffer solution is 500 ml, and ionic strength is 0.1.


















Black rubber stopper


Curved matel tube,
connected to syringe


Cell holder

Cuvette ILight proof box






I I
PMT K
I'
t I







Figure 3-1: Schematic diagram of setup for chemiluminescent measurement in static
mode.

















CHAPTER 4
THE CHEMICAL PROPERTIES OF
POTASSIUM SUPROXIDE

Comparison Studies


Since it is a well-documented oxidant for the chemiluminescent reactions of

luminol, H202 was used as a reference to evaluate K02's capability as an oxidizer

producing chemiluminescence.

At the same concentration (1 M), 40 pll each of aqueous H202 solution (a) and

K02 (b) were reacted with 2.5 ml of 10'6M luminol. The chemiluminescences were

recorded by a flatbed recorder. The signal profiles are shown in Figure 4-1. At least three

injections were used to obtain the average peak heights for (a) and (b). The intensities in

term of peak heights can be calculated as

Ia=53.5/ 100 x 0.003 x 100 mv = 0.1605 mv

I 70.5 /100 x 0.1 x 500 mv = 35.25 mv

I1, I. = 35.25 / 0.1605 = 219.6

Here, 53.5 is the peak height in centimeters, 100 is the scale height in centimeters which

corresponds to the signal scale of 100 mv, and 0.003 is the sensitivity of the

photomultiplier tube. Figure 4-1 clearly demonstrates that the chemiluminescence from











the reaction of luminol and KO2 emitted faster, disappeared faster with greater increased

intensity over the reaction with H202. Using the same experimental parameters, further

examinations shown that 5.0 x 10' M luminol was detectable with H202, but 5.0 x 10"

M with KG2. The quicker chemiluminescence from the superoxide system seems more

desirable for quantition of reactions emitting chemiluminescences, and KG2 probably has

the greater potential to make P-lactam antibiotics chemiluminescent compounds.

These results indicate that the reaction mechanism with KG2 is different from that

with H202. To increase the chemiluminescence by more than two magnitude, a significant

lowering of activity energy is necessary. It seems that the superoxide anion decomposes

into other reactive species.

Solvent Selection for Potassium Superoxide

KG2 solutions with a concentration of 0.1 M were made by dissolving 177.5 mg

of KG2 in 25 ml of various solvents just before the chemiluminescent experiments. These

solutions were evaluated using 10'5 M luminol injections made every 15 minutes. The

experimental parameters were the same as in the comparison experiments, and the

stability of chemiluninescent emissions was evaluated.

Clearly, water is not a suitable solvent for KG2 because of its extremely high

reactivity towards KG2. In CH3CN, KG2 was able to oxidize luminol and emit

chemiluminescences with similar intensities and peak profiles to K2 in H20. However,

the signal reproducibility was not good. Although CH3CN is miscible with H20, the











mixing status is more variable following injection than is the mixing of the two solutions

in water.

CH3OH was also investigated as a solvent for KO2. The intensity of

chemiluminescence grew abruptly, but disappeared gradually. This is probably due to the

relative low rate of releasing KG2 from CH3OH into the water. Even though the short-

term signals were reproducible, the chemiluminescent emissions of repeated injections

decreased significantly over time, as shown in Figure 4-2.

As nonpolar solvents, n-butanol and methylene chloride can barely dissolve KG2,

and no chemiluminescent signal was observed. Furthermore, these solvents are not

completely miscible with water.

Unfortunately, all of these solvents can not preserve the activity of K02 over two

hours. Even when KO2 was dissolved in the mixed solvents like CH3CN-CH3OH,

CH3CN-H20, and CH3OH-H20, a similar loss of activity of KO2 was observed. The

typical correlation between the time and chemiluminescent intensity is shown in Figure

4-3.

To increase the solubility of KO2 in CH3CN, 18-crown-6-ether was added to give

a saturated solution. After standing over night and centrifuged as described in chapter 3,

the saturated KG2 reagent was tested by reacting with luminol. The results were very

positive, the chemiluminescent intensity was greatly increased due to the high

concentration of KO2, and the signal was stable for at least 24 hours. More encouragingly,

fresh KG2 and crown ether can be added periodically to keep the KG2 amount constant,






50




and the reactivity of this reagent was sustained. The concentration of K02 was found to

be 0.4386 M as determined by atomic emission spectroscopy (Zeeman 5100, Perkin-

Elmer, Norwalk, CT, courtesy to Dr. Bergeron).






































a b



Figure 4-1: The chemiluminescence of luminol following the oxidation by (a) H202 and
(b) K02. Experimental parameters: High Voltage (HV) of Photomultiplier Tube (PMT) is
910 volts, Time Constant is 0.25, Suppression Background is set at high, recorder speed
is I cm/min. Sensitivity of PMT and Signal Scale of recorder is adjusted according to
varied intensities. In (a) Sensitivity = 0.003, Signal Scale = 100 my. In (b) Sensitivity
0.1, Signal Scale = 500 mv.
















































Figure 4-2: The chemiluminescence of luminol reacting with KO2 in CH3OH. (a) at
starting time, (b) two hours later, (c) three hours later.








600

500

400

300

200

100

0


I I I I I


1 2 3 4 5 6 7 8 9 10 11
Time period (15 nin each)



Figure 4-3: The correlation between time and the chemiluminescent intensities of luminol after repeating the injections
of 50 tl of 1:1 CH3CN-CH3OH solution containing 0.1M K20 into 2.5 ml of 10-9 M luminol every 15 minutes.
















CHAPTER 5
CHEMILUMINESCENCES AND P3-LACTAM ANTIBIOTIC


Introduction


In the last few years, quantitative studies of 3-lactam antibiotics by

chemiluminescences were based on either their enhancement [100, 101 ] or inhibition

[102] of the luminol system. In this chapter the chemiluminescent emission following the

interaction of KO2 with 13 13-lactam antibiotics was investigated as a potential analytical

tool.

The chemiluminescent intensities of antibiotics were determined at optimized

parameters (pH 4.5-7.0. 950 volts across the PMT). 50 ml of saturated K02 solution was

injected into 2 ml of 1.0 mg/ml antibiotics in 100 ml deionized water. All aqueous

antibiotic solutions were prepared just before the detection to minimize the hydrolysis

effects. The materials and methods were as described in the chapter 3.

Six of the antibiotics studied gave no measurable signal, while the other seven

emitted differently intensified chemiluminescence. The chemiluminescent intensities of

the antibiotics are compared in Figure 5-1. The strongest chemiluminescence was

observed with penicillin G. The linear dynamic range of 0.01-0.1 mg/ml is shown in

Figure 5-2. Dicloxacillin in the concentration range 0.2 to 1.0 mg/ml also gave a linear











relationship, as shown in Figure 5-3. 30 % H202 had been used to oxidize all the

antibiotics but no signal was observed.

Probe of Maximum Emission Wavelength

The maximum emission wavelength of the chemiluminescence following the

oxidation of 10 mg/ml penicillin G by the saturated KO2 reagent was measured at 535 m

by using a SPEX Photocounting Fluorescence Detector Courtesy of Dr. Winefordner.

After turning off the excitation source, the fluorescence detector was ready for

chemiluminescence determination. The injection of K02 was made at emission

wavelengths of 400, 450, 520, 530, 540, 600, 700, and 800 nm. The intensity observed at

535 nm was the highest. The chemiluminescenct spectrum between 530-590 nm is shown

in Figure 5-4. To validate the detecting system, the spectrum of 10' M luminol (Figure 5-

5) was observed by reacting it with 30% H202, this data is consistent with the literature

[77]. The experiments here demonstrate that the maximum emission from the reaction

between penicillin G and K02 is at 535 nim.

Discussion


There are some unique characteristics produced by the reactions of jP-lactam

antibiotics with K02. First, in these experiments seven of the thirteen P-lactam antibiotics

generated chemiluminescences at pH 4.5-7.0 rather in the more usual alkaline media. It

should be remembered that these antibiotics will hydrolyze at extremely acidic or basic

pH values [103]. Second, the maximum emission wavelength of chemiluminescence











following the oxidation of penicillin G by KG2 was observed at 535 nm, which is

different from the 425 nm of 3-aminophthalate (from luminol) [77], 275-425 nm range of

S2 [89], 260-480 rum of SO, [91]; but is closer to the 420-500 nm range of N-

methylacridone (from lucigene) [84] or 562 rim of oxyluciferin (from luciferin) [87].

Third, the chemiluminescences of the emitting antibiotics disappeared very quickly, the

typical full width at half maximum (FWHM) was 2 sec (Figure 5-6); the slowest

emission, from ampicillin, had a FWHM of 7 sec (Figure 5-7) and probably a diode-array

detector is needed to observe the full spectrum.

However, there are still several questions remaining to be answered.

1. When penicillin G was dissolved in an aprotic solvent like CH3CN, even after

oxidation by the KG2 saturated solution, no chemiluminescent signal was observed. In

dry CH3CN, KO2 exists in the form of superoxide anion radical. This may indicate that

without dismutation reactions with water or other protic solvents, superoxide can not be

transferred into the reactive species, and so the question arises what is the reactive species

generated from the superoxide dismutation reactions? Since H.02 is not able to generate

chemiluminescence after oxidizing any of the 3-lactam antibiotics, a possible candidate

of this reactive specie is 02. Further experiments are conducteded to prove its existence.

2. Why do the six antibiotics including azetidinone, sulbactam, clavulanic acid,

methicillin, penicillin V, and cefotaxime not produce chemiluminescence in the reaction?

Why do the other seven antibiotics gave chemiluminescences of different intensities?

What are the chemiluminescent intermediates and through which pathways do they go?





57




3. Is it possible that in this reaction the hydroxyl radical has been generated, i.e. is

the oxygen toxicity theory based on the Haber-Weiss reaction the right assumption?

To answer these questions, the degradation products of P-lactam antibiotics in the

reactions with KG2 needed to be identified, and a DO2 experiment needed be designed to

prove the existence of 02 in the superoxide dismutation reaction.








0.8
0.7 -
0.6-
=L
""-W 0 .5 -
4 0.4-

Sc 0.3 -
-J
o 0.2-
00
0.1-
0-III
c~ c: C o C CU~ 9
0- -11- M Q 000 C *-- 45
EX 0 CO > U C C. -
CO E E=
CU 0 "a~ l.0

Figure 5-1: The comparison of chemiluminescent intensities of thirteen 13-lactam antibiotics following the oxidation
by K02.












y = 8.3921e-3 + 4.5490x
pA


RA2 = 0.994


Figure 5-2: The calibration curve of penicillin G.


0.4



0.3



0.2



0.1



0.0 -
0.00


0.02 0.04 0.06 0.08 0.10

Concentration (mg/ml)


0.12











0.4 y = 3.9366e-2 + 0.42075x RA2 = 0.994

pA


0.3



0.2




0.1




0.0 ,
0.0 0.2 0.4 0.6 0.8 1.0 1.2


Concentration (mg/ml)



Figure 5-3: The calibration curve of dicloxacillin.












961.


n

t 646,286-
e



t 43t03.857
y



( 0
c 215.489-




5.2 io 540 Sio54 549 540 540
tavelensqth (nm) t. 8
928.888 cps

Figure 5-4: The photocounting chemiluminescent spectrum (intensity vs. wavelength) of penicillin G reacting with
superoxide.












433.


I
n
t323. 143-



I
t 215. 429-
y


C 187.714-
P/\





Wavele1gf.)i640 z,,
)~a 9.0 8928z94



Figure 5-5: The photocounting chemiluminescent spectrum (intensity vs. wavelength) of luminol reacting with
hydrogen peroxide.



































0 20 40 60
Seconds

Figure 5-6: The typical chemiluminescent spectrum (intensity vs. time) of f3-lactam antibiotics.
















.2




4



0-









0 100 2(
Seconds

Figure 5-7: The chemiluminescent spectrum (intensity vs. time) of ampicillin.


200)
















CHAPTER 6
MECHANISTIC STUDIES

Penicillin G was chosen as an example to investigate the mechanism of the

chemiluminescence of P-lactam antibiotics interacting with potassium superoxide.

Various experiments were conducted on the aqueous penicillin G sample and the sample

containing the products generated by oxidizing penicillin G with potassium superoxide.

These experiments include nuclear magnetic resonance (NMR) assay, thin-layer

chromatography (TLC), and high performance liquid chromatography (HPLC) using a

ultraviolet (UV) spectrophotometer, a photodiode array (PDA) detector, and a mass

spectrometry (MS) detector. In all these experiments, solid KG2 was utilized instead of

the saturated K02 reagent (KO2-acetonitrile-18-crown-6-ether) in order to simplify the

experiments. A DO2 experiment was also run to prove the generation of singlet oxygen in

the suproxide dismutation reaction.

The Deuteration Experiment


Stauff et al. [104, 105] proposed that the spontaneous disproportion of superoxide

generating '02, rather than ground state oxygen ('02), as shown in equation (1). This

proposal had received support from the calculations of Knoppenol [106].
spontaneous H202 + 1O2 (1)
02 + 02 +2H+ (











Khan [94, 95] had also proposed that superoxide is a precursor of '0., although he

believes this occurred by a direct electron transfer reaction [equation (2)].

02 a- e- + 102 (2)

Nilsson and Kearns [98], however, based on experiments injecting D20 rather than H20

into the dry dimethyl sulfoxide (DMSO) containing K02, rejected the suggestion that

superoxide serves as a direct precursor of '02. The major question remaining in their

investigation is the fact that even though it can be generated from KO2 after injection of a

small amount of D20 or H20, '02 will be mostly surrounded by DMSO, not by D20 or

H20 because both react with superoxide through the dismutation reaction. Comparing the

30 ptsec lifetime of '02 in DMSO with its 20 ptsec in D20 and its 2 .sec in H20 it is not

surprising that no difference was observed. Also, whether the trace amounts of singlet

oxygen produced in this system is enough to oxidize the '02 acceptor tetramethylene

(TME) is uncertain, and thus it is possible that no oxidation signal of TME could be

detected in the KG2 system adopted by Khan's experiments.

Since a relatively strong chemiluminescence had been obtained in the reaction of

aqueous penicillin G interacting with K02, D20 was used instead of H20 to dissolve

penicillin G, and to examine the deuterium effect on the chemiluminescent intensities. In

view of the fact that in D20 '02 has more time to react with penicillin G, an enhancement

of chemiluminescence supports the concept that the '02 is generated in the superoxide

dismutation reaction. 4.0 mg/ml of penicillin G solutions were prepared in pure D20,

70% D20-30% H20, 50% D20- 50% H20, and pure H20. The chemiluminescences of 2











ml of these solutions were tested by injection of 50 gtl of K02, which is contained in a

saturated solution made of 18-crwon-6-ether and acetonitrile. The experimental

parameters were set as the following: Gain (PMT/HV) is 8.8, Sensitivity is set at auto,

Time Constant is 5, and Suppression Background is set at high. The differently

intensified chemiluminescences are described in Table 6-1, and the chemiluminescent

profiles of penicillin G in the mixture solvents with varied D20 to H20 ratios (from pure

D20-the top to pure H20-the bottom) are demonstrated in Figure 6-1. Inspection of Table

6-1 clearly shows an approximately 9-fold enhancement of chemiluminescent emission of

penicillin G by replacing H20 with D20 as the solvent. Also lowering the ratio of D20 to

H20 was correlated with decreased intensities of chemiluminescences. Another

observation really attracts attention. When the high voltage of photomultiplier tube

(PMT) was adjusted to 980 volts, the time constant was set at 0.25 to catch the weak and

short-life emissions. After the first injection of KG2 into the solution of penicillin G in

pure D20, the huge numbers of photoelectrons collected by the PMT outscaled the

detector. Characteristically, even after a second, or third injection of KG2 into the same

reacting cell, a significant signal was still observed, although their intensities dramatically

decreased. This is totally different from the case of the penicillin G in pure H20, where a

much weaker chemiluminescence from the first injection of KG2 was detectable, but no

signal was observed after the second injection of KG2. This phenomenon further suggests

that '02 can be generated in the superoxide dismutation reaction. In this detection system,

the syringe needle was connected to a long, curved metal tube, which is immersed in the











center of the cell to prevent the light leakage caused by injection. After injection, KG2

diffuses radially from the center to the cell wall. Most penicillin G reacts with the KG2

and the concentration of penicillin G is decreased, especially in the center. KO from the

second or third injection has to travel relatively longer distance to react with the

remaining penicillin G. The singlet oxygens produced from the second or third injection

of KG2 into the D20 therefore have more opportunity to react than those in H20 because

of the longer lifetime of '02 in D20. In the contrast, '02 formed in H20 can not travel far

enough to react with penicillin G away from the center. The repeated injections is an

advantage of investigations in the static mode.

The results presented in this experiment strongly reinforce that in the contrast of

the Haber-Weiss reaction, the '02 can be generated by superoxide dismutation reaction,

which can occur either spontaneously or when catalyzed by SOD with great rate. It

indicates that superoxide maybe must not have time to further react with hydrogen

peroxide to produce hydroxyl radicals and cause oxygen toxicity, as described by the

Haber-Weiss reaction. Even though it is catalyzed by transition metals, which are

available in biological systems, the Haber-Weiss reaction will face competition from the

Fenton type reaction [equation (3), (4)].


02+FeTm Fe +02 (3)


Fe"1 + H202 so Fe1' + OH + OH











Equation (4) is consistent with the fact that plenty of H202 has been generated almost

instantly by the superoxide dismutation reaction. It is already claimed that the evidence

for the formation of OH is overwhelming [107, 108,109, 110]. This does not, of course,

preclude the formation of reactive oxygen species in additional to OH, for example '02 in

equation (3). Taking account of the fact that hydroxyl radicals may also come from the

Fenton type reaction as well as the Haber-Weiss reaction, and the 102 is capable of

damaging the same biological components as is superoxide, not all the oxygen toxicity

should be rationalized by the Haber-Weiss reaction. The toxic species should be specified

in order to better understand the pathology, which causes the diseases, and to develop

new therapeutic strategies as well as new antioxidants. In the experiments discussed

above, it is strongly suggested that '02 well deserves more attention, and it is probably

the prime oxidant in various diseases caused by oxygen toxicity.

NMR Consideration


Two sample of 15 mg of penicillin G ( or penicillin V) in 0.75 ml deuterium oxide

were made, one of them was reacted with 10 mg K02. The 300 MHz 'H-NMR, 75 MHz

"3C-NMR and Attached Proton Test (APT) were obtained by a GEMINI-300 NMR

spectrometer at room temperature. 3-(Trimethylsilyl) propionic-2, 2, 3, 3-d4 acid, sodium

salt (TSP, Aldrich) was used as internal NMR standard and the HDO peak appeared

around 4.8 ppm from TSP. These NMR assays are used to detect the structural alternation











before and after the oxidation of penicillin G or penicillin V by KO, and to check the

purity of these two drugs.

NMR Spectra of Penicillin G

Figure 6-2 is the proton NMR spectrum of potassium penicillin G in D20. The

interpretation of the spectrum is given below.


HO.
0


HN-HNSN
10
penicillin G 1

14


Chemical Shift (ppm)

1.501 (s), 1.570 (s)
3.615, 3.630 (ABq, J=14 Hz)
4.247 (s)
5.430 (d, J=3.9 Hz)
5.493 (d, J=3.9 Hz)
7.278-7.392 (m)


Figure 6-3 depicts the 13C-NMR spectrum of potassium penicillin G in D20. The

assignments are tabulated below.


Chemical Shift (ppm)

26.594, 30.843
42.088
58.190
64.518
66.779
73.305
127.619
129.137,129.486
134.600
174.345, 174.497, 174.786


Position

1,2
10
7
3
6
4
14
12, 13
11
5,8,9


Positin

1,2
10
4
6, 7
6, 7
12, 13, 14










The results here of 'H-NMR and "3C-NMR spectra are similar to these presented

in a study of penicillins and cephalosporins [111].

Compared with Distortionless Enhancement by Polarization Transfer (DEPT),

APT is less sensitive and does not distinguish between CH3 and CH peaks (both down) or

between CH2 and quaternary C peaks (both up). However, in the case of penicillin G and

penicillin V, the partial information obtained by APT is sufficient for interpretations. As

shown in Figure 6-4, the CH3 peaks at position 1, 2 and CH peaks at position 4, 6, 7, 12,

13, 14 are down. Since the stronger peaks at 127.619 ppm, 129.137 ppm, and 129.486

ppm obviously belong to the aromatic carbons, the other down peaks are easily identified.

For the up peaks, the peaks at 174.786 ppm, 174.497 ppm, and 174.345 ppm are clearly

from carbonyl groups; the peaks at 134.600 ppm and 64.518 ppm refer to the quaternary

carbon of aromatic ring and alkanes respectively. Subsequently the peak at 42.008 ppm is

identified as CH2 at position 10.

The degradation products of penicillin G generated by reaction with potassium

superoxide were also examined by NMR spectroscopy without isolation as shown in

Figure 6-5 ('H-NMR spectrum), and Figure 6-6 ('aC-NMR spectrum). The APT spectra

of degradation products were not readable and so are not presented. Interestingly, the

NMR signal at position 10 in Figure 6-6 seems to have disappeared. Probably penicillin

G lost the CH2 at position 10 after oxidation by K02. However, there is a peak at 42.589

ppm close to 42.088 ppm (position 10 in Figure 6-3) in the '3C-NMR spectrum, which

may suggest that the electronic environment has changed but CH2 still exists. More










experiments are needed to clarify the status of the methylene, which is between the

carbonyl and benzyl moiety on penicillin G.

Meanwhile the comparison between the NMR spectra of the phenyl ring before

and after the oxidation by KO, draws attention. If the hydroxyl radical is generated by the

interaction between superoxide and hydrogen peroxide, which is produced by the

superoxide dismutation reaction, the hydroxylate aromatic compounds should be

identified by NMR spectra since the electrophilic hydroxyl radical adds readily to the

aromatic nucleus (K = 3 8 x 109M-' S-) [ 112]. No such hydroxylate compounds are

demonstrated in Figure 6-5 and Figure 6-6. There is no disubstitution or chemical shift of

the aromatic ring to suggest the hydroxylation. Hence, no conclusion can be drawn that

the hydroxyl radical is formed by the Haber-Weiss reaction in this superoxide system.

NMR Spectra of Penicillin V

As shown in Figure 5-1, penicillin G emitted the strongest chemiluminescence

after the oxidation by K02 and surprisingly no chemiluminescent signal was observed

with penicillin V under the same reacting conditions. Considering the fact that the only

structural difference between penicillin G and penicillin V is the "phenoxyl methyl"

moiety of penicillin V, the oxidation of potassium penicillin V in D20 was examined by

NMR spectroscopy by comparing the spectra following the oxidation of penicillin G and

penicillin V by K02.

The proton NMR spectrum of penicillin V in D20 is shown in Figure 6-7, and the

interpretation is described as below.














HN-,//O
H3C --/S...H//--.90
H32"._ -!o--
,4 13

OH
penicillin V


Chemical Shift (ppm)

1.500, 1.513 (s)
4.260 (s)
4.469 (ABq, J=16.2 Hz)
5.520 (d, J=4.2 Hz)
5.569 (d, J=3.9 Hz)
6.837-7.273 (m)


Figure 6-8 is the carbon NMR of penicillin V in D20. The peaks are assigned as

the followings, which are in agreement with the literature values [113].


Chemical Shift (ppm)

26.548, 31.238
57.492
64.670
66.491a
66.613a
73.260
114.841
122.384
130.002
156.908
170.643
174.285b


Position

1,2
7


4
12
13
14
11
5
8.9


,the peaks with chemical shifts of 66.491 and
66.613 ppm are overlapping and they are
differentiated by APT (Figure 6-9).
'the peaks of carbons at position 8 and 9 are
overlapping at 174.285 ppm.


Position

1,2
4
10
6,7
6,7
12, 13, 14











In case of penicillin V, APT is very useful to identify carbon 3 at 66.492 ppm (up)

and carbon 6 at 66.597 ppm (down) as shown in Figure 6-9. The degradation products of

penicillin V after the oxidation by KG2 were also studied by proton NMR (Figure 6-10)

and carbon NMR (Figure 6-11). Both showed no structure alternation on the aromatic

ring, and contrary to penicillin G the CH2 at position 10 remains intact.

Conclusion

As discussed in chapter 1 (equation 7), the quantum yield of chemiluminescence

is associated with (R, the yield of product; DEs. the number of molecules entering the

exited states; and OF, the fluorescence quantum yield. Therefore, the generation of

chemiluminescent products is the first step to emit chemiluminescence. It seems that

unlike penicillin G, penicillin V did not carry out this first step and no chemiluminescent

signal was observed. The NMR data here suggests that the CH2 at position 10 be

correlated with the chemiluminescent emissions, and no hydroxylate aromatic compound

was found to support the possibility that hydroxyl radical was generated by the Haber-

Weiss reaction in the current superoxide system.

TLC Analysis


Introduction

The TLC procedures have been used to allow the simple and rapid separation and

identification of the different (spontaneous, chemical and enzymatic) degradation

products of penicillins and cephalosporins [114]. After testing several spray reagents, Lin










and Kondo [115] found that the iodine-azide reagent was the best for detection of

penicillin G, producing a yellow color on the TLC plate; the vanillin-phosphoric acid

mixture was the best for the detection of streptomycin (brown) and dihydrostreptomycin

(brown); ninhydrin was best for kanamycin (purple) and fradiomycin (purple). Since the

NMR spectra had indicated that numerous degradation products of penicillin G were

generated after the oxidation by KG2 (Figure 6-5 and 6-6), these three reagents (iodine-

azide, vanillin-phosphoric acid, and ninhydrin) were utilized in the TLC analysis to

separate the hydrolysis products from the oxidation products or other derivatives of

penicillin G. It is anticipated that sufficient samples would be obtained to conduct

experiments using mass spectrometry after separation and purification.

Experimental

TLC plates (1 inch x 3 inches) of silica gel were used for separation. Five

different solvent systems were tested: (A) n-butanol-water-ethanol-acetic acid

(5:2:1.5:1.5); (B) n-butanol-water-acetic acid (4:1:1); (C) acetone-acetic acid (19:1); (D)

acetone-water (17:1); (E) methanol-water (50:50). Among them, solvent system (B) was

the best for separating the degradation products of penicillin G. Each TLC plate was

sprayed consecutively with the following spray reagents: (a) ninhydrin, (b) iodine-azide,

(c) vanillin-phosphoric acid. The iodine-azide reagent was prepared by dissolving 0.5076

g iodine (Sigma) and 58.9 mg sodium azide (Fisher) into 200 ml chloroform to make a

solution of 0.01 N iodine-0.02 % azide. Vanillin (Sigma) 0.9992 g was dissolved in 100

ml phosphoric acid and this solution was diluted 1:1 with methanol to form the vanillin-










phosphoric acid reagent. Ninhydrin 0.2010 g was dissolved in 5 ml of 10 % acetic acid,

and added to 95 ml of n-butanol to produce the ninhydrin reagent. Since spraying

vanillin-phosphoric acid mixture did not mark any new spot, as shown in Table 6-2, this

reagent was not used in the separation procedures.

Results and Discussions

Each of the aqueous penicillin G sample (sample PG) 10 mg/ml and the sample

containing degradation products of penicillin G (sample PGKO2) was made as in the

NMR experiments. Both of them were adjusted to pH 12.00 and were spotted on the same

TLC plate. Each spot was dried with pressured air and the spotting was repeated four

times to increase the amount of sample in each spot. After the first elution in the solvent

system (B), the TLC plate was dried by a hot plate, then ninhydrin and iodine-azide were

sprayed consecutively on the TLC plate. Three spots were found originating from sample

PGKO2, two spots from sample PG. The colors and the moving distances of these spots

were shown in Figure 6-12. Obviously the second spot from sample PGKO2 and the first

spot from sample PG have the same Rf value (distance solute moves/distance solvent

front moves). After the third elution, three spots of sample PGKO were found to have the

same Rf values as those of sample PG, as shown in Figure 6-12. This indicates that while

it was oxidized by K02, the aqueous penicillin G was hydrolyzed as well because of the

strong basicity of superoxide in water [116]. Therefore, the first three spots of sample

PGKO2 after the third elution were believed to be produced from the reactions other than

hydrolysis.










To prepare for the mass spectroscopy experiments, 100 mg/ml of sample PGKO2

was made and put on one of the 36 TLC plates in a line spots. Each spot was dried by

pressured air and the spotting was repeated four times. The plate was sprayed after the

third elution, 4/5 of the plate was covered by a piece of paper and the other 1/5 was

sprayed by ninhydrin and iodine-azide, as shown in Figure 6-13. The invisible three lines

on the 4/5 part of the TLC plate was located by the sprayed 1/5 part of the plate, then

each line on the 4/5 part was cut off separately. After the three desired block on all the 36

plates were cut off and separated into three samples, each of them was washed by

methanol and the silica gel was filtered. The subsequent solution was concentrated into

three approximate 1.5 ml solution (solution 1, 2, and 3). Each solution was repeatedly

spotted 12 times on a TLC plate (Figure 6-13) and developed in the solvent system (B).

After the first elution and the two reagents were sprayed, two spots were found from

solution 1, one from solution 2, and none from solution 3. Possibly solution 3 results

from the running solvents or from the solvent of sample PGKO,, as its spots were

overlapping with the solvent front. These four spots were named 11, 12, 13, and 3.

Following the same procedures of mass production as discussed before, four solid sample

were finally obtained after filtering, and drying and were analyzed by Fast Atom

Bombardment (FAB)-mass spectroscopy. The average lengths of the time consumed

during each run are listed in Table 6-3.

One peak of m/z 105 with quite low intensity was found in one of the four

samples, as shown in Figure 6-14. This is most likely ph-C=O, however, in the other











three sample spectra it appears that they was enough sodium salts in the samples to

suppress other signals. The sodium salts may come from the TLC plates or from other

reagents. Although it is consistent with the results of NMR assay, in which the CH2 seems

lost after the aqueous penicillin G was oxidized by K02, the peak m/z 105 in mass

spectrum of spot 13 seems unreliable and better separation methods are needed.

HPLC Detection


HPLC is widely utilized to analyze the degradation products of penicillin G in

aqueous solutions because of its specificity, ease of use and high sample capacity.

However, no product generated by the oxidation of penicillin G had been studied by

HPLC. In this section the products of aqueous penicillin G following the oxidation by

potassium superoxide is investigated by HPLC using a C 18 column.

Experimental

The HPLC experiments are conducted with the following equipment:

1) Beckman Model 1 IOA pump

Beckman Model 153 UV detector

Beckman Instrument Inc., Fullerton, CA.

2) Hewlett-Packard 3392A integrator

Hewlett-Packard Co., Palo Alto, CA.

3) Waters 501 HPLC pump

WatersM 1996 Photodiode Array Detector











WatersT 717 plus Autosampler

Millipore Corporation, Milford, MA.

4) Exsils ODS column, 150 by 4.6 mm dimension, 7 gm particle size, 100 A pore size.

Keystone Scientific, Inc., Bellefonte, PA.

5) Helium compressed gas, Gainesville Welding Supply, Gainesville, FL.

Various reference compounds were injected and the retention times were

measured in order to identify the unknown peaks, which were produced by the

degradation of penicillin G following the oxidation by potassium superoxide. Hydrogen

peroxide was obtained from Fisher. 3,4-dihydroxybenzoic acid, 2, 4-dihydroxybenzoic

acid, 2, 3-dihydroxybenzoic acid, phenylacetic acid, p-hydroxyphenyl acetic acid, and

benzaldehyde were bought from Aldrich. Benzoic acid was obtained from Sigma.

P-hydroxybenzoic acid was obtained from Eastman (Eastman Organic Chemicals,

Rochester, NY). Salicylic acid was obtained from Mallinckroot (Mallinckroot Chemical

Works, St. Lois, New York, and Montreal). All the chemicals, including the reference

compounds and the samples, were completely dissolved using an ultrasonic cleaner. All

the reagents utilized by the HPLC system were filtered and degassed for about 10

minutes. Prior to the experiments, the HPLC setup was washed by water, 50:50

water:methanol, and methanol consecutively, each for half-hour. The UV detector was set

at 254 un, 0.08 absorbance unit full scale. The flow rate was 1.0 ml/min and the pump

pressure was kept below 4000 psi. The 2 mg/ml of sample PG and PGKO2 were prepared

as in chapter 3.










Results and Discussion

30% MeOH-70% phosphate buffer (pH4.15) is good for the separation of the

hydrolysis products of penicillin G in aqueous solutions, but is not good for oxidized

sample, which is consistent with the observation of Vadino et al. [117]. Acetonitrile 0.1%

in pH 5.66 phosphate buffer was found useful for the detection of the oxidation products

of penicillin G as shown in Figure 6-15. Unfortunately, none of the retention times of

unknown peaks of sample PGKO, matched up with that of the reference compounds as

shown in table 6-4. Since the same C 18 column was not available, the experiments using

photodiode array detector (PDA) were conducted on a different ExilR ODS column.

Similar results were obtained as shown in Table 6-4.

Conclusion

1. When the sample PGKO2 stayed overnight huge peaks appeared hours after the

injections. Were polymers generated?

2. No benzoic acid or its derivatives were identified by this kind of detection and maybe

the amount generated is so small that it is out of the detecting limit of the UV

detector.

3. UV detector is not suitable for the detection of the mixture caused by oxidation and

hydrolysis. Mass spectrometer is the obvious choice.

4. The HPLC buffer of sodium salts must be replaced by ammonium buffers or other

nonmetal-ion solutions since in the TLC analysis vast amounts of sodium salts

suppress the signal for mass spectroscopy.












HPLC/ESI-MS and HPLC/APCI-MS Analysis of the Degradation of Penicillin G
Following the Oxidation by Potassium Superoxide

Introduction

Electrospray ionization mass spectrometry (ESI-MS) was first introduced by

Yamashita and Fenn [118, 119] in 1984. At approximately the same time, a very similar,

independent development was reported by Aleksandrov and coworkers [120]. Actually,

electrospray as a source of gas-phased ions and their analysis by mass spectrometry was

proposed much early by Dole [121 ], however Dole's experiments were too narrowly

focused on the detection of polymeric species, which are not themselves ionized in

solution, and the experimental results were not convincing [122].

Since the end of 1980s, ESI-MS has developed at a tremendous pace and

established itself as an outstanding method for biochemical applications. It has permitted

new possibilities for mass spectrometric analysis of high-molecular-weight compounds of

all types, including proteins, nucleotides, and synthetic polymers. Other analytical

techniques do not provide the same level of detailed information regarding molecular

weights and structures from extremely small amounts of material. Compared to the other

mass spectrometric ionization techniques, ESI has three superior features. First, ESI has

the truly unique ability to produce extensively multiply charged ions, which may be

analyzed on virtually all types of mass spectrometers. Second, the samples analyzed by

ESI-MS must be introduced as a liquid phase. This results in a natural compatibility of











ESI with many types of separation techniques. Third, the extreme "softness" of ESI

process allows the preservation in the gas phase of noncovalent interaction between

molecules as well as the study of three-dimensional conformation. For earlier ionization

methods such as fast atom bombardment (FAB) and plasma desorption, abundant energy

is applied in a highly localized fashion over a short time. These methods lead not only to

ion desolvation but also to fragmentation or even net ionization, that is, the creation of

ions from neutrals. As for ESI-MS, in which the desolvation is achieved gradually by

thermal energy at a relatively low temperature as it is a far softer technique.

There are three major steps in the production of gas-phase ions from electrolyte

ions in solution by electrospray: (1) production of the charged droplets at the ES capillary

tip; (2) shrinkage of the charged droplets by solvent evaporation and repeated droplet

disintegrations, leading ultimately to very small highly charged droplets capable of

producing gas-phase ions; and (3) the actual mechanism by which gas-phase ions are

produced from the very small and highly charged droplets. Kebarle and Tang [123]

described the first two steps in Figure 6-16. As shown in this schematic representation of

the charged-droplet formation, a high voltage of 2-3 KV is supplied to the metal capillary

which are typically 0.2 mm o.d. and 0.1 mm i.d. and located 1-3 cm from the counter-

electrode. Penetration of an imposed electric field into the liquid leads to formation of an

electric double layer in liquid. Enrichment of the surface of the liquid by positive

electrolyte ions leads to destabilization of the meniscus and formation of cone and jet