Citation
Assessment of the photochemistry of 4'-aminomethyl 4,5',8-trimethylpsoralen in a biological matrix

Material Information

Title:
Assessment of the photochemistry of 4'-aminomethyl 4,5',8-trimethylpsoralen in a biological matrix
Creator:
Williams, Renee Eileen
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
99 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Photochemistry ( lcsh )
Blood -- Analysis ( lcsh )
Bloodborne infections -- Prevention ( lcsh )
Blood -- Analysis ( fast )
Bloodborne infections -- Prevention ( fast )
Photochemistry ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 95-99).
Thesis:
Department of Chemistry
Statement of Responsibility:
by Renee Eileen Williams.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
48713371 ( OCLC )
ocm48713371
Classification:
LD1190.L46 2001m .W54 ( lcc )

Downloads

This item has the following downloads:


Full Text
ASSESSMENT OF THE PHOTOCHEMISTRY OF
4-AMINOMETHYL 4,5%8-TRIMETHYLPSORALEN
IN A BIOLOGICAL MATRIX
by
Renee Eileen Williams
B.S., Brigham Young University, 1995
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Chemistry
2001


This Thesis for the Master of Science
degree by
Renee Eileen Williams
has been approved
by
Hfzo 10 ,
Date


Williams, Renee (M.S., Chemistry)
Assessment of the Photochemsitry of 4-Aminomethyl 4,5,8-Trimethylpsoralen in a
Biological Matrix
Thesis directed by Assistant Professor Ellen J. Levy
ABSTRACT
The issue of blood safety has taken high priority in the wake of transfusion associated
diseases such as HIV and hepatitis. Research efforts in recent years have focused on
developing technologies to destroy the infectious capability of these pathogens in
blood products prior to infusion into the patient. One approach currently under
investigation employs a photosensitizer and UV light to sterilize pathogens that
harbor in donated blood products. The photosensitizer has an affinity for DNA and
will intercalate between base pairs. Upon exposure to UV light the photosensitizer
absorbs the light energy. In this excited state the photosentizer can form covalent
bonds or cause bond scission with surrounding material. This photochemistry causes
DNA lesions rendering the pathogen incapable of replicating within a host. The
compound 4-aminomethyl-4,5,8-trimethylpsoralen (AMT) has been explored as a
potential photosensitizer for this sterilization application. Studies demonstrate
acceptable kill of challenge organisms but do not report the photochemistry for AMT
molecules that do not bind to target DNA in a biological matrix.
The purpose of this thesis is to report photoproducts within a biological matrix and
examine additional AMT targets in human serum. Studies demonstrate that 4 major
degradation products are detected post illumination: AMT furan:pyrone heterodimer,
AMT pyronerpyrone homodimer, 6-aminoacetyl-7-hydroxy-4,8,-dimethylcoumarin
and a fourth compound which has many spectral similarities to 6-aminoacetyl-7-
hydroxy-4,8,-dimethylcoumarin but could not be positively identified. Estimates for
the recovery of AMT can only account for 50% of the initial concentration in the form
of photoproducts. Additional assessments show that approximately 50% of the AMT
is bound to serum proteins. A large portion of the AMT may be bound to albumin but
additional protein targets are present which should be investigated.
in


Some toxicity information can be found for AMT but not for the photoproducts.
Cursory studies indicate that the cytotoxicity and mutagenicity of AMT is greater than
the photoproducts. Additional research must be conducted prior to clinical use of a
technology including AMT to ensure that patients are not exposed to carcinogens and
mutagens in an effort to decrease any risk of disease transmission.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Signed
IV


ACKNOWLEDGEMENT
This thesis was a collaborative effort on many fronts.
I would like to express my appreciation to the Chemistry Department of the
University of Colorado at Denver for their faith in me to accomplish an offsite
project.
In addition I want to acknowledge the support of GAMBRO BCT, INC. for providing
the project idea and the means to accomplish the research.
The assistance of Dr. Matthew Platz has been valuable in unveiling the AMT
photochemistry.
I would like to thank the following laboratory groups from the Chemistry Department
at the University of Utah for accepting an orphan and making available their research
facilities for this project: Dr. Cynthia J. Burrows group, Dr. Dale Poulter group, Dr.
Matthew Sigman group and the Mass Spectrometry Laboratory.


CONTENTS
Figures....................................................................viii
Tables........................................................................x
Abbreviations................................................................xi
Chapter
1. Safety of Donated Blood for Transfusion..................................1
1.1 Introduction.............................................................1
2. Analysis of AMT Photoproducts by HPLC with Photodiode
Array Detection........................................................11
2.1 Introduction............................................................11
2.2 Background on Photosensitizer Selection.................................12
2.3 Experimental............................................................13
2.4 Results.................................................................15
2.5 Discussion..............................................................30
3. Analysis of AMT Photoproducts by HPLC Mass Spectrometry.................42
3.1 Introduction............................................................42
3.2 Experimental............................................................43
3.3 Results.................................................................44
3.4 Discussion..............................................................51
4. Analysis of AMT Treated Plasma Proteins.................................58
vi


4.1 Introduction.............................................................58
4.2 Experimental.............................................................60
4.3 Results..................................................................61
4.4 Discussion...............................................................66
5. Analysis of Protein Function of AMT Treated Plasma.......................73
5.1 Introduction.............................................................73
5.2 Experimental.............................................................74
5.3 Results..................................................................76
5.4 Discussion...............................................................78
6. Cytotoxicity of AMT......................................................80
6.1 Introduction.............................................................80
6.2 Experimental.............................................................82
6.3 Results and Discussion...................................................82
7. Toxicity Concerns........................................................86
7.1 Introduction.............................................................86
7.2 Literature Review........................................................88
References....................................................................95
vii


FIGURES
Figure
2.1 AMT in Media without Illumination.....................................17
2.2 AMT in Media with Illumination........................................18
2.3 AMT Spectrum from Literature..........................................21
2.4 AMT Spectrum from HPLC Analysis.......................................22
2.5 AMT in Plasma Extract without Illumination............................23
2.6 AMT in Plasma Extract with Illumination...............................24
2.7 Spectrum of HMT Furan:pyrone Heterodimer..............................33
2.8 Spectrum of Peak at 4 Minutes....................................... 34
2.9 Spectrum of HMT Pyrone:pyrone Homodimer...............................35
2.10 Spectrum of Peak at 7 Minutes........................................36
2.11 Spectrum of Peak at 3.7 Minutes......................................39
2.12 Spectrum of Peak at 14 Minutes.......................................40
2.13 Coumarin Spectrum from Literature....................................41
3.1 Representative Chromatogram of Illuminated Sample.....................45
3.2 Representative Chromatrogram of Illuminated Sample....................46
3.3 Chromatrogram of Concentrated Plasma Extract Illuminated Sample.......49
3.4 6-Aminoacetyl-7-Hydroxy-4,8-Dimethylcoumarin Formation................55
4.1 Chromatogram of Plasma Blank..........................................62
viii


4.2 Chromatogram of AMT in Plasma-Not Illuminated.......................63
4.3 Chromatogram of AMT in Plasma-Illuminated...........................64
IX


TABLES
Table
2.1 AMT in Media................................................19
2.2 AMT in Plasma...............................................25
2.3 AMT in Plasma...............................................26
2.4 AMT in Plasma...............................................28
2.5 AMT in Plasma Concentrated Extract.....................................29
3.1 Masses by Retention Time...............................................47
3.2 Masses by Retention Time...............................................48
3.3 Masses by Retention Time...............................................50
4.1 Scintillation Detector Results from Plasma Protein Analysis............65
4.2 Prevalent Human Plasma Proteins........................................68
5.1 Protein Function Data..................................................77
6.1 Reactivity Grades for Direct Contact Test..............................81
x


ABBREVIATIONS
AIDS Acquired Immunodeficiency Syndrome
AMT 4-Aminomethyl 4,5 ,8-Trimethylpsoralen
AMU Atomic Mass Units
Avg Average
BSA Bovine Serum Albumin
eV electron Volts
HIV Human Immunodeficiency Virus
HMT 4-hydroxymethyl 4,5,8-Trimethylpsoralen
HSA Human Serum Albumin
mAU Milliabsorbence Units
PHS Public Health Service
PUVA Psoralen plus A ultraviolet light
TA Transfusion Associated
UV Ultra violet
V Volts
xi


1.
Safety of Donated Blood for Transfusion
1.1 Introduction
Science continues to search for a suitable replacement for blood. The search is
daunting, as any replacement must perform many complex functions. Blood brings
fuel and oxygen to all cells, carries away waste, transports molecules and even
functions as a communication pathway for the body. When disease and accidents
dilute the efficiency of blood, it must be recharged quickly or life is threatened.
Ensuring enough reserve blood to meet the needs of the community and a nation such
as the U.S. is no small task. Cooperation on many fronts is required for success.
Physicians and clinicians must use the blood products wisely so as not to reduce the
supply unnecessarily. Recruitment for donors is a high priority for the blood banking
clinics to sustain and provide a fresh supply. More than 12 million units of blood are
collected on an annual basis.1 Blood supply management requires much thought as
the blood products required by hospitals and clinics have a finite shelf life, some on
the order of a few days. Along with ensuring a viable blood supply, blood banks must
also ensure blood products are safe and free of disease. Donor contaminated blood
can transfer life threatening conditions when infused into a patient.
1


During the 1980s after acquired immunodeficiency syndrome (AIDS) was
diagnosed, it was found that nearly half of the U.S. hemophiliac population contracted
the disease without an identified means of exposure.2 Up to this time the identified
risk factors for the development of AIDS included sharing of drug needles or sexual
intercourse with persons who harbored the AIDS virus. Not until 1983 when the
human immunodeficiency virus (HIV) was finally isolated was it clear that AIDS was
transmitted by a blood-bome pathogen, as were hepatitis and syphilis. Prior to the
isolation of the virus, patients were only known to contract AIDS after receiving
blood donated by infected individuals.
The occurrence of HIV transmission via blood products reached epidemic proportions
prior to the implementation of steps by health authorities and industry to minimize the
risk of transmission in the blood supply.4 Most of the delay in reducing risk was due
to the mysterious nature of the disease. Since HIV attacks the T-cells of the immune
system, patients with advanced stages of AIDS were diagnosed with diseases such as
pneumonia, Kaposis sarcoma or other opportunistic infections. AIDS patients were
not responding to medication regimens commonly used to treat these diseases. Only
until clinical evidence demonstrated a weakened immune system were physicians able
to define clear symptoms of HIV infection.
2


Patients contracting HIV by receiving contaminated, blood products became known as
transfusion associated (TA) AIDS patients. It was observed that TA AIDS patients
demonstrated an increased susceptibility to the disease.2 AIDS patients who
contracted the disease through high risk behaviors showed clinical manifestations on
the average of seven years post-infection. The median time period in which TA-
AIDS patients demonstrated similar symptoms was much sooner, only 29 months
post-infection. Upon further study of these cases, theories regarding why T A-ADDS
patients were at an increased risk began to emerge. The prevalent theory for such a
high vulnerability for transfusion recipients is three fold: their need for high volumes
of blood, multiple exposures to contaminated products and the fact that they may be
infected with a more virulent strain of HIV. Ironically, these patients developed
AIDS from the very fluid they needed to preserve their own lives.
Due to the fatal nature of HIV and the increased vulnerability of transfusion patients
for AIDS onset, public concern for the risk of infection from the blood supply was
very high. When enough information regarding the cause of AIDS was known,
measures taken by the blood banking industry that dramatically reduced the risk of
patient infection.4,5 AIDS was diagnosed in 1982; in 1983 the United States Public
Health Service (PHS) issued a statement recommending that persons at risk for HIV
infection refrain from donating blood or plasma. An ELISA assay to detect HIV in
3


blood was made available for clinical use in 1985. The PHS announcement in
combination with6 donor education and careful screening of the blood reduced the
rate of infection dramatically.
HIV4 was not the first disease to be transmitted through the blood supply. Infectious
viruses such as hepatitis and influenza were known to be present in donated blood
prior to the 1980s. It was not until the advent of HIV in the blood supply that efforts
to reduce TA disease transmission were implemented. Following the HIV ELISA
assay, additional tests were implemented for routine screening of donated blood.
Currently no less than seven different viral screens are performed on each donated
unit of blood: syphilis, hepatitis B and C HIV types I and II, and human T-cell
lymphotropic virus I and II. Also, in cases of patient susceptibility a screen for
cytomegalovirus is performed. Patients who are immuno-compromised such as those
undergoing extensive chemotherapy, renal transplantation, bone marrow
transplantation, and liver transplantation are more susceptible to infection by
cytomegalovirus and require additional testing on blood products they will receive.5
Testing of blood components combined with donor screening have helped to make the
U.S. blood supply as safe as it has ever been. A 2000 report7 stated that the risk of
transfusion-transmitted viral infections is extremely low. Effectiveness has improved
4


to +99.9%. The risk of HIV infection was reported to be 1 in 677,000, 1 in 641,000
for human T-lymphotropic virus, 1 in 103,000 for hepatitis C, and 1 in 63,000 for
hepatitis B virus.
With all of these measures the risk of blood borne pathogen transfer via transfusion
has not been completely eliminated.5 Screening is only as effective as the honesty of
the donors regarding high-risk behaviors. There are vulnerabilities of the assays as
well. A window of time exists between the time of infection to when the viruses
become prevalent enough in the blood to be detected by the assay. For each virus
assay that time frame varies. In the case of HIV the window of time from infection to
detection is 45 days. The estimated risk of passing an infected unit of screened
blood during that period is 1:225,000.
Technological processes are under investigation to further decrease the risk of TA
disease transmission. The next step to improve blood safety is to commercialize a
technology to inactivate viruses and bacteria in blood.1 A direct method to destroy
and/or compromise the agent of infection is a preferred method to ensure blood safety.
This treatment may be performed on all blood products slated for use in clinics and
hospitals and would prove to be a safeguard against false negative test results. This
5


includes products, which are contaminated and have titers below the detection of the
assay, as well as future infectious pathogens yet undetermined.
Q 1A
One such technology, a solvent/detergent procedure has been implemented in
various parts of the world. For this treatment process, 1% of tri(n-butyl)phosphate
and 1% triton X are added to the donated blood product and then maintained at 30C
for 4 hours. There are several drawbacks to this technique.
1. The procedure may only be used on plasma products (contain no cellular material
only proteins) as the detergents disrupt cell membranes.
2. The solvent/detergent mixture must be removed using a laborious soybean oil
extraction procedure and chromatographic purification techniques prior to clinical
use.
3. This process is performed on large pooled plasma products (from 1000 to 20000
units per batch). Pooling plasma from multiple donors increases the risk of
transfusion reactions.
4. The solvent/detergent is only effective against enveloped viruses. Non-enveloped
viruses such as parvovirus B19 will still be infectious after the procedure.
5. Up to 20% of the plasma volume is lost during the process.
6


Another approach used instead of solvent/detergent for pathogen inactivation involves
photosensitization. The photosensitizer would have to meet very strict criteria to be
useful for this application.1 The photosensitizer must be sufficiently water soluble for
application in a blood matrix, have a high affinity for DNA and a very low affinity for
cell membranes and proteins found in blood. In addition, the compounds must be
non-toxic to the patient, and be effective under a gentle treatment process to maintain
the viability of the blood.
This sensitizer would be added to the blood product and then the mixture would be
exposed to specific wavelengths of light for a defined period of time. During
illumination the photosensitizer absorbs the light. In this excited state the sensitizer
transfers the excess energy which initiates chemical reactions causing catastrophic
damage to surrounding material. These lesions render DNA incapable of replicating
within the host and causing infection.
One group of photosensitizers currently under investigation for the inactivation
procedure is the psoralens."14 Psoralens belong to a group of compounds called
furocoumarins. Furocoumarins are heterocyclic molecules having a furan ring
condensed to a coumarin (a bicyclic ring system involving benzyl and pyrone ring
moieties). Psoralens have known photosensitive properties in the UV region and have
7


been used to treat skin conditions for centuries. In more recent decades psoralens have
, , 1 c I o .
been used to treat psoriasis and vitiligo. A diagram of the psoralen structure is
given below showing the three fused rings. Various substituents such as hyroxyl,
methyl and amino groups are added to the rings to affect the binding affinity of
psoralens as well as their solubility in water. The ring structure is planar and has
similar characteristics to the DNA bases.
4' 5 4
The biological activity of psoralens comes from their affinity for pyrimidine bases14
specifically in the form of hydrogen bonding, aromatic ring stacking and van der
Waals interactions.19 Psoralens intercalate into the helix and primarily interact with
thymidine in DNA and uridine in RNA; minor interactions with cytosine also
occur.14,15 Upon exposure to UVA light (340 400 nm), cyclobutane formation
involving either the C4 and C5 of the furan ring or C3 and C4 of the pyrone ring on
the psoralen molecule with thymidine occurs.20,21
8


With the knowledge of the photoactivity and mechanism of lethality in the treatment
of skin disorders, investigations during the late 1970s began to examine the
effectiveness of psoralens for the inactivation of viruses and bacteria for blood
products. Many psoralens have been investigated for blood product sterilization
applications: 8-methoxypsoralen, 4,5,8-trimethylpsoralen, and 4hydroxymethyl
4,5,8-trimethylpsoralen.14 Another psoralen which has shown considerable promise
for use in blood products 4-aminomethyl-4,58-trimethylpsoralen (AMT).
Methylation of AMT increases its photoreactivity with DNA while the primary amine
on C4 increases its hydrophilic nature sufficient for application in a blood matrix.
Psoralen treatment procedures have demonstrated acceptable kill rates for many
pathogens.13,22"24 These studies assess the efficacy of psoralens in sterilizing human
pathogens but do not focus on the impact residual psoralens in the blood products
may have on the patients who receive them. Nor has there been adequate attention
given to the photochemistry of psorlens within a blood matrix. Identification of
photoproducts in blood and a review of the toxicity of such compounds are absent
from this body of literature.
Concerns regarding the toxicity of the psoralen compounds and photoproducts must
be addressed. Historically in the treatment of skin disorders, the adverse effects of
9


psoralen treatment including erythema, edema, genotoxicity, increased risk of skin
cancer and cataracts have been well documented.u>'17 When these compounds are
considered for introduction directly into the blood stream, bypassing much of the
bodys own defense mechanisms, psoralen photoproducts must be characterized.
Once the degradation profile has been studied, rigorous toxicity screening of the
parent compound and significant photoproducts must be investigated prior to market
approval of such a treatment process. It must be clearly demonstrated that any
carcinogenic or toxic effects of the sterilized product do not overshadow the potential
benefit of the treatment process.
The purpose of this thesis is to report the data and results of experiments conducted to
determine the photochemistry of a psoralen compound (AMT) in a blood matrix. A
review of the known toxicity of the parent molecule and significant photoproducts
will follow the presentation of the AMT photochemsitry.
With the presentation of these experimental results, it is hoped that the need for
additional metabolic and toxic information about psoralens and decomposition
products will be recognized and that the pertinent studies are performed prior to
offering psoralen-treated blood products for routine use in hospitals and clinics.
10


2. Analysis of AMT Photoproducts by HPLC with Photodiode Array
Detection
2.1 Introduction
Psoralen photoproducts have been studied in a variety of solvents such as ethanol,
water, and methylene chloride. From these investigations it becomes evident that the
25 28
photoproducts generated are very dependent upon the experimental conditions.
Upon illumination psoralen dimers have been identified at concentrations of 2.5 mM
in methylene chloride while no dimer was detected in aqueous solution at 20 pM.
Singlet oxygen has also been observed to direct the photodegradation of psoralens.
When a singlet oxygen quencher was added to a matrix containing 8-methoxypsoralen
*yc.
the distribution of photoproducts was significantly shifted, additional studies have
demonstrated the impact of hydrogen ion concentration29 on the degradation products.
With all of the psoralen photoproducts that have been identified in the literature, the
photochemistry determined in a study will only be pertinent to the experimental
design used to acquire the data. A review of psoralen photochemistry and
photoproducts found in the literature will not be sufficient alone but must compliment
actual results obtained from samples which have undergone the treatment process
being examined.
11


Psoralen photoproducts have been identified in the literature from an aqueous matrix
at pH 7 but no studies have been published from data collected in a blood matrix.3
For purposes of understanding the photochemistry that occurs during the sterilization
of blood products, information collected in a blood matrix is crucial as the
complexity of reactions has the potential to increase dramatically
2.2 Background on Photosensitizer Selection
For the sensitizer to be effective in a blood sterilization application it must meet a
rigorous set of criteria. It must have much higher affinity for DNA than for proteins
or cell membranes that will also be present in the blood.1,14 Many molecules which
have a high DNA binding affinity are hydrophobic in nature. Proteins and cell
membranes also have environments that attract hydrophobic molecules; finding a
molecule that is exclusively selective for DNA poses a significant challenge. Higher
hydrophobicity in a molecule tends to decrease water solubility. Application for
blood requires some water solubility so that the agent may be available to associate
with the virus target to achieve inactivation.14 Among psoralens, AMT has shown
promise. Methylation of AMT increases its photoreactivity with DNA while the
primary amine on C4increases psoralen solubility sufficient for application in a blood
matrix.
12


AMT
2.3 Experimental
AMT (170 pM, Sigma Chemical) was dissolved in solution of purified water and
salts. The media content was sodium chloride (115 mM, Aldrich), sodium acetate (30
mM, Aldrich), sodium citrate (10 mM, Spectrum), monobasic sodium phosphate
(11.4 mM, Fischer), dibasic sodium phosphate (14.6 mM, Spectrum). The sodium
phosphate monobasic and dibasic ratio was buffered to pH 7.4.
Once fully dissolved, 30 mL of solution was placed into PVC bags (Charter Medical,
P/N 13101 150 mL transfer bag). The bag was placed in a UVP illuminator (UVP,
Inc.) equipped with twenty 365 nm light bulbs. Bulbs were arranged with 10 above
the samples and 10 underneath. Samples were exposed between 1 minute 20 seconds
and 2 minutes 15 seconds for an energy dose of 2 to 3 J/cm Flux readings were
taken on various days to determine day to day variation of UV dose at 365 nm. An
13


average flux reading was used to calculate the time required to obtain the desired UV
dose.
Post exposure non-illuminated and illuminated samples and blanks (media solution
without AMT) were analyzed by HPLC (Hewlett Packard). A Cl8 column (Hewlett
Packard, Eclipse XDB-C18 P/N 99096902) was used to achieve separation of the
mixture. Column temperature was 30C. Initial mobile phase consisted of 40%
acetonitrile and 60% 5 mM hexanesulfonic acid (Fischer) in water (pH adjusted to
3.2). Acetonitrile concentration increased to 100% in 15 minutes. Flow rate was 1.0
mL/min and the run time 15 was minutes. Detector wavelength was set to 254 nm.
UV profiles of all detected peaks were taken from 190 to 400 nm.
The single donor human plasma used for these experiments was expressed from
whole blood (BBMBC, Denver, CO). The data presented in this chapter was
generated using a unit of 0+ blood containing a generic formulation of citrate,
dextrose and phosphate for purposes of preserving shelf life.
The media formulation outlined above was also used for plasma experiments except
the concentration of AMT was increased. AMT concentration in media was 227 to
340 pM to achieve a concentration of 150 to 250 pM AMT in the final plasma
14


sample. Media was mixed with plasma at a ratio of 2 to 1. A total of 30 mL was
placed into the charter medical transfer bags. Illumination dose used for plasma was
between 2 and 5 J/cm2 (1 minute 20 seconds to 3 minutes 25 seconds).
To remove proteins prior to HPLC analysis, plasma samples were extracted through
HLB cartridges (Waters, HLB #WAT094226). Columns were conditioned with 1
mL of methanol followed by 1 mL of water. Then 2 mL of plasma sample were
added followed by 1 mL of wash solution (5% methanol in water) and analytes were
removed from column with 1 mL of 98% methanol 2% acetic acid. This solution was
diluted 1:1 with water prior to HPLC analysis. For concentrated extracts, 2 and 3
times the plasma was loaded onto the column and eluted in the same amount of
methanol/acetic acid solution.
The HPLC method was modified for the analysis of the plasma extracts. An isocratic
method with a modified mobile phase was implemented. The mobile phase changed
to 75% 10 mM ammonium acetate (Fischer) in water (pH adjusted to 3.5) and 25%
acetonitrile (Aldrich). No other HPLC method parameters were changed.
2.4 Results
AMT was first illuminated in a media matrix to determine the photoproducts in the
absence of plasma proteins. Figures 2.1 and 2.2 are representative chromatograms of
15


AMT in media. Table 2.1 contains the areas and retention times for peaks observed
in the chromatograms. Only those peaks that appeared or significantly increased post
UV exposure were listed in Table 2.1.
16


Figure 2.1 AMT in Media without Illumination
AMT in Media no Illumination
This figure shows the chromatogram for AMT in media with out illumination. AMT
elutes at 3.5 minutes.
17
13.0


Figure 2.2 AMT in Media with Illumination
AMT in Media Illumination
This figure shows the chromatogram for AMT in media with illumination. AMT
elutes at 3.5 minutes. Degradation peaks appear at 3, 4.6 and 6 minutes. The
absorbance scale on the Y-axis is much smaller than that found in Figure 2.1. The
range in Figure 2.2 is from 0 to 70 milliabsorbance units (mAU).
18


Table 2.1 AMT in Media
Sample 3 Retention Time (minutes) 3.5 4.6 6
Blank
Not illuminated 1888.58
Illuminated
1 255.45 182.88 56.75 23.63
2 259.09 182.80 66.61 23.81
3 279.30 192.02 93.24 29.51
4 269.71 158.65 76.51 24.25
5 268.11 132.93 85.63 29.22
6 259.94 108.96 72.95 22.53
7 245.24 121.51 63.00 17.62
8 267.13 206.90 95.82 24.18
9 272.72 192.36 91.13 23.64
10 278.30 221.85 99.31 25.16
This table contains the areas and retention times of pertinent peaks found in AMT
media chromatograms. Samples were prepared and analyzed on 6/26/00. The UVP
illuminator was used and all samples received a dose of 3 J/cm2. The sample labeled
Blank did not contain AMT and was not exposed to UV light. The sample labeled
Not illuminated contained AMT but was not exposed to UV light. The samples
labeled 1 through 10 were replicates containing AMT and were exposed to UV light.
The AMT peak appears at 3.5 minutes. Three additional peaks appear post-
illumination. They elute at 3, 4.6 and 6 minutes.
19


The AMT elutes at 3.5 minutes when using the hexanesulfonic acid method. This
peak has the largest area counts in the control and the UV profile for peak eluting at
3.5 minutes matches literature spectrum for AMT. Figure 2.1 is a spectrum of AMT
spectra obtained from the literature30 and Figure 2.2 is a spectrum of the peak
appearing at 3.5 minutes obtained from HPLC analysis. Psoralens exhibit three
characteristic regions of absorption maxima: the 225 region, the 250 nm region and
the 300 nm region.
20


Figure 2.3 AMT Spectrum from Literature
WAVELENGTH, nm
This spectrum only shows the absorbance profile from 275 to 400 nm for AMT. The
absorbance maximum is shown at 300 nm. The absorbance profile slopes to zero
before 400 nm.
21


Figure 2.4 AMT Spectrum from HPLC Analysis
UV Spectrum of AMT
Wavelength (nm)
In contrast to the AMT spectrum from the literature, this figure also shows the AMT
spectrum below 275 down to 200 nm. Absorbance maxima are observed at 210, 250,
300 nm. The absorbance profile slopes to zero before 400 nm.
22


Figure 2.5 AMT in Plasma Extract without Illumination
AMT in Plasma-Control
This figure shows the AMT peak eluting at 5.5 minutes.
23
11.8


At>*orinc* (mAU)
Figure 2.6 AMT in Plasma Extract with Illumination
AMT in P lasm a*IHum Inated
Tim*
This figure shows the AMT peak eluting at 5.5 minutes and 4 degradation products
eluting at 3.7, 4, 7 and 14 minutes.
24


Table 2.2 AMT in Plasma
Sample Retention Time (minutes) 3.7 4 5.5 7 14
Blank
Plasma Blank
Not illuminated* 4198.8

Illuminated
1 59.2 195.8 1039.3 111.4 46.8
2 62.1 202.9 973.2 119.2 53.9
3 52.5 223.2 892.9 118.2 65.4
4 66.2 213.2 738.0 121.1 67.6
1* 26.2 95.4 1174.9 121.5 61.4
3* 30.5 98.5 797.8 140.5 85.8
This table contains the areas and retention times of important peaks found in the
chromatograms from the AMT plasma extract samples. Samples were prepared on
8/17/00 and analyzed on 10/02/00. The UVP illuminator was used and samples 1-4
received an illumination dose of 2 J/cm2. The sample labeled Blank consisted of
media that did not contain AMT and was not exposed to UV light. The sample
labeled Plasma blank consisted of a plasma extract that did not contain AMT and
was not exposed to UV light. The sample labeled Not illuminated consisted of
plasma extract containing AMT that was not was not exposed to UV light. The
samples labeled 1 through 4 were replicate plasma extracts containing AMT. The
AMT peak appears at 5.5 minutes. Four additional peaks appear post-illumination.
They elute at 3.7, 4, 7 and 14 minutes. *AMT peak in the not-illuminated sample did
was not as large as expected. To understand if the extraction process was performed
improperly, the not-illuminated sample and AMT 1 and 3 were extracted and
analyzed a second time on 2/20/00.
25


Table 2.3 AMT in Plasma
Sample Retention Time (minutes) 3.7 4 5.5 14
Blank
Not illuminated 2908.4

Illuminated
1 34.95 43.4 136.9 139.6
2 37.7 41.2 87.7 160.9
3 28.95 37.25 120.6 133.25
This table contains the areas and retention times of important peaks found
chromatograms from the AMT plasma extract samples. Samples were prepared and
analyzed on 2/20/01. The Research illuminator was used and all samples received an
illumination dose of 2 J/cm2. The sample labeled Blank consisted of media that did
not contain AMT and was not exposed to UV light. The sample labeled Not
illuminated consisted of plasma and AMT but was not exposed to UV light. The
samples labeled 1 through 3 were replicate plasma extracts containing AMT. The
AMT peak appears at 5.5 minutes. Three additional peaks appear post-illumination.
They elute at 3.7, 4, and 14 minutes. Data shown for each sample in this table is an
average of two data points. The peak at 7 minutes may not have appeared in this data
because a different illuminator was used. Smaller peak areas for degradation products
were also observed in this data set than for samples illuminated in the UVP.
26


AMT under the aqueous ammonium nitrate/acetonitrile method elutes at 5.5 minutes.
The UV profile for this peak matches that of the peak eluting at 3.5 minutes in the
hexane sulfonic acid method.
To determine the repeatability of the degradation profile in plasma using the same
illumination device, a final experiment was conducted. For this experiment samples
of AMT in media and plasma were prepared as well as blanks of media and plasma.
The injection volume for this sample analysis was increased from 20 uL to 60 uL to
improve instrument sensitivity. The data is listed in Table 2.4 to follow.
Plasma samples from Table 2.4 were extracted a second time and concentrated by a
factor of 2 or 3. This was done to minimize solvent interference in the UV profiles
for each AMT photoproducts. This would ensure a more accurate assessment of
structural characteristics depicted in the UV profile. Data from the analysis of
concentrated extracts are listed in Table 2.5.
27


Table 2.4 AMT in Plasma
Sample Retention Time (minutes)
3.7 4.0 5.4 7 14
Blank
AMT media not illuminated 9.9 76.8 3799.4
AMT media illuminated 22.7 2296.9 486.2 485.17 82.3
Plasma blank
AMT plasma not illuminated 5390.7
AMT plasma illuminated 209.7 831.9 1199.8 206.5 146.0
AMT plasma illuminated 90.1 662.5 1152.1 215.8 172.4
This table contains the areas and retention times of important peaks found in the
chromatograms for AMT plasma extracts and media samples. Samples were prepared
2/26/01 and analyzed on 3/08/01. The UVP illuminator was used to deliver a 2
J/cm2.dose to all illuminated samples. The sample labeled Blank consisted of
media that did not contain AMT and was not exposed to UV light. AMT media not
illuminated and AMT media illuminated consisted of samples containing AMT in
media with and without UV exposure, respectively. The sample labeled AMT
plasma not illuminated consisted of plasma and AMT but was not exposed to UV
light. The samples labeled AMT plasma illuminated were replicate plasma extracts
containing AMT. The AMT peak appears at 5.5 minutes. Four additional peaks
appear post-illumination. They elute at 3.7, 4, 7, and 14 minutes.
28


Table 2.5 AMT in Plasma Concentrated Plasma Extracts
Sample 3.7 Retention Time (minutes) 4 5.5 7 14
Plasma blank 60.8 36.0
AMT plasma control 12050
AMT plasma illum+ 2118.8 227.8 3498.0 866.6 394.6
AMT plasma illum+ 2252.1 265.4 3692.4 958.3 450.1
This table contains the areas and retention times of important peaks found in the
chromatograms from the concentrated plasma extracts. Samples were prepared on
2/20/01 and analyzed on 3/08/01. The UVP illuminator was used to deliver a dose of
2 J/cm2 to all illuminated samples. The sample labeled Plasma blank consisted of a
plasma extract that did not contain AMT and was not exposed to UV light. This
extract had a concentration factor of 2. The sample labeled AMT plasma not
illuminated consisted of plasma and AMT but was not exposed to UV light. This
extract also had a concentration factor of 2. The samples labeled AMT plasma -
illuminated were replicate plasma extracts containing AMT post-UV exposure.
These samples had a concentration factor of 3. The AMT peak appears at 5.5
minutes. Four additional peaks appear post-illumination. They elute at 3.7,4, 7 and
14 minutes.
29


2.5 Discussion
Data collected in these experiments consistently show 4 degradation products for all
experiments where experimental conditions including the analytical separation
method are kept constant. From the UV profiles the peaks presented in Table 2.1 can
be matched with peaks appearing in subsequent tables. The UV profiles for peaks in
Table 2.1 appearing at 3, 4.6 and 6 minutes match closely with peak profiles from
Tables 2.2-2.5 appearing at 4, 7 and 14 minutes respectively. The repeatability of the
UV profile further suggests the consistency of AMT photochemistry in this
experiment.
An evaluation of peak areas pre and post-illumination from Tables 2.1-2.5 reveal that
between 75 to 90% of the AMT is destroyed during photolysis. Comparing peak
areas of AMT in media and plasma pre and post-illumination shows that more AMT
survives in plasma than in media for the same UV dose (see Table 2.4). This suggests
that photodegradation of AMT is faster in media than in plasma. Plasma proteins
also absorb in the UV and would have the effect of retarding AMT decomposition
during illumination.
30


Peak areas for degradation products are much greater in media samples than in plasma
extracts, when UV dose and illuminator are constant. By summing all the identified
photoproduct peak areas of AMT, 90%+ of the initial AMT peak area is found in the
post illumination chromatograms from media data in Table 2.4. Only 30 to 60% of
the initial AMT peak area is present post-illumination in the plasma extracts.
Even though the same dose was given to plasma and the media samples, additional
photochemistry is occurring in plasma that has not been observed by this analytical
technique. Interactions between AMT and plasma proteins probably occur and may
account for the reduction in photoproduct generation. AMT bound to protein would
not have been captured in the plasma extract and been absent from the HPLC
analysis. A more direct assessment of plasma proteins will be necessary to determine
if AMT protein interactions occur.
Even though the concentration of AMT photoproducts in plasma is much less than in
media, the distribution of photoproducts formed during illumination remains
consistent. The peak area at 4 minutes is the largest of all AMT degradation products
accounting for 60% and 13% of the initial AMT peak area in media and plasma,
respectively. The next largest peak area for a degradation product is the peak at 7
minutes in both media and plasma. This peak accounts for 13% of the initial AMT
31


peak area in media and 4% in plasma. The peaks at 3.7 and 14 minutes account for
less than 3% of the initial AMT peak area in both media and plasma.
Clues to the structure of the AMT photoproducts can be obtained from studying the
UV profiles of each peak. Interestingly, two of the UV profiles of the AMT
photoproduct peaks are similar to profiles obtained for photoproducts of other
psoralen compounds. A psoralen similar to AMT called 4-hydroxymethyl-4,5,8-
trimethylpsoralen (HMT) was exposed to UV light at concentrations of 10 to 100 pM
for a dose of 18 to 36 J/cm2 (240-300 nm).33 The difference between AMT and HMT
is that where the former has an NH2 group the latter has an hydroxyl group.
Results of the HMT study demonstrated the formation of two psoralen dimers: a
furanrpyrone heterodimer and a pyrone:pyrone homodimer. The profile of the furan:
pyrone heterodimer is similar to the peak at 4 minutes, while the pyrone:pyrone
homodimer spectrum is very similar that of the peak at 7 minutes. None of the
spectra presented in this paper were similar to the spectra obtained for the 3.7 and 14
minute peaks.
32


Figure 2.7 Spectrum of HMT Furan:pyrone Heterodimer
This spectrum shows absorbance maxima at 210, 230, 260, and 340 nm for the HMT
furanrpyrone heterodimer. The Y-axis is presented in units of absorbance.
33


Absorbance (mAU)
Figure 2.8 Spectrum of Peak at 4 Minutes
UV Spectrum of 4 Min Peak
250
200
150
100
50
0
\
\
190
220
250
280
310
340
370
400
Wavelength (nm)
This figure shows similar absorbance maxima to Figure 2.8. They are observed at
210, 225, 260, and 330 nm.
34


Figure 2.9 Spectrum of HMT Pyrone:pyrone Homodimer
This figure shows absorbance maxima at 225 with a plateau around 2600 nm for the
HMT pyrone:pyrone homodimer. Absorbance profile slopes toward 0 between 300
and 400 nm. The Y-axis is presented in units of absorbance.
35


Absorbance (mAU)
Figure 2.10 Spectrum of Peak at 7 Minutes
UV Spectrum of 7 Min Peak
190
220
250
280
310
340
370
400
Wavelength (nm)
This figure shows the absorbance maxima at 220 with a sloping plateau at 250.
Absorbance profile slopes toward 0 about 300 nm.
36


It is interesting to note how similar the spectra for the peaks appearing at 3.7 and 14
minutes are to each other when they appear in the chromatogram over ten minutes
apart. Spectra from these two peaks may be superimposed almost exactly suggesting
the two compounds have a very similar structure. As the retention times differ so
greatly, the polarity of these compounds must be very different. The compound
appearing at 14 minutes may be a dimer of the compound eluting at 3.7 minutes. If
two molecules were linked at positions that did not alter the chromophore, such a
spectral similarity may be observed for the two compounds.
UV profiles with absorbance above 300 nn indicate extensive conjugation. A relative
absorbance maximum around 300 nm is characteristic of psoralens whereas for
coumarin that maximum is shifted up to 320. Coumarins also exhibit a maximum
around 270 nm; the psoralen maximum is usually much stronger and appears around
250 nm. The structure of coumarin is given below and an UV spectrum of the
compound is located in Figure 2.13.32
5 4
Coumarin with atoms numbered
37


Spectra for the peaks eluting at 3.7 and 14 minutes have closer similarities to the
coumarin spectrum than the AMT spectrum (Figure 2.4). Photochemistry
surrounding formation of coumarin products from AMT would involve rupture of the
furan ring.
Data presented in Chapter 2 has given many clues as to the number of products
obtained during illumination as well as some of the structural characteristics of those
photoproducts. Key to the identification of some of these photoproducts will be the
determination of the masses associated with the compounds. Mass spectrometer data
will be presented in the next chapter.
38


Figure 2.11 Spectrum of Peak at 3.7 Minutes
UV Spectrum of 3.7 Min Peak
This figure shows three absorbance maxima: 210, 264, and 345 nm.
39


Absorbance (mAU)
Figure 2.12 Spectrum of Peak at 14 Minutes
UV Spectrum of 14 Min Peak
This figure shows three absorbance maxima: 212, 260, and 350 nm.
40


Figure 2.13 Coumarin Spectrum from Literature
The figure is presented in units of absorbance on the Y-axis and wavelength (nm) on
the X-axis. In this spectrum absorbance maxima are observed at 270 and 320 nm.
41


3. Analysis of AMT Photoproducts by HPLC Mass Spectrometry
3.1 Introduction
The number of AMT degradation products was identified in Chapter 2 but only very
qualitative statements may be made about the identity of the photoproducts without
additional analysis. In mass spectrometry, the mass of an unknown compound may be
determined without comparing analytical results with known compounds for positive
identification. The mass spectrometer is calibrated for mass independently of the
specific sample being analyzed. Associating a mass to each peak identified in the
HPLC photodiode array analysis may be a powerful piece of information to determine
the structure of the AMT photoproducts generated during this sterilization procedure.
Structural characteristics gained from the UV profile and mass for each peak can also
be combined with additional information to ensure proper peak identification. The
photochemistry expected to occur in solution can be extremely powerful in assessing
the likelihood of formation for any proposed structure. A review1 of psoralen
reactions most likely to occur in an aqueous environment as well as the bonds parent
molecule which would most likely to break under UV exposure has been published in
the literature. From such information, a structure as well as a mechanism can be
42


presented to describe the AMT photochemistry occurring in the blood matrix during
illumination.
In this chapter, the masses associated with the 4 degradation products identified in
Chapter 2 will be presented. Structures of the degradation products will be
presented as well as reaction mechanisms to support formation of these structures in
solution.
3.2 Experimental
The HPLC used in this study was a 2690 separations module (Waters) equipped with
a 2487 dual absorbance detector (Waters). Following the UV detector a Quattro II
triple Quadrupole Mass Spectrometer (Micromass) was connected. The MassLynx
3.2 (Micromass) operating system was used to control and communicate between the
components. Settings for the mass spectrometer were cone voltage 5V, capillary
voltage 2.8 kV, and extractor voltage 3V. In addition the desolvation temperature
was 250C, the source block temperature was 80C, and the ionization energy was set
to 0.7 keV. This is a z-spray spectrometer and the analysis was run in positive ion
mode. In positive ion mode the molecules are bombarded with the ionized reagent
gas CH5+. Molecules are protonated under this technique. Usually 1 proton sticks to
the molecule to form a charged species (M+l)+ that can be selected in the quadrupole
magnet for detection.
43


Sample extraction and the HPLC method used for these analyses are outlined in detail
in chapter 2. The same sample extracts which were run for HPLC photodiode array
analyses were also run for mass spectrometry analysis.
3.3 Results
Retention times for the same peaks were shorter by 0.5 minutes in the HPLC mass
spectrometry analysis than for the HPLC photodiode array analysis. For reasons of
clarity, the retention times for the peaks presented in Chapter 2 will be used for the
mass data instead of the retention times observed in this analysis. Figures 3.1 and 3.2
contain representative chromatograms for this analysis.
44


Figure 3.1 Representative Chromatogram of Illuminated Sample
2000Q753
100-,
3 Q.34
%
2000Q753
4.00
2.97
1.78 3.44
100-,
%A
j 0.96 3.80
2000Q753
100-,
%-
2000Q753
100-,
2.76 3.54
3.85
6.12
ri
5.40 ; ;
A / \ 6-79 7.57
5.40
9.029.64 10.57
Scan ES+
515
3.74e
11.961.2'37 13.6214.44
Scan ES+
258
9.93e6
i \ g 54
J V__ \ _7.11 _8-09 11.08.41.34_12.32 ___________
Scan ES+
5.40 241
A 7.39e7
Scan ES+
248
2.17e6
%
0.18 3.54 ; 6 18 -r a* 8.81 Q 79
1.78 2.61 4.42 5.76 9.27 9 /9 1Q.98
2000Q753
100,
4.00
%-j
Qi
6.18
2000Q753
100-,
%-
0-t==
2000Q753
2.36
3.07
3.25 3,g2 4.90
5.67
i\
12.27 12-7413.9214.7o
Scan ES+
498
1 46e7
254nm
An1
10.00e5
"13.07
This figure shows the mass traces for 515, 258, 241, 248 and 498 m/z. The last signal
shown in this chromatogram is the UV trace collected at 254 nm.
45


Figure 3.2 Representative Chromatogram of Illuminated Sample
20000753
100-
Scan ES+
13.56 290
A 2.31e6
i\
/ ) 13.77
1.632.56^2.82 3.J
5.04 6.18 7.31
7.67 9.07 Q __ _ 12.99
9.58-10.31 11.70
2000Q753
100-.
0.80 1-21
2.71
\ 2.82
!!!
Scan ES+
601
2-75e5
13.51
12.53 l
fl 14.60
, 2.51 3.33 3.90 S 6.23 8.ai 1036 i J l 14j6
U \ ^ i8-1^9.27i A /p3* /W'i3/e iU14"
. .a w^V- ifi''n-y W t V'
2000Q753
Scan ES+
This figure shows additional mass traces for the sample in Figure 3.1 at 290 and 601
m/z. The last signal shown in this chromatogram is the UV trace collected at 254 nm.
46


Table 3.1 Masses by Retention Time
Sample 3.7 4 5.5 7 14
Not Illuminated 241

Illuminated
AMT 1 248 498, 515 241, 258, 515 498,515 290
AMT 2 248 498, 515 241, 258, 515 498,515 290
AMT 3 248 498, 515 241, 258, 515 498,515 290
AMT 4 248 498, 515 241, 258, 515 498,515 290
This table shows the mass signals associated with the peaks observed in the UV
chromatogram. All masses listed have units of m/z and are an M+l positive ion. The
data in this table were obtained from samples prepared on 8/17/00.
47


Table 3.2 Masses by Rel tention Time
Sample 4 5.5 7 14
AMT media not illuminated 241, 258, 299, 515
AMT media illuminated 498, 515 241, 258, 515 498, 515 290
AMT plasma illuminated 498 241, 258, 515 498, 515 290
AMT plasma illuminated 498 241, 258, 515 498,515 290
AMT plasma not illuminated 241, 258, 299, 515
This table shows the mass signals associated with the peaks observed in the UV
chromatogram. All masses listed have units of m/z and are an M+l positive ion. The
data in this table were obtained from samples prepared on. 02/26/01.
48


Figure 3.3 Chromatogram of Concentrated Plasma Extract Illuminated Sample
7 (3X) Renee Williams (02/26/01)
01Q254
100-,
%-
0-
1.422.30 , 333

13.62
7.52 8.34 8.50 10:00 10.93 12;27
Scan ES+
601
6.47ef
14.08
01Q254
100t
%-
oi-
01Q254
100-,
3 33 3.75
*1 I ''1-Tl ITI r
5.07
%-
0-
3.13
3.76
2.37

6.03
__/V
6.47
/
Scan ES+
13.67 29C
f\ 1.15e7
I \
254nrri
An1
5.61 e5
13.43
.'TN .
This figure shows the appearance of a mass signal at 601 m/z for the peak eluting at
14 minutes in the concentrated AMT plasma extract. The mass signal at 290 m/z also
increases for the same peak in the concentrated extract. The last signal shown in this
chromatogram is the UV trace collected at 254 nm.
49


Table 3.3 Masses by Retention Time or Concentrated Samples
Sample 3.7 4 5.5 7 14
AMT plasma illuminated 248,231 498, 578 241, 258, 515 515 290, 331, 601
AMT plasma illuminated 248, 231 498, 578 241, 258, 515 515 290, 331, 601
AMT plasma not illuminated 241, 258, 299, 515
This table shows the mass signals associated with the peaks observed in the UV
chromatogram for the concentrated plasma extracts. Additional mass signals of
interest appear for the peak eluting at 14 minutes. All masses listed have units of m/z
and are an M+l positive ion. The data in this table were obtained from samples
prepared on. 02/26/01.
50


3.4 Discussion
In the control, three mass peaks are observed for the AMT peak: 241 (M + 1), 258
(M + 1) and 515 (M+l) m/z. A fourth mass of 299 m/z appears for the control
samples in tables 2 and 3. The molecular weight of AMT is 257. The mass at 258
m/z is the molecular ion peak of AMT. The strong signal at 241 is 17 mass units less
than 258 suggesting a loss of an OH group from the molecule. This indicates that the
most stable ion form of AMT in the mass spectrometer has one less hydrogen and
oxygen on the molecule. The mass at 515 is double the mass at 257 (molecular
weight of AMT) and indicates dimer formation within the mass spectrometer. The
299 m/z only appears in the AMT samples when the concentration of the molecule is
high. This suggests that the molecule is combining with another molecule in the mass
spectrometer and is not a true representation of the mass of the molecule. Acetonitrile
constitutes 25% of the mobile phase and has a molecular weight of 41. Adding 41
mass units to 258 equals 299 explaining the appearance of the 299 m/z peak.
Reviewing the illuminated samples, a mass of 515 consistently corresponds to peak
eluting at 7 minutes. In chapter 2, it was shown that the UV profile of the peak
eluting at 7 minutes matched that of a psoralen homodimer presented in the
51


literature.33 As stated above a mass of 515 (M+l), is the mass of a molecule twice the
weight of AMT. This evidence confirms that the peak eluting at 7 minutes is an AMT
dimer. By UV profile and the mass data the peak eluting at 7 minutes has been
identified as the AMT pyrone:pyrone homodimer.
Also presented in chapter 2 was the similarity between the UV profile of the peak
eluting at 4 minutes and the furanrpyrone heterodimer. The main mass associated
with this peak is 498 m/z. Interestingly, this mass is 17 units less than 515. A loss of
17 mass units was also seen for the major peak of AMT. The pyronerpyrone
homodimer does not have this same phenomenon. The difference between the
p)Tone:pyrone homodimer and the other two molecules is that the both pyrone ring
moities are protected through dimer formation. In the furan:pryone dimer only one
pyrone moiety is protected. AMT also has an unprotected pyrone moiety. The
reduction of the molecular mass in the mass spectrometer must be associated with the
rupture of the pyrone ring and subsequent loss of OH. Using this reasoning, the peak
eluting at 4 minutes has been identified as the AMT furan:pyrone heterodimer by both
UV spectral match and mass.
A paper reviewing photooxidation studies on 8-methoxypsoralen characterized a
dimer link at the C3 position of the molecule.25 The C3 position of AMT is not
52


substituted and is a possible site for a dimer linkage. Psoralens are also known to
dimerize at the 45 double bond of the furan ring. For both dimers, [2+2] cyclo-
addition is the most probable reaction mechanism for cyclobutane ring formation
between the AMT molecules. Figures of the dimers are presented below. These
structures were adapted to AMT from dimers presented for HMT.33
o
AMT Pyrone:Pyrone Homodimer
53



AMT Furan:Pyrone Heterodimer
A 248 m/z is associated with the peak eluting at 3.7 minutes. A structure has been
presented to have a mass of 247 below. This structure is formed by the scission of the
4,5 double bond in the furan ring. Rupture of this ring at the 4,5 position has been
reported for other psoralens. It is thought that this reaction occurs via a dioxetane
formation involving singlet oxygen attack to the double bond. Substituting electron
rich groups such as methyls on to the double bond enhances this reaction. Both the 4
and 5 positions are substituted with such groups. This molecule would have
increased hydrophilic character due to the hydroxyl group and would be expected to
elute before AMT. This mass peak appears before the AMT masses and is associated
with the peak at 4 minutes in the HPLC chromatogram. The proposed reaction
mechanism involving singlet oxygen attack of AMT is given below.
54


Figure 3.4 6-Aminoacetyl-7-Hydroxy-4,8-Dimethylcoumarin Formation
Singlet oxygen is presented in this figure attacking the psoralen molecule causing the
furan ring to rupture in formation of a coumarin compound.
55


Mass 248 does not appear in the data presented in table 3 as the 3.7 minute peak is too
weak. When samples 6 and 7 are concentrated a mass of 248 as well as 231 appears
for the peak. The pyrone ring is unprotected in this molecule as well and may rupture
upon entry into the mass spectrometer. The repeatable retention time, mass data and
UV profile for this peak observed in chapter 2 demonstrates that the same compound
has been generated in both data sets. The peak eluting at 3.7 minutes has been
identified as 6-aminoacetyl-7-hydroxy-4,8-dimethylcoumarin.
ch3
6-Aminoacetyl-7-Hydroxy-4,8-Dimethylcoumarin
A mass of290 m/z is associated with the 14 minute peak. A weaker mass of 331 m/z
appears in the data presented in Table 3.5 for the concentrated extracts. A mass of
331 is 41 mass units higher than 290 and is consistent with acetonitrile combining
with the 290 molecule as was seen for the AMT molecule. A mass of289 does not
correspond with any known psoralen degradation product. The structure of this
compound probably has similarities to 6-aminoacetyl-7-hydroxy-4,8-
dimethylcoumarin as the spectral profile for the two peaks is so similar. The
56


concentrated samples also show a small peak at 601 m/z for the peak eluting at 14
minutes. This mass is probably closer to the molecular weight of the molecule but a
structure with the mass of 600 atomic weight units (AMU) cannot be determined from
6-aminoacetyl-7-hydroxy-4,8,-dimethylcoumarin. Additional characterization of this
compound will be required for positive identification.
With 3 of the 4 major degradation products identified, experiments will be reviewed
in Chapter 4 that begin to assess the AMT/protein interactions. From the data
presented in Chapter 2 a significant portion of the AMT, on the order of 50%, can be
expected to interact with the plasma proteins.
57


4.
Analysis of AMT Treated Plasma Proteins
4.1 Introduction
With the major AMT photoproducts identified in Chapters 2 and 3, the focus of the
research will now turn to the protein component of the samples. A cursory attempt at
a mass balance of AMT in plasma post illumination was presented in Chapter 2.
From that analysis it was estimated that approximately 50% of the AMT was missing
post UV exposure. Since the samples were not inoculated with pathogens and the
plasma had been tested for known infectious agents, AMT must have additional
targets within plasma besides the DNA. The major component in plasma besides
water is proteins. Research studies have been conducted to determine if psoralens
will also bind to proteins such as bovine serum albumin (BSA) and human serum
albumin (HSA) in addition to DNA.33"36
Strong associations between proteins and psoralens have been documented in the
literature which occur without light energy for excitation.33,34 In photochemistry, this
non-covalent association is often called dark binding. Dark binding of psoralens and
proteins has been strong enough to interfere with the determination of covalent
binding of psoralens to proteins. High affinity dark binding has been observed
between lipoproteins and furocoumarins and has been measured by several analytical
58


techniques including gel exclusion chromatography, difference spectrometry,
33
equilibrium dialysis of radio-labeled furocoumarins, and fluorescence quenching.
Results from the fluorescence quenching experiments suggest that the tryptophan
residue of human serum albumin is involved in binding with psoralens indicating 1
psoralen binding site per molecule of human serum albumin.
Under the influence of light, enough energy is present for psoralens to form covalent
bonds with proteins. Photobinding of psoralens is postulated to occur by two separate
pathways. Either the excited furocoumarin binds directly to the protein or an excited
furocoumarin forms, probably via an excited singlet oxygen mechanism,
photoproducts that covalently bind to the protein. This theory was presented after
reviewing data where 8-methoxypsoralen in combination with BSA was illuminated
in the presence and absence of oxygen with varying molar ratios and illumination
times. Binding was observed both where 8-Methoxypsoralen was pre-irradiated or
irradiated in the presence of BSA.
In an effort to determine the extent of AMT binding to plasma proteins during the
sterilization process, an analysis of the plasma proteins in conjunction with
radiolabeled AMT was designed and executed.
59


Structure of Tritiated AMT
4.2 Experimental
HPLC used was a 2690 separations module (Waters) equipped with a 441 UV
detector (Waters). In series behind the UV detector was a flow scintillation analyzer
(Packard, 500TR series Ser #423464). Flow 1 (Packard, Ver 3.6.1) software was used
for the operation of the equipment and data manipulation. The chromatography
achieved using a size exclusion column (Zorbax GF250, Hewlett Packard, P/N
884973-901). This stationary phase is silica based and has diol linkages. The silanol
groups are clad with zirconium to reduce any analyte charge interactions with the
stationary phase. The significant interaction between the analytes and the stationary
phase is through entry into the pores. Therefore retention time is determined by size.
Mobile phase consisted of 130 mM sodium chloride, 20 mM potassium chloride, and
50 mM sodium phosphate, at pH 7. The flow rate was 1 mL/min and the column
temperature was 30C. The detector wavelength was set to 210 nm. Analysis time
was 20 minutes and injection volume was 100 uL.
60


Samples were prepared as outlined in Chapter 2 with the following exceptions. Two
one mL aliquots of tritiated AMT (Cerrus, 0.4 C/mmol) were added to 10 mg of AMT
(Sigma, #A4330). At the time of the last analysis in 1995 the activity of the AMT was
0.6 mC/mL. Therefore approximately 1.2 mC of AMT was added to the media
mixture. The concentration of AMT in the final plasma samples was 250 pM.
4.3 Results
Representative chromatograms of this analysis are presented in Figures 4.1 to 4.3.
Scintillation counts from the sample chromatograms are listed in Table 4.1.
61


Figure 4.1 Chromatogram of Plasma Blank
This figure contains the chromatogram from the analysis of plasma without AMT
present and without UV exposure. The radioactivity counts with respect to time are
shown in the top graphic with the UV chromatogram underneath. The UV
chromatogram was obtained at 254 nm. The Y scale for the radioactivity counts
ranges from 0 to 13 and shows the noise level of the detector. A large protein peak is
observed in the UV chromatogram beginning at 1.5 minutes.
62


Figure 4.2 Chromatogram of AMT in Plasma-Not Illuminated
Radiomatic 500TR v3.60/3.60 S/N:423464 User: SUE RunFile: RENE0003 Date: 11/10/2000 10:51AM Page:l
Report Format File: INT_PRT
This figure contains the chromatograms from the analysis of plasma with AMT
present and without UV exposure. The radioactivity counts with respect to time are
shown in the top graphic with the UV chromatogram underneath. The UV
chromatogram was obtained at 254 nm. The Y scale for the radioactivity counts
ranges from 0 to 1970. A large protein peak is again observed in the UV
chromatogram beginning at 1.5 minutes.
63


Figure 4.3 Chromatogram of AMT in Plasma-Illuminated
Radiomatic SOOTH v3.60/3.60 S/N:423464 User: SUE RunFile: RENE0004 Date: 11/10/2000 10:53AM Page:l
Report Format File: INT_PRT
This figure contains the chromatograms from the analysis of plasma with AMT and
after 2 J/cm2 of UV exposure. The radioactivity counts with respect to time are
shown in the top graphic with the UV chromatogram underneath. The UV
chromatogram was obtained at 254 nm. The Y scale for the radioactivity counts
ranges from 0 to 1516. A large protein peak is again observed in the UV
chromatogram beginning at 1.5 minutes.
64


Table 4.1 Scintillation Detector Results from Plasma Protein Analysis
Sample Description Counts by peak retention time 2.5 min 3 min 5 min 11 min Total counts %Bound
1,8 saline blank Nd Nd Nd Nd
2,9 plasma blank Nd Nd Nd Nd
3,10 Not illuminated Nd 2050 1308 42226 45584 7.37
4,11 plasma 1 1639 6801 2169 7373 17982 59.00
5,12 plasma 2 1651 6441 2366 9294 19808 53.08
6,13 plasma V def 1620 6599 1322 7268 16808 56.76
7,14 plasma VUI def 2167 5688 1656 6929 16440 57.85
This table contains the scintillation counts from pertinent peaks in the analytical
chromatograms. The sample labeled saline blank consisted of salt solution without
AMT or plasma and was not illuminated. The sample labeled plasma blank did not
contain AMT and was not exposed to UV light. The sample labeled Not
illuminated contained radioactive AMT but did not receive a dose of UV light. The
remaining samples labeled plasma 1,2 V def and VIII def all received an UV dose
of 2 J/cm Plasma 1 and 2 were replicate samples while plasma V def and plasma
VIII def were plasma samples deficient in factor V and factor VIH, respectively.
The term Nd indicates a peak was not detected. The total radioactivity counts
decrease by over half post-UV exposure. It not clear why this has occurred but may
indicate that the tritium atom may be released from the AMT molecule during
illumination.
65


4.4 Discussion
Several interesting observations maybe noted upon initial review of the data
contained in Table 4.1. The peak appearing at 11 minutes is the unbound AMT. This
peak has the largest peak area in the chromatogram and is the last to elute. AMT is
much smaller than the proteins and unbound AMT should elute later than AMT
associated with proteins in the control sample. The observation of two additional
peaks besides the unbound AMT peak indicates that dark binding is detectable by this
technique. This fact is very surprising as some protein denaturing is expected under
these analytical conditions. AMT protein associations have remained intact
demonstrating a strong affinity with at least one, maybe two proteins.
The radioactive peaks at 2.5, 3 and 5 minutes are believed to be associated with
proteins within the plasma. It is difficult to determine exactly which proteins are
responsible for the binding unless experiments were conducted with proteins
individually to confirm both retention time and AMT association. The UV trace
shows one large peak with a long tail beginning about 1.5 minutes. Plasma proteins
range in size from 50,000 to 900,000 daltons (see Table 4.2). The column used in this
analysis is capable of achieving baseline separation between proteins with such a
large span in size. The fact that most of the sample elutes at the void volume
66


indicates that the proteins are not interacting with the stationary phase, as intended.
They do not appear to be entering the stationary pores of the column, which would
retain the proteins longer and extend the time before they would be detected.
67


Table 4.2 Prevalent Human Plasma Proteins
Protein Size (Daltons) Mean Concentration (mg/dl) % of Total Plasma Protein
Albumin 65,000 4200 64
IgM 900,000 1,100 17
Fibrinogen 340,000 350 5
Coagulation Factors 56,000-330,000 16 >1
IgG 160,000 366 5
Transferrin 80,000 280 4
IgA 170,000 240 4
Data presented in this table was obtained from a hematology reference book.37
68


One reason for such poor separation and low interaction with the column may be
protein denaturing. If the proteins become denatured during the analysis they would
not enter the size exclusion pores and would have a much lower retention time than
anticipated for their size. Using this analytical method, HSA is expected to have a
retention time closer to 8 minutes. No protein peak is observed so late in the UV
chromatogram.
To achieve more definitive results the analytical method should be refined to ensure
the tertiary and quaternary integrity of the proteins is maintained throughout the
analysis. In this manner, proteins could be separated and identified by retention time.
In addition, radioactive peaks could be aligned with more defined peaks in the UV
trace.
The information presented in Table 4.1 demonstrates the amount of AMT bound to
protein increases from 7% in the control to over 50% in the illuminated samples.
Another study published on covalent AMT/protein binding following UV
illumination reported only 36% AMT bound.23 Ultrafiltration was used for detection
in this study. In this technique the sample is passed through a pore filter prior to the
radioactivity assessment. The filter is designed to trap large molecules in the upper
chamber while allowing small molecules to pass through. Unbound AMT would pass
69


through the filter. Since over half of the radioactivity is not observed post-
illumination, some of the AMT/protein adduct may not been measured by the
ultrafiltration technique. Based upon the information presented in this study, the ratio
of AMT bound versus unbound to protein appears to be greater than 36% and may
have been underestimated by ultrafiltration analysis previously reported.
In Chapter 2 it was postulated that at least half of the AMT was not accounted for
post illumination. AMT strongly or covalently associated with protein was eliminated
from sample during the extraction process. Plasma proteins are too large and
hydrophilic to stick to the solid phase extraction cartridge. As a result they are
washed away prior to the final elution step which captures all the small hydrophobic
molecules in the sample. Data presented by this analysis corroborates information in
Chapter 2 accounting for the missing AMT. AMT interactions with plasma protein
are very significant accounting for half the initial concentration within the sample.
Psoralens have been shown to interact with albumin in previous studies. These
studies suggested that 80 to 90% of all psoralen binding to protein is with albumin. It
is likely that a significant portion of AMT protein binding in plasma occurs with
albumin as well. The peak eluting at 5 minutes accounts for 60% of the bound AMT
signal. Due to the reproducibility of the peak areas and retention times in all samples
70


it is not likely that albumin is present in all three protein peaks. It is very reasonable
that 40% of the bound AMT is with proteins other that albumin.
Albumin accounts for 60% of plasma protein and performs the function as a non
specific protein carrier. Albumin is a very versatile protein can interact with a variety
of different molecules requiring transport in the blood stream. Due to its function and
prevalence in the blood it is likely that AMT associates with this protein in plasma
and may become irreversibly bound during the sterilization process. It has been
suggested that AMT may have affinity for tryptophan. Human serum albumin
(HSA) has one tryptophan residue in its active form and is a potential AMT binding
site.38
The purpose for analyzing the factor deficient plasma as to understand if coagulation
factors were interacting with AMT. When either factor V and VEH were missing from
the plasma the counts in the 5 minute peak appear to decrease. Additional replicate
analyses would be required to determine if the decrease is significant. Any
assessment of coagulation factor/AMT interaction in the plasma will be difficult as
they constitute less than 1 % of plasma protein. If a source of individual coagulation
factors could be located, examination of the AMT interaction could be carried out
directly with these proteins. Factor V and VHI deficient plasma was chosen as those
71


factors have a wider therapeutic benefit. This initial data indicates that the AMT may
interact with these factors. The decrease in therapeutic effectiveness of factor VIII in
chapter 4 also suggests and AMT interaction. This data is not definitive however, and
direct analysis is recommended for solid confirmation.
The data presented in this study demonstrates that half of the AMT added to plasma is
bound to protein post illumination. Although possible AMT targets were explored in
this exercise no definitive information was supplied to identify which proteins have
formed AMT adducts. Prior to the use of such a sterilization process on blood
products in the hospital or clinic additional studies are recommended to determine the
major AMT bound proteins and the impact those adducts may have on the therapeutic
benefit of the plasma product.
Protein quality data will be presented in Chapter 5 to assess the impact of AMT and
the illumination process upon the coagulation proteins within the plasma product.
72


5.
Analysis of Protein Function of AMT Treated Plasma
5.1 Introduction
As plasma has many therapeutic applications, any process intended to enhance the
safety of the blood product by reducing disease transmission must also maintain the
efficacy of the product for the clinical setting. One of the main benefits plasma
supplies to the medical community is in the treatment of clotting disorders such as
hemophilia. There are 12 proteins called clotting factors involved in the complex
process of clot formation.37 The clinical manifestation of hemophilia can be caused
by a deficiency in one of several or a combination of clotting factors. The type of
hemophilia a patient may have is dependent upon which proteins do not function
properly causing the clotting cascade to break down. For treatment, specific clotting
factors from as many as 1000 different units of plasma are combined and
concentrated.10 Patients receive a life sustaining dose of the specific factors they need
at regular intervals from the concentrated supply.
If the sterilization process reduces the functionality of the clotting factors below a
clinically acceptable level, implementing the process would not be a practical solution
to the disease transmission dilemma. Factor quality of the plasma must be assessed
before and after AMT administration and before and after UV exposure. In this
73


chapter the coagulation protein activity experiments and data are presented. For this
study coagulation factors with the most therapeutic benefit were examined
individually. This data will not directly determine which proteins may have AMT
interactions but will indicate which proteins are most affected by the sterilization
process. In addition, the data collectively will assess the overall impact of the process
on the plasma.
5.2 Experimental
Ten illuminated plasma samples in addition to a control and a blank were prepared as
outlined in Chapter 2. These samples were not extracted prior to analysis. The blank
consists of plasma without AMT added and without illumination. The Not
illuminated sample contained AMT but was not exposed to illumination. Samples
y
were irradiated for a dose of 3 J/cm or 5 J/cm Following illumination, 2mL
aliquots of each sample were submitted for protein activity assessment.
Coagulation analyzer (AMAX Sigma 190 Plus) was used to determine clotting
factor activities within plasma samples. Sample run included analysis of controls to
ensure instrument was operating within specifications and that results were accurate.
The coagulation cascade is complex and includes many reactions. The proteins
analyzed in this experiment include antithrombin, prothrombin, and various factors
from 2 to 11.
74


Coagulation factors are assessed one at a time to determine their activity. An aliquot
of each plasma sample is placed in a sample cup for the instrument. The instrument is
equipped with various reagents depending upon the coagulation factors of interest to
be tested. The instrument takes a small portion of the sample and introduces it into a
reaction well. Then a reagent plasma, deficient in the factor being tested, is
introduced into the reaction well. When the starter reagent is added to the reaction
well to initiate the clotting cascade the timer begins counting.
There are two different detection methods employed to determine the clotting time.
For most factors the optical detection method is used. A beam of light is projected
through the reaction well. Once the light transmission through the cell has decreased
by 10 to 30 milliequivalence (me) absorbance units clot formation is complete. In the
case of F2 or fibrinogen, a mechanical method of detection is used. At the initiation
of the analysis, a metallic ball inside the reaction vessel is in contact with a magnet.
As the clot forms the metallic ball is pulled away from the magnet. The end of clot
formation is signaled once the metallic ball is no longer in contact with the magnet.
For both detection methods, the clotting time for each protein is compared to nominal
values and presented as a functionality percentage of the nominal value. In the case
75


of F2 the clotting time is converted to mg/dL through the use of a standard curve.
The longer it takes to reach the specified endpoint signal the lower the protein quality
in the sample.
5.3 Results
Data for this experiment are located in Table 5.1
76


Table 5.1 Protein Function Data
Sample Pro- thrombin (%) (F2) (mg/dL) FV (%) FVII (%) FVIII (%) FX (%) FXI (%) Anti- thrombin (%)
blank 98 34 27 26 25 37 33 36
Not illuminated 103 35 28 21 22 38 27 34
3 J 1 76 30 20 14 10 26 19 34
3 J 2 67 31 22 14 11 28 19 33
3 J 3 82 31 21 14 12 27 19 35
3 J 4 75 30 19 14 11 26 19 33
3 J 5 76 30 21 14 11 27 19 34
5 J-1 53 26 16 11 8 21 16 35
5 J-2 78 28 21 14 11 26 19 34
5 J 3 85 32 21 15 11 29 20 37
5 J 4 78 29 21 14 10 26 19 33
5 J 5 77 28 21 14 11 27 19 35
Avg3 J 75 30 21 14 11 27 19 34
% blank 77 89 76 54 44 72 58 94
Avg 5 J 74 29 20 14 10 26 19 35
% blank 76 84 74 52 41 70 56 97
This table shows the results of protein quality assessments of coagulation factors from
plasma samples with and without AMT as well as with and without UV exposure.
The sample labeled blank did not contain AMT nor was it exposed to UV light.
The sample labeled Not illuminated contained AMT but was not exposed to UV
light. The samples labeled 3J 1-5 contained AMT and received an UV dose of 3
J/cm2 while the samples labeled 5J 1-5 also contained AMT and received an UV dose
of 5 J/cm2. F2 also known as fibrinogen is reported in mg/dL instead of %. F2 is
commonly reported with these units and only reflects the concentration of active
protein within the sample. Averages of the 3 J and 5 J samples were calculated and
divided by the blank values to determine the impact of the sterilization process on the
coagulation factors. Average values were listed as a percentage of the blank values in
Table 5.1.
77


5.4 Discussion
A review of Table 5.1 shows some change in activity for all factors tested post
illumination. Interestingly, the control also showed some moderate decline in factors
VII, VX1 and maybe VIII. AMT may combine with factors to reduce protein function
or initiate reactions that reduce coagulation performance. Additional samples would
be required to determine a statistically significant decline.
In the post illumination samples a 50% decrease in protein function was observed in
factors VUI, VH and XI. The effect of illumination upon antithrombin was very small
(3%). For the blood banking industry, a reduction in activity of 50% is considered
significant and defines the lower acceptable activity limit. All other factors showed a
decline of 15 to 25% in activity. Of all the factors examined in this study, a decline in
factor Vm would be expected before any of the other proteins as it is the most labile.
This factor is also one of the most therapeutically useful as a majority of hemophiliacs
have a factor VIII deficiency.
There does not seem to be a difference between the samples illuminated for 3 and 5
J/cm This result suggests the protein damage may have occurred rapidly upon the
initiation of UV exposure.
78


The data in this study show the sterilization process to have negative impact upon
protein quality. The impact of AMT alone has not been definitively shown but any
impact is moderate. Factors VII, Vm and XI suffered the most damage and may not
be considered of therapeutic use post sterilization. All remaining proteins, except
antithrombin, suffered some damage but activity levels were higher than for factors
VII, Vm and XI. Antithrombin did not show a significant decrease in activity level
post sterilization.
The remainder of the thesis will focus on the toxicity issues associated with AMT.
Chapter 6 will review the results of a cytotoxicity screen performed on AMT in media
pre and post illumination.
79


6.
Cytotoxicity of AMT
6.1 Introduction
To obtain a general understanding of AMT and AMT photoproduct toxicity a direct
contact cytotoxicity test was performed. A cytotoxicity test is designed to determine
the biological reactivity of mammalian cell cultures following contact with the
material under test. Extreme measure of the test is to determine the extent of
apoptosis or cell death that may occur as a result of contacting the material.
Cytotoxicity is determined by microscopic evidence of malformation, degeneration,
sloughing or lysis of cells, or a moderate to severe reduction in cell layer density.
An immortalized murine cell line is used in this study. A monolayer of cells is
uniformly spread across the test plate. The substance to be tested is applied topically
to the test plate. This test is called out in the United States Pharmacopeia 24 and is
titled Biological Reactivity Tests, In Vitro <87>.39
For this test, cells are inspected periodically after administration of the test solution.
A score greater than 2 constitutes a test failure. Table 6.1 outlines the grading scale
used for this test as well as the criteria for determining each grade.
80


Table 6.1: Reactivity Grades for Direct Contact Test
Grade Reactivity Conditions of all cultures
0 None Discrete intracytoplasmic granules, no cell lysis
1 Slight Not more than 20% of cells are round, loosely attached and without intracytoplasmic granules; occasional cell lysed cells are present
2 Mild Not more than 50% of the cells are round and devoid of intracytoplasmic granules; extensive cell lysis and empty areas between cells
3 Moderate Not more than 70% of the cell layers contain rounded cells and/or are lysed.
4 Severe Nearly complete destruction of the cell layers.
81


6.2 Experimental
Strain L929 mouse fibroblast cells were cultured up to a confluent monolayer on a 35
mm diameter plate and exposed 2 mL of media sample. Observations were made at
not less than 24 and 48 hours for toxicity. Extracts were diluted 1 part AMT sample
to 2 parts of commercial cell media prior to addition to the cell layers.
AMT was prepared in media as outlined in Chapter 2. Concentration of AMT in
media was approximately 170 pM. One 30 mL aliquot of AMT in media was
exposed to 3 J/cm2. Two samples were submitted for cytotoxicity testing: one
illuminated sample and one non-illuminated sample.
The AMT media samples were mixed 1 part to 2 parts of commercially available cell
culture media prior to administration on the cell layers. This was done to ensure cell
lysis was not caused by an imbalance between intra and extra-cellular tonicity.
6.3 Results and Discussion
The control received a scale of 4 after 24 and 48 hours while the illuminated sample
received a score of 0 after both time points.
82


A UV dose of 3 J/cm2 was chosen to produce complete photodegradation of AMT in
the hopes that a worst case concentration of photoproducts would be generated. It
was not discovered until after the cytotoxicity assessment was performed that the
additional J/cm2 of energy destroyed not only the AMT but the photoproducts as well,
Data where AMT in media was illuminated for 3 J/ cm2 showed lower levels of
2
photoproduct generation when compared to similar samples exposed to only 2 J/cm
(see Tables 2.1 and 2.4).
AMT dimer photoproducts were present 1/5 and 1/10 the concentration when
illuminated for the longer dose. Heterodimer constitutes 60% of initial AMT
concentration in the 2 J/cm2 samples but in this sample accounted for only 14%.
Homodimer accounted for 12% of initial AMT concentration in the samples with
lower dose but only present at 1% after 3 J/cm of exposure. Thecoumann
compounds were also reduced significantly. Since the two coumarin compounds
compare with less than 3% of the initial AMT concentration their toxicological
contribution to the sample would expect to be low in a worst case scenario.
However, the AMT control still produced a toxic response in the murine cells even
after dilution with the commercial cell culture media. Therefore AMT at a
83


concentration of 57 uM is still considered toxic by the results of this test but a
concentration of less than 6 uM does not produce a toxic response in this cell line.
AMT has also demonstrated mutagenic behavior in the absence of UV light. A
23
mutagenicity study was performed on platelet suspensions containing AMT.
Platelet suspensions containing up to 160 pM AMT were tested against the
Salmonella/Mammalian Microsome Reverse Mutation Assay. Illuminated and non-
illuminated samples were tested. In samples without UV exposure or where the
residual AMT concentration was high, a significant number of mutations were
detected above background levels. The number of observed mutations decreased with
AMT concentration. Results of this AMES test show that AMT is mutagenic whereas
the photodecomposition products demonstrate no mutagenic activity. The
mutagenesis is- postulated to occur via AMT binding to nucleic acids.
Cytotoxicity tests are performed because they are rapid and straightforward with
endpoints which can be measured reproducibly and accurately.40 The disadvantage to
such tests, however, is that the assays are not mechanistically based and do not
usually provide information as to how and why the chemicals tested cause irritation.
In addition, it is difficult to use cytotoxicity test results as a predictor of in vivo
assessments.
84


AMT was shown to be cytotoxic to the cells examined in this assay. AMT has also
been shown to be genotoxic in the literature. Even though AMT induced a toxic
response in these cells, such data should be combined with a more strenuous
toxicological regiment to understand the mechanism of toxicity. Assays with a more
mechanistic design such as a liver assay using cytochrome P450 as a marker would
help to determine the potential metabolic interactions of AMT in a living system.
Since the concentration of AMT photoproducts in the illuminated were not a worst
case representation, a comparison of the toxicity of the sample pre and post UV
exposure cannot be made. Additional toxicological assessment of the AMT dimers is
especially recommended as they have the greatest peak areas of all the photoproducts
in the sterilized blood matrix.
85


7.
Toxicity Concerns
7.1 Introduction
Psoralen compounds have been examined for use in blood sterilization technologies
specifically for their mutagenic and toxic effects on DNA in the presence of UV light.
Finding a compound suitable for inhibition of virus and bacteria within blood is a
challenge, as the compounds that are effective have to meet a very rigorous set of
requirements. As outlined in Chapter 2, it must have much high DNA affinity, low
affinity for proteins and cell membranes and sufficient solubility for application in a
blood matrix.1 AMT, in particular, has meet some of the criteria as it has sufficient
water solubility and shown a high DNA affinity through demonstration of acceptable
levels of viral inactivation.
Much has been learned about how the type and placement substituents on the psoralen
molecule affect its ability to produce lethal affects with UV exposure. Research
studies have found that substituents on the psoralen molecule dramatically impact its
ability to form DNA adducts. Methylation of psoralen molecule, in general, increases
dark binding affinity, the quantum yield of photoaddition and the quantum yield of
photobreakdown of the compound. Dark binding affinity describes the ability for the
molecule to intercalate in the DNA and the strength of the non-covalent interaction.
86


Substituents on the C4 position are thought to cause steric interference with the
thymidine C5 methyl group as psoralens with C4 substituents form less than 2%
pyrone side adduct.
Substituents at the 3 and 4 positions appear to decrease the photoactivity of the
psoralen as steric effects exist with the methyl group on the 5 position of the
thymidine.19 There is evidence that molecules with only bifunctional unsaturated
groups at 3,4 and 4,5 positions produce lethal and mutagenic effects when added to
bacterial cultures supporting the crosslinking theory. Monoadducts and even
diadducts or interstrand crosslinking can occur when psoralen is relatively free from
steric hinderence at both the furan and pyrone sides of the molecule.
Psoralen affinity for DNA will be maintained for molecules that survive illumination
and are subsequently infused into a patient. From the experiments presented in this
thesis, anywhere from 15 to 60 pM of AMT is present in plasma post UV exposure
and will bind to patient DNA once infused. The impact of dark binding by psoralens
on DNA replication and protein expression in humans is not mentioned in the
literature. Presumably the effects of dark binding would be minor in humans for this
particular application given the milligrams each patient would receive from each
87


procedure. Due to the toxic nature of psoralens and AMT specifically these impacts
should be investigated prior to implementing such a technology in the field.
7.2 Literature Review
For much of the toxicological literature reviewed hereafter, studies involving 8-
methoxypsoralen will be presented. Historically 8-methoxypsoralen has been the
drug used to treat the skin conditions mentioned previously and is connected with the
side effects observed clinically. Toxicology studies using AMT specifically are not
present in the literature due to the lack of medicinal therapies involving its use. For
purposes of understanding psoralen toxicity in general, toxicological studies involving
all psoralens will be presented and statements relating to AMT will be made where
possible.
The toxic nature of psoralens in humans has been reported in the literature. As was
mentioned in the introduction, patients who had been treated for conditions such as
vitiligo and psoriasis with psoralens manifested complications such as erythema
(redness of the skin due to congestion of the capillaries), edema (swelling),
genotoxicity, risk of skin cancer and cataracts.30 In Chapter 6, data were presented
demonstrating both cytotoxic and mutagenic behavior of AMT. Psoralens such as 8-
methoxypsoralen have also been under investigations for liver toxicity.41
88


In many toxicology assessments of drugs, impairment or changes of hepatic activity
are examined as the liver performs the function of sequestering or breaking down
substances for elimination. If a new drug is toxic to the body it often manifests itself
in the liver first. The kidneys are also involved in detoxification and may be included
in a toxicology assessment as well. The liver and the kidneys constitute the primary
ways of eliminating substances from the body. If the substance is not sequestered by
either of these mechanisms, the compound may often be stored in the fat cells or
within cell membranes especially for those compounds that are hydrophobic. If the
substance is not eliminated, and exposure extended over time long term toxic effects
may be exhibited.
In the case of psoralens it does appear that breakdown and elimination does occur
hepatically. Both furan and pyrone ring opened urinary metabolites of 8-
methoxypsoralen were detected subsequent to cytochrome P450 (liver enzyme)
mediated oxidative attack on the molecule.41 It has been shown that 8-
methoxypsoralen is metabolically activated by both rat and human liver cytochrome
P450. The metabolites of this reaction in turn then inactivate P450. Through
investigations of coumarin and other psoralen derivatives it has been suggested that a
both furan and pyrone ring metabolites contribute to the inactivation of the enzyme.
Methylation of the furan ring appears to retard enzyme inhibition and presumably
89


Full Text

PAGE 1

ASSESSMENT OF THE PHOTOCHEMISTRY OF 4' -AMINO METHYL 4,5' ,8TRIMETHYLPSORALEN IN A BIOLOGICAL MATRIX by Renee Eileen Williams B.S. Brigham Young University, 1995 A thesis submitted to the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Master of Science Chemistry 2001

PAGE 2

This Thesis for the Master of Science degree by Renee Eileen has been approved by / i>, I Date

PAGE 3

Williams, Renee (M.S., Chemistry) Assessment of the Photochemsitry of 4' -Aminomethyl 4,5' ,8Trimethylpsoralen in a Biological Matrix Thesis directed by Assistant Professor Ellen J. Levy ABSTRACT The issue of blood safety has taken high priority in the wake of transfusion associated diseases such as HIV and hepatitis. Research efforts in recent years have focused on developing technologies to destroy the infectious capability of these pathogens in blood products prior to infusion into the patient. One approach currently under investigation employs a photosensitizer and UV light to sterilize pathogens that harbor in donated blood products. The photosensitizer has an affinity for DNA and will intercalate between base pairs. Upon exposure to UV light the photosensitizer absorbs the light energy. In this excited state the photosentizer can form covalent bonds or cause bond scission with surrounding material. This photochemistry causes DNA lesions rendering the pathogen incapable of replicating within a host. The compound 4' -aminomethyl-4,5' ,8-trimethylpsoralen (AMT) has been explored as a potential photosensitizer for this sterilization application. Studies demonstrate acceptable kill of challenge organisms but do not report the photochemistry for AMT molecules that do not bind to target DNA in a biological matrix. The purpose of this thesis is to report photoproducts within a biological matrix and examine additional AMT targets in human serum. Studies demonstrate that 4 major degradation products are detected post illumination: AMT furan:pyrone heterodimer, AMT pyrone:pyrone homodimer, 6-aminoacetyl-7-hydroxy-4,8,-dimethylcoumarin and a fourth compound which has many spectral similarities to 6-aminoacetyl-7hydroxy-4,8,-dimethylcoumarin but could not be positively identified. Estimates for the recovery of AMT can only account for 50% of the initial concentration in the form of photoproducts. Additional assessments show that approximately 50% of the AMT is bound to serum proteins. A large portion of the AMT may be bound to albumin but additional protein targets are present which should be investigated. lll

PAGE 4

Some toxicity information can be found for AMT but not for the photoproducts. Cursory studies indicate that the cytotoxicity and mutagenicity of AMT is greater than the photoproducts. Additional research must be conducted prior to clinical use of a technology including AMT to ensure that patients are not exposed to carcinogens and mutagens in an effort to decrease any risk of disease transmission. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. Signed IV

PAGE 5

ACKNOWLEDGEMENT This thesis was a collaborative effort on many fronts. I would like to express my appreciation to the Chemistry Department of the University of Colorado at Denver for their faith in me to accomplish an offsite project. In addition I want to acknowledge the support of GAMBRO BCT, INC. for providing the project idea and the means to accomplish the research. The assistance of Dr. Matthew Platz has been valuable in unveiling the AMT photochemistry. I would like to thank the following laboratory groups from the Chemistry Department at the University of Utah for accepting an orphan and making available their research facilities for this project: Dr. Cynthia J. Burrows group, Dr. Dale Poulter group, Dr. Matthew Sigman group and the Mass Spectrometry Laboratory

PAGE 6

CONTENTS Figures ..... ... ........... .... ........... .......... .... ....... .... ...... ........ ... .... ......... ... . ....... ....... viii Tables .......... ....................... . ................................................... ........................ ... ........ x Abbreviations .................................. .......................... ... ........................ ... ............... xi Chapter 1. Safety ofDonated Blood for Transfusion ......... ................ . . .................... ........ 1.1 Introduction .............. .................................................. ........ ............. ... ........... ... 1 2. Analysis of AMT Photoproducts by HPLC with Photodiode Array Detection ... ..... ....... .... ... .... .... ....... ..... . ... ... .............. .... ..... ..................... 11 2.1 Introduction ........... . ........................ ..... ................... ...... ................................... ll 2.2 Background on Photosensitizer Selection . .......... ........................................... 12 2.3 Experimental ... .... .... ..... .... .......... . . . .... .... ....... ..... ... ..... ....... . ... .................. l3 2.4 Results ............. .......... ................................................ ... .................................. l5 2.5 Discussion ......... . ... ................... ..... ..... ............ ............................................... 30 3. Analysis of AMT Photoproducts by HPLC Mass Spectrometry ........ . ..... ...... .42 3.1 Introduction .. .... .................... .... ........... .... .......................................... ............ 42 3.2 Experimental .......... .................. ........... ................. ... . . .................... ................ 43 3.3 Results ...... ................................. ..... .... ........................................................... 44 3.4 Discussion ... ...... .......................... ..... . ..................................... ...................... 51 4. Analysis of AMT Treated Plasma Proteins ........... ........ ................................. 58 VI

PAGE 7

4.1 Introduction .......... . ........................ .... ....................... .............. ........ ...... ........ 58 4 2 Experimental . . ..... ... ... ... ........... . ....... ........... ...... ... .... .... . ... ... . ..... . ... . ......... 60 4.3 Results ..... .......... .... ....... ...... ........... .... ......... ... ........ .... .............. .......... ... . .... 61 4.4 Discussion ............ ... . ................... ......................... ...... ... ................. ... . .......... 66 5. Analysis of Protein Function of AMT Treated Plasma ....... . .. .. ... ....... ... .......... 73 5.1 Introduction ..... .................. ............. .... ...................... ............................ ... ....... 73 5.2 Experimental ... ........ ........ .......... ... ............ ............... ................... . ............... 74 5.3 Results .... ... . ... . ..... ..... ......... ... ........ ..... ... ... . . ............. ... ... .. ... . .... ....... .......... 76 5 .4 Discussion ......... ...... ........ . .............. .......... . .......... .... ............ ....... .... . ........ 7 8 6. Cytotoxicity of AMT ........ ........ ......... ..... .... ............ ..... ............... ......... ... ........ 80 6 1 Introduction . . .... .... ............. ...... .... . ....... . ... .......... ............... . ..... ... ........... ... 80 6.2 Experimental ..... .............................. ............ ........... ..... .................... ....... ........ 82 6 3 Results and Discussion ........... ............................. .. ... ........ .......... ... .... .... .... ... 82 7 Toxicity Concems ... . ... . ..................... .......... ...... ......... ... .......... . . ............ .... 86 7.1 Introduction ........ .... . .................... ....... ........ ........... .... ............................. ........ 86 7.2 Literature Review ..... ......... ...... ..... . ... .............. .... . ... ................. .... ....... ......... 88 R e ferences .... ...................... . ......... ...... . ... ........ ................. ..................................... 95 Vll

PAGE 8

FIGURES Figure 2 1 AMT in Media without Illumination ........ ... .... ....................... ................... ... l7 2.2 AMT in Media with Illumination ..................................................................... 18 2.3 AMT Spectrum from Literature ........................................................................ 21 2.4 AMT Spectrum from HPLC Analysis .... ... ....... ............................................... 22 2.5 AMT in Plasma Extract without Illumination .................................................. 23 2.6 AMT in Plasma Extract with Illumination ............ ... . ................................ .... 24 2 7 Spectrum ofHMT Furan : pyrone Heterodimer ....... ......................................... 33 2.8 Spectrum of Peak at 4 Minutes ................................. .... . .................. ............. 34 2.9 Spectrum ofHMT Pyrone:pyrone Homodimer ........... ............. ....................... 35 2.10 Spectrum ofPeak at 7 Minutes .... ................................................................... 36 2.11 Spectrum of Peak at 3.7 Minutes ..... ... .................................... ............. ........... 39 2.12 Spectrum ofPeak at 14 Minutes ..... ........ ............ .. .... ................... ... .... ......... .40 2.13 Coumarin Spectrum from Literature ......................... .......... ........................... .41 3.1 Representative Chromatogram ofllluminated Sample ......................... .......... .45 3.2 Representative Chromatrogram ofllluminated Sample ............ ..... . ..... ......... .46 3.3 Chromatrogram of Concentrated Plasma Extract Illuminated Sample ............ .49 3.4 6-Aminoacetyl-7-Hydroxy-4,8-Dimethylcoumarin Formation ........................ 55 4.1 Chromatogram ofPlasma Blank .... ... ....................... ........................................ 62 Vlll

PAGE 9

4.2 Chromatogram of AMT in Plasma-Not Illuminated ........... ............... ......... .... 63 4.3 Chromatogram of AMT in Plasma-Illuminated .............. ..... ......... ................. 64 IX

PAGE 10

TABLES Table 2.1 AMT in Media ............................. .................. .............. ............. .................... 19 2.2 AMT in Plasma ........... .......................................................... .......................... 25 2.3 AMT in Plasma ...................... ............. ....................................... .................... 26 2.4 AMT in Plasma .... ............. ... ... .... ................ ... ............................... .................. 28 2.5 AMT in Plasma Concentrated Extract .............................................................. 29 3 .1 Masses by Retention Time .... .... ....... .................................. ............ ............... .4 7 3.2 Masses by Retention Time ......................... ... ........ .......................................... .48 3.3 Masses by Retention Time ... ............... ............................................................ 50 4.1 Scintillation Detector Results from Plasma Protein Analysis ............... .......... 65 4.2 Prevalent Human Plasma Proteins .................................................................... 68 5.1 Protein Function Data ................ ............. ................... ........................ ........... 77 6.1 Reactivity Grades for Direct Contact Test ................ ....................................... 81 X

PAGE 11

AIDS AMT AMU Avg BSA eV HN HMT HSA mAU PHS PUVA TA uv v ABBREVIATIONS Acquired Immunodeficiency Syndrome 4, -Aminomethyl 4,5 ',8-Trimethylpsoralen Atomic Mass Units Average Bovine Serum Albumin electron Volts Human Immunodeficiency Virus 4, -hydroxymethyl 4,5' ,8Trimethylpsoralen Human Serum Albumin Milliabsorbence Units Public Health Service Psoralen plus A ultraviolet light Transfusion Associated Ultra violet Volts Xl

PAGE 12

1. Safety of Donated Blood for Transfusion 1.1 Introduction Science continues to search for a suitable replacement for blood. The search is daunting, as any replacement must perform many complex functions. Blood brings fuel and oxygen to all cells, carries away waste, transports molecules and even functions as a communication pathway for the body. When disease and accidents dilute the efficiency of blood, it must be recharged quickly or life is threatened. Ensuring enough reserve blood to meet the needs of the community and a nation such as the U.S. is no small task. Cooperation on many fronts is required for success. Physicians and clinicians must use the blood products wisely so as not to reduce the supply unnecessarily. Recruitment for donors is a high priority for the blood banking clinics to sustain and provide a fresh supply. More than 12 million units ofblood are collected on an annual basis. 1 Blood supply management requires much thought as the blood products required by hospitals and clinics have a finite shelflife, some on the order of a few days. Along with ensuring a viable blood supply, blood banks must also ensure blood products are safe and free of disease. Donor contaminated blood can transfer life threatening conditions when infused into a patient. 1

PAGE 13

During the 1980's after acquired immunodeficiency syndrome (AIDS) was diagnosed, it was found that nearly half of the U.S. hemophiliac population contracted the disease without an identified means of exposure. 2 Up to this time the identified risk factors for the development of AIDS included sharing of drug needles or sexual intercourse with persons who harbored the AIDS virus. Not until 1983 when the human immunodeficiency virus (HN) was finally isolated was it clear that AIDS was transmitted by a blood-borne pathogen, as were hepatitis and syphilis. Prior to the isolation of the virus, patients were only known to contract AIDS after receiving blood donated by infected individuals? The occurrence ofHN transmission via blood products reached epidemic proportions prior to the implementation of steps by health authorities and industry to minimize the risk oftransmission in the blood supply.4 Most ofthe delay in reducing risk was due to the mysterious nature of the disease. Since HN attacks the T-cells ofthe immune system, patients with advanced stages of AIDS were diagnosed with diseases such as pneumonia, Kaposi's sarcoma or other opportunistic infections. AIDS patients were not responding to medication regimens commonly used to treat these diseases. Only until clinical evidence demonstrated a weakened immune system were physicians able to define clear symptoms of HN infection 2

PAGE 14

Patients contracting HN by receiving contaminated blood products became known as transfusion associated (T A) AIDS patients. It was observed that T A AIDS patients demonstrated an increased susceptibility to the disease. 2 AIDS patients who contracted the disease through high risk behaviors showed clinical manifestations on the average of seven years post-infection. The median time period in which TA AIDS patients demonstrated similar symptoms was much sooner, only 29 months post-infection. Upon further study of these cases, theories regarding why TA-AIDS patients were at an increased risk began to emerge. The prevalent theory for such a high vulnerability for transfusion recipients is three fold: their need for high volumes of blood, multiple exposures to contaminated products and the fact that they may be infected with a more virulent strain ofHN. Ironically, these patients developed AIDS from the very fluid they needed to preserve their own lives. Due to the fatal nature ofHN and the increased vulnerability of transfusion patients for AIDS onset, public concern for the risk of infection from the blood supply was very high. When enough information regarding the cause of AIDS was known, measures taken by the blood banking industry that dramatically reduced the risk of patient infection.4 5 AIDS was diagnosed in 1982; in 1983 the United States Public Health Service (PHS) issued a statement recommending that persons at risk for HN infection refrain from donating blood or plasma. An ELISA assay to detect HN in 3

PAGE 15

blood was made available for clinical use in 1985. The PHS announcement in combination with6 donor education and careful screening of the blood reduced the rate of infection dramatically. HN4 was not the first disease to be transmitted through the blood supply. Infectious viruses such as hepatitis and influenza were known to be present in donated blood prior to the 1980's It was not until the advent ofHN in the blood supply that efforts to reduce T A disease transmission were implemented. Following the HN ELISA assay, additional tests were implemented for routine screening of donated blood. Currently no less than seven different viral screens are performed on each donated unit of blood: syphilis, hepatitis Band C, HN types I and ll, and human T-cell lymphotropic virus I and ll. Also, in cases of patient susceptibility a screen for cytomegalovirus is performed. Patients who are immuno-compromised such as those undergoing extensive chemotherapy, renal transplantation, bone marrow transplantation, and liver transplantation are more susceptible to infection by cytomegalovirus and require additional testing on blood products they will receive.5 Testing of blood components combined with donor screening have helped to make the U.S. blood supply as safe as it has ever been. A 2000 report7 stated that the risk of transfusion-transmitted viral infections is extremely low. Effectiveness has improved 4

PAGE 16

to +99.9%. The risk ofHN infection was reported to be 1 in 677,000, 1 in 641,000 for human T -lymphotropic virus, 1 in 103,000 for hepatitis C, and 1 in 63 000 for hepatitis B virus. With all of these measures the risk of blood borne pathogen transfer via transfusion has not been completely eliminated. 5 Screening is only as effective as the honesty of the donors regarding high-risk behaviors There are vulnerabilities of the assays as well. A window of time exists between the time of infection to when the viruses become prevalent enough in the blood to be detected by the assay. For each virus assay that time frame varies. In the case ofHN the window oftime from infection to detection is 45 days. The estimated risk of passing" an infected unit of screened blood during that period is 1:225,000. Technological processes are under investigation to further decrease the risk ofT A disease transmission. The next step to improve blood safety is to commercialize a technology to inactivate viruses and bacteria in blood. 1 A direct method to destroy and/or compromise the agent of infection is a preferred method to ensure blood safety. This treatment may be performed on all blood products slated for use in clinics and hospitals and would prove to be a safeguard against "false negative" test results. This 5

PAGE 17

includes products, which are contaminated and have titers below the detection of the assay, as well as future infectious pathogens yet undetermined. One such technology, a solvent/detergent procedure8 10 has been implemented in various parts of the world For this treatment process, 1% oftri(n-butyl)phosphate and 1% triton X are added to the donated blood product and then maintained at 30C for 4 hours. There are several drawbacks to this technique 1. The procedure may only be used on plasma products (contain no cellular material only proteins) as the detergents disrupt cell membranes. 2 The solvent/detergent mixture must be removed using a laborious soybean oil extraction procedure and chromatographic purification techniques prior to clinical use. 3. This process is performed on large pooled plasma products (from 1000 to 20000 units per batch). Pooling plasma from multiple donors increases the risk of transfusion reactions. 4. The solvent/detergent is only effective against enveloped viruses. Non-enveloped viruses such as parvovirus B 19 will still be infectious after the procedure. 5. Up to 20% of the plasma volume is lost during the process. 6

PAGE 18

Another approach used instead of solvent/detergent for pathogen inactivation involves photosensitization. The photosensitizer would have to meet very strict criteria to be useful for this application. 1 The photosensitizer must be sufficiently water soluble for application in a blood matrix, have a high affinity for DNA and a very low affinity for cell membranes and proteins found in blood. In addition, the compounds must be non-toxic to the patient, and be effective under a gentle treatment process to maintain the viability of the blood This sensitizer would be added to the blood product and then the mixture would be exposed to specific wavelengths of light for a defined period of time. During illumination the photosensitizer absorbs the light. In this excited state the sensitizer transfers the excess energy which initiates chemical reactions causing catastrophic damage to surrounding material. These lesions render DNA incapable of replicating within the host and causing infection. One group of photosensitizers currently under investigation for the inactivation procedure is the psoralens.1114 Psoralens belong to a group of compounds called furocoumarins. Furocoumarins are heterocyclic molecules having a furan ring condensed to a coumarin (a bicyclic ring system involving benzyl and pyrone ring moieties). Psoralens have known photosensitive properties in the UV region and have 7

PAGE 19

been used to treat skin conditions for centuries In more recent decades psoralens have been used to treat psoriasis and vitiligo.151 8 A diagram of the psoralen structure is given below showing the three fused rings. Various substituents such as hyroxyl, methyl and amino groups are added to the rings to affect the binding affinity of psoralens as well as their solubility in water The ring structure is planar and has similar characteristics to the DNA bases 4' 5 4 5' Psoralen with atoms numbered The biological activity ofpsoralens comes from their affinity for pyrimidine bases14 specifically in the form of hydrogen bonding aromatic ring stacking and van der Waals interactions.19 Psoralens intercalate into the helix and primarily interact with thymidine in DNA and uridine in RNA; minor interactions with cytosine a lso occur .1415 Upon exposure to UV A light (340 400 run) cyclobutane formation involving either the C4' and C5' of the furan ring or C3 and C4 of the pyrone ring on the psoralen molecule with thymidine occurs?0 2 1 8

PAGE 20

With the knowledge of the photoactivity and mechanism of lethality in the treatment of skin disorders, investigations during the late 1970's began to examine the effectiveness ofpsoralens for the inactivation of viruses and bacteria for blood products.13 Many psoralens have been investigated for blood product sterilization applications: 8-methoxypsoralen, 4 ,5' ,8-trimethylpsoralen, and 4 hydroxymethyl 4,5 ',8 trimethylpsoralen.14 Another psoralen which has shown considerable promise for use in blood products 4' -aminomethyl-4,5'8-trimethylpsoralen (AMT) Methylation of AMT increases its photoreactivity with DNA while the primary amine on C4' increases its hydrophilic nature sufficient for application in a blood matrix. Psoralen treatment procedures have demonstrated acceptable kill rates for many pathogens 13 22 2 4 These studies assess the efficacy of psoralens in sterilizing human pathogens but do not focus on the impact residual psoralens in the blood products may have on the patients who receive them Nor has there been adequate attention given to the photochemistry of psorlens within a blood matrix. Identification of photoproducts in blood and a review of the toxicity of such compounds are absent from this body of literature. Concerns regarding the toxicity of the psoralen compounds and photoproducts must be addressed. Historically in the treatment of skin disorders, the adverse effects of 9

PAGE 21

psoralen treatment including erythema, edema, genotoxicity, increased risk of skin cancer and cataracts have been well documented 1 r'.17 When these compounds are considered for introduction directly into the blood stream, bypassing much of the body's own defense mechanisms, psoralen photoproducts must be characterized Once the degradation profile has been studied, rigorous toxicity screening of the parent compound and significant photoproducts must be investigated prior to market approval of such a treatment process. It must be clearly demonstrated that any carcinogenic or toxic effects ofthe sterilized product do not overshadow the potential benefit ofthe treatment process. The purpose of this thesis is to report the data and results of experiments conducted to determine the photochemistry of a psoralen compound (AMT) in a blood matrix. A review of the known toxicity of the parent molecule and significant photoproducts will follow the presentation of the AMT photochemsitry. With the presentation ofthese experimental results, it is hoped that the need for additional metabolic and toxic information about psoralens and decomposition products will be recognized and that the pertinent studies are performed prior to offering psoralen-treated blood products for routine use in hospitals and clinics. 10

PAGE 22

2. Analysis of AMT Photoproducts by HPLC with Photodiode Array Detection 2.1 Introduction Psoralen photoproducts have been studied in a variety of solvents such as ethanol, water, and methylene chloride. From these investigations it becomes evident that the photoproducts generated are very dependent upon the experimental conditions .2528 Upon illumination psoralen dimers have been identified at concentrations of 2.5 mM in methylene chloride while no dimer was detected in aqueous solution at 20 f.LM.28 Singlet oxygen has also been observed to direct the photodegradation of psoralens. When a singlet oxygen quencher was added to a matrix containing 8-methoxypsoralen the distribution of photoproducts was significantly shifted?6 additional studies have demonstrated the impact ofhydrogen ion concentration29 on the degradation products . With all of the psoralen photoproducts that have been identified in the literature, the photochemistry determined in a study will only be pertinent to the experimental design used to acquire the data. A review of psoralen photochemistry and photoproducts found in the literature will not be sufficient alone but must compliment actual results obtained from samples which have undergone the treatment process being examined. 11

PAGE 23

Psoralen photoproducts have been identified in the literature from an aqueous matrix at pH 7 but no studies have been published from data collected in a blood matrix.3 For purposes of understanding the photochemistry that occurs during the sterilization of blood products, information collected in a blood matrix is crucial as the complexity of reactions has the potential to increase dramatically 2.2 Background on Photosensitizer Selection For the sensitizer to be effective in a blood sterilization application it must meet a rigorous set of criteria. It must have much higher affinity for DNA than for proteins or cell membranes that will also be present in the blood.1 14 Many molecules which have a high DNA binding affinity are hydrophobic in nature. Proteins and cell membranes also have environments that attract hydrophobic molecules; finding a molecule that is exclusively selective for DNA poses a significant challenge Higher hydrophobicity in a molecule tends to decrease water solubility. Application for blood requires some water solubility so that the agent may be available to associate with the virus target to achieve inactivation.14 Among psoralens, AMT has shown promise. Methylation of AMT increases its photoreactivity with DNA while the primary amine on C4'increases psoralen solubility sufficient for application in a blood matrix. 12

PAGE 24

0 AMT 2.3 Experimental AMT (170 Sigma Chemical) was dissolved in solution of purified water and salts. The media content was sodium chloride (115 mM, Aldrich), sodium acetate (30 mM, Aldrich), sodium citrate (1 0 mM, Spectrum), monobasic sodium phosphate (11.4 mM, Fischer), dibasic sodium phosphate (14.6 mM, Spectrum). The sodium phosphate monobasic and dibasic ratio was buffered to pH 7 .4. Once fully dissolved, 30 mL of solution was placed into PVC bags (Charter Medical, PIN 13101 150 mL transfer bag). The bag was placed in a UVP illuminator (UVP, Inc.) equipped with twenty 365 nm light bulbs. Bulbs were arranged with 10 above the samples and 10 underneath. Samples were exposed between 1 minute 20 seconds and 2 minutes 15 seconds for an energy dose of2 to 3 J/cm2 Flux readings were taken on various days to determine day to day variation ofUV dose at 365 nm. An 13

PAGE 25

average flux reading was used to calculate the time required to obtain the desired UV dose. Post exposure non-illuminated and illuminated samples and blanks (media solution without AMT) were analyzed byHPLC (Hewlett Packard) A CIS column (Hewlett Packard, Eclipse XDB-C18 PIN 99096902) was used to achieve separation ofthe mixture. Column temperature was 30C. Initial mobile phase consisted of 40% acetonitrile and 60% 5 mM hexanesulfonic acid (Fischer) in water (pH adjusted to 3.2). Acetonitrile concentration increased to 100% in 15 minutes. Flow rate was 1.0 mL/min and the run time 15 was minutes Detector wavelength was set to 254 nm. UV profiles of all detected peaks were taken from 190 to 400 nm. The single donor human plasma used for these experiments was expressed from whole blood (BBMBC, Denver, CO). The data presented in this chapter was generated using a unit of 0+ blood containing a generic formulation of citrate, dextrose and phosphate for purposes of preserving shelf life. The media formulation outlined above was also used for plasma experiments except the concentration of AMT was increased. AMT concentration in media was 227 to 340 J.lM to achieve a concentration of 150 to 250 J.lM AMT in the final plasma 14

PAGE 26

sample. Media was mixed with plasma at a ratio of2 to 1. A total of 30 mL was placed into the charter medical transfer bags Illumination dose used for plasma was between 2 and 5 J/cm2 (1 minute 20 seconds to 3 minutes 25 seconds). To remove proteins prior to HPLC analysis, plasma samples were extracted through HLB cartridges (Waters, HLB #W AT094226). Columns were conditioned with 1 mL of methanol followed by 1 mL of water Then 2 mL of plasma sample were added followed by 1 mL of wash solution (5% methanol in water) and analytes were removed from column with 1 mL of 98% methanol 2% acetic acid. This solution was diluted 1:1 with water prior to HPLC analysis. For concentrated extracts, 2 and 3 times the plasma was loaded onto the column and eluted in the same amount of methanoVacetic acid solution The HPLC method was modified for the analysis of the plasma extracts. An isocratic method with a modified mobile phase was implemented. The mobile phase changed to 75% 10 mM ammonium acetate (Fischer) in water (pH adjusted to 3.5) and 25% acetonitrile (Aldrich). No other HPLC method parameters were changed 2.4 Results AMT was first illuminated in a media matrix to determine the photoproducts in the absence of plasma proteins. Figures 2.1 and 2 2 are representative chromatograms of 15

PAGE 27

AMT in media. Table 2.1 contains the areas and retention times for peaks observed in the chromatograms Only those peaks that appeared or significantly increased post UV exposure were listed in Table 2.1. 16

PAGE 28

Figure 2.1 AMT in Media without Illumination AMT in Media no Illumination 450 --------------........... .... --------! 400 350. 300 150 100 50 0 . . q q q q 0 0 0 N N N M M m m Time (Minutes) This figure shows the chromatogram for AMT in media with out illumination. AMT elutes at 3.5 minutes. 17

PAGE 29

Figure 2.2 AMT in Media with Illumination AMT in Media Illumination 70 ------------------------------------60 I 50 "" "" '" "' ., .. Q. c. !!! ::1 5 <:: ::<
PAGE 30

Table 2.1 AMT in Media Sample Retention Time (minutes) 3 3.5 4.6 6 Blank Not illuminated 1888.58 Illuminated 1 255.45 182 88 56.75 23.63 2 259.09 182.80 66.61 23 .81 3 279.30 192.02 93.24 29.51 4 269.71 158 65 76 .51 24.25 5 268 .11 132.93 85.63 29.22 6 259.94 108 .96 72.95 22.53 7 245.24 121.51 63.00 17.62 8 267.13 206.90 95.82 24.18 9 272.72 192.36 91.13 23.64 10 278.30 221.85 99 31 25.16 This table contains the areas and retention times of pertinent peaks found in AMT media chromatograms. Samples were prepared and analyzed on 6/26/00 The UVP illuminator was used and all samples received a dose of3 J/cm2 The sample labeled "Blank did not contain AMT and was not exposed to UV light. The sample labeled "Not illuminated" contained AMT but was not exposed to UV light. The samples labeled 1 through I 0 were replicates containing AMT and were exposed to UV light. The AMT peak appears at 3.5 minutes Three additional peaks appear post illumination. They elute at 3, 4 6 and 6 minutes. 19

PAGE 31

The AMT elutes at 3.5 minutes when using the hexanesulfonic acid method. This peak has the largest area counts in the control and the UV profile for peak eluting at 3.5 minutes matches literature spectrum for AMT. Figure 2.1 is a spectrum of AMT spectra obtained from the literature30 and Figure 2 2 is a spectrum of the peak appearing at 3.5 minutes obtained from HPLC analysis. Psoralens exhibit three characteristic regions of absorption maxima: the 225 region, the 250 run region and the 300 nm region 20

PAGE 32

Figure 2.3 AMT Spectrum from Literature ILl (.) z ex CD a: 0.2 g 0 1 CD ex 275 300 325 WAVELENGTH, nm 350 375 This spectrum only shows the absorbance profile from 275 to 400 nm for AMT. The absorbance maximum is shown at 300 nm. The absorbance profile slopes to zero before 400 run. 21

PAGE 33

Figure 2.4 AMT Spectrum from HPLC Analysis UV Spectrum of AMT 1000 ----------------------------------------. 900 BOD 700 :; 1 :! i! 0 400 300 200 100 190 210 230 250 270 290 310 330 350 370 390 Wavelength (nm) In contrast to the AMT spectrum from the literature, this figure also shows the AMT spectrum below 275 down to 200 run. Absorbance maxima are observed at 210,250, 300 run. The absorbance profile slopes to zero before 400 run. 22

PAGE 34

Figure 2.5 AMT in Plasma Extract without Illumination AMT in Plasma-Control 600 ... -------------500 400 200. 100 0 . . . q q q 0 0 0 N N N M M V V W W W m m m e Time (minutes) This figure shows the AMT peak eluting at 5.5 minutes. 23

PAGE 35

Figure 2.6 AMT in Plasma Extract with Illumination AMT In Plasmalllumlnated 600 ----. -----500 ; This figure shows the AMT peak eluting at 5.5 minutes and 4 degradation products eluting at 3.7, 4, 7 and 14 minutes. 24

PAGE 36

Table 2.2 AMT in Plasma Sample Retention Time (minutes) 3.7 4 5.5 7 14 Blank Plasma Blank Not illuminated* 4198 8 Illuminated 1 59.2 195 8 1039.3 111.4 46.8 2 62.1 202 9 973.2 119.2 53.9 3 52 5 223.2 892.9 118 2 65.4 4 66.2 213.2 738.0 121.1 67 6 1* 26 2 95.4 1174.9 121.5 61.4 3* 30. 5 98.5 797.8 140.5 85.8 This table contains the areas and retention times of important peaks found in the chromatograms from the AMT plasma extract samples. Samples were prepared on 8/17/00 and analyzed on 10/02/00. The UVP illuminator was used and samples 1-4 received an illumination dose of 2 J/cm2 The sample labeled "Blank" cons i sted of media that did not contain AMT and was not exposed to UV light. The sample labeled "Plasma blank" consisted of a plasma extract that did not contain AMT and was not exposed to UV light. The sample label e d "Not illuminated" consisted of plasma extract containing AMT that was not was not exposed to UV light. The samples labeled 1 through 4 were replicate plasma extracts containing AMT. The AMT peak appears at 5 5 minutes. Four additional peaks appear post-illumination They elute at 3.7, 4, 7 and 14 minutes. AMT peak in the not-illuminated sample did was not as lar ge as expected. To understand if the extraction process was performed improperly, the not illuminated sample and AMT 1 and 3 were extracted and analyzed a second time on 2/20/00. 25

PAGE 37

Table 2.3 AMT in Plasma Sample Retention Time (minutes) 3.7 4 5.5 14 Blank Not illuminated 2908.4 Illuminated 1 34.95 43.4 136.9 139.6 2 37.7 41.2 87.7 160.9 3 28.95 37 25 120 6 133.25 This table contains the areas and retention times of important peaks found chromatograms from the AMT plasma extract samples. Samples were prepared and analyzed on 2/20/01. The Research illuminator was used and all samples received an illumination dose of2 J/cm2 The sample labeled "Blank" consisted of media that did not contain AMT and was not exposed to UV light. The sample labeled "Not illuminated" consisted of plasma and AMT but was not exposed to UV light. The samples labeled 1 through 3 were replicate plasma extracts containing AMT. The AMT peak appears at 5 5 minutes. Three additional peaks appear post-illumination. They elute at 3.7, 4, and 14 minutes. Data shown for each sample in this table is an average of two data points. The peak at 7 minutes may not have appeared in this data because a different illuminator was used. Smaller peak areas for degradation products were also observed in this data set than for samples illuminated in the UVP. 26

PAGE 38

AMT under the aqueous ammonium nitrate/acetonitrile method elutes at 5.5 minutes. The UV profile for this peak matches that of the peak eluting at 3.5 minutes in the hexane sulfonic acid method. To determine the repeatability of the degradation profile in plasma using the same illumination device, a final experiment was conducted. For this experiment samples of AMT in media and plasma were prepared as well as blanks of media and plasma. The injection volume for this sample analysis was increased from 20 uL to 60 uL to improve instrument sensitivity The data is listed in Table 2.4 to follow. Plasma samples from Table 2.4 were extracted a second time and concentrated by a factor of2 or 3. This was done to minimize solvent interference in the UV profiles for each AMT photoproducts. This would ensure a more accurate assessment of structural characteristics depicted in the UV profile. Data from the analysis of concentrated extracts are listed in Table 2 5. 27

PAGE 39

Table 2.4 AMT in Plasma Sample Retention Time (minutes) 3.7 4.0 5.4 7 14 Blank AMT media not illuminated 9 9 76.8 3799.4 AMT media illuminated 22.7 2296.9 486.2 485.17 82.3 Plasma blank AMT plasma not illuminated 5390.7 AMT plasma illuminated 209.7 831.9 1199 8 206.5 146.0 AMT plasma illuminated 90. 1 662.5 1152.1 215.8 172.4 This table contains the areas and retention times of important peaks found in the chromatograms for AMT plasma extracts and media samples. Samples were prepared 2/26/01 and analyzed on 3/08/01. The UVP illuminator was used to deliver a 2 J/cm2.dose to all illuminated samples. The sample labeled "Blank" consisted of media that did not contain AMT and was not exposed to UV light. "AMT media not illuminated" and "AMT media illuminated" consisted of samples containing AMT in media with and without UV exposure, respectively. The sample labeled "AMT plasma not illuminated" consisted of plasma and AMT but was not exposed to UV light. The samples labeled "AMT plasma illuminated" were replicate plasma extracts containing AMT. The AMT peak appears at 5.5 minutes. Four additional peaks appear post-illumination. They elute at 3.7, 4, 7, and 14 minutes. 28

PAGE 40

Table 2.5 AMT in Plasma Concentrated Plasma Extracts Sample Retention Time (minutes) 3.7 4 5.5 7 14 Plasma blank* 60.8 36.0 AMT plasma control* 12050 AMT plasma ilium+ 2118.8 227.8 3498.0 866.6 394.6 AMT plasma illum + 2252.1 265.4 3692.4 958.3 450.1 This table contains the areas and retention times of important peaks found in the chromatograms from the concentrated plasma extracts. Samples were prepared on 2/20/01 and analyzed on 3/08/01. The UVP illuminator was used to deliver a dose of 2 J/cm2 to all illuminated samples. The sample labeled "Plasma blank" consisted of a plasma extract that did not contain AMT and was not exposed to UV light. This extract had a concentration factor of2. The sample labeled "AMT plasma not illuminated" consisted of plasma and AMT but was not exposed to UV light. This extract also had a concentration factor of 2. The samples labeled "AMT plasma illuminated" were replicate plasma extracts containing AMT post-UV exposure. These samples had a concentration factor of 3. The AMT peak appears at 5.5 minutes. Four additional peaks appear post-illumination They elute at 3.7, 4, 7 and 14 minutes. 29

PAGE 41

2.5 Discussion Data collected in these experiments consistently show 4 degradation products for all experiments where experimental conditions including the analytical separation method are kept constant. From the UV profiles the peaks presented in Table 2.1 can be matched with peaks appearing in subsequent tables The UV profiles for peaks in Table 2.1 appearing at 3, 4.6 and 6 minutes match closely with peak profiles from Tables 2.2-2.5 appearing at 4, 7 and 14 minutes respectively. The repeatability of the UV profile further suggests the consistency of AMT photochemistry in this experiment. An evaluation of peak areas pre and post illumination from Tables 2.1-2.5 reveal that between 75 to 90% of the AMT is destroyed during photolysis. Comparing peak areas of AMT in media and plasma pre and post-illumination shows that more AMT survives in plasma than in media for the same UV dose (see Table 2.4). This suggests that photodegradation of AMT is faster in media than in plasma Plasma proteins also absorb in the UV and would have the effect of retarding AMT decomposition during illumination. 30

PAGE 42

Peak areas for degradation products are much greater in media samples than in plasma extracts, when UV dose and illuminator are constant. By summing all the identified photoproduct peak areas of AMT, 90%+ ofthe initial AMT peak area is found in the post illumination chromatograms from media data in Table 2 4. Only 30 to 60% of the initial AMT peak area is present post-illumination in the plasma extracts. Even though the same dose was given to plasma and the media samples additional photochemistry is occurring in plasma that has not been observed by this analytical technique Interactions between AMT and plasma proteins probably occur and may account for the reduction in photoproduct generation AMT bound to protein would not have been captured in the plasma extract and been absent from the HPLC analysis. A more direct assessment of plasma proteins will be necessary to determine if AMT protein interactions occur. Even though the concentration of AMT photoproducts in plasma is much less than in media, the distribution of photoproducts form e d during illumination remains consistent. The peak area at 4 minutes is the largest of all AMT degradation products accounting for 60% and 13% ofthe initial AMT peak area in media and plasma, respectively. The next largest peak area for a degradation product is the peak at 7 minutes in both media and plasma. This peak accounts for 13% of the initial AMT 31

PAGE 43

peak area in media and 4% in plasma. The peaks at 3.7 and 14 minutes account for less than 3% of the initial AMT peak area in both media and plasma. Clues to the structure of the AMT photoproducts can be obtained from studying the UV profiles of each peak. Interestingly, two of the UV profiles of the AMT photoproduct peaks are similar to profiles obtained for photoproducts of other psoralen compounds. A psoralen similar to AMT called 4' -hydroxymethyl-4,5 ,8-trimethylpsoralen (HMT) was exposed to UV light at concentrations of 10 to 100 l-!M for a dose of 18 to 36 J/cm2 (240-300 nm) .33 The difference between AMT and HMT is that where the former has an NH2 group the latter has an hydroxyl group Results of the HMT study demonstrated the formation of two psoralen dimers : a furan:pyrone heterodimer and a pyrone:pyrone homodimer. The profile ofthe furan: pyrone heterodimer is similar to the peak at 4 minutes, while the pyrone:pyrone homodimer spectrum is very similar that of the peak at 7 minutes. None of the spectra presented in this paper were similar to the spectra obtained for the 3 7 and 14 minute peaks. 32

PAGE 44

Figure 2.7 Spectrum ofHMT Furan:pyrone Heterodimer 0.60 0.40 0 .20 200 300 This spectrum shows absorbance maxima at 210, 230, 260, and 340 run for the HMT furan : pyrone heterodimer. The Y -axis is presented in units of absorbance. 33

PAGE 45

Figure 2.8 Spectrum of Peak at 4 Minutes 250 ..... -... 200 5' 150 < .. !l c 100 50 UV Spectrum of 4 Min Peak -----------------------------.... ---------""' \ \ '-----190 220 250 260 310 340 370 Wavelength (nm) This figure shows similar absorbance maxima to Figure 2.8. They are observed at 210, 225, 260, and 330 nm. 34 400

PAGE 46

Figure 2.9 Spectrum of HMT Pyrone: pyrone Homodimer 0.10 200 400 This figure shows absorbance maxima at 225 with a plateau around 2600 run for the HMT pyrone:pyrone homodimer. Absorbance profile slopes toward 0 between 300 and 400 run. The Y -axis is presented in units of absorbance. 35

PAGE 47

Figure 2.10 Spectrum of Peak at 7 Minutes UV Spe c trum of7 Min Peak 3 5 ,-... -------:----------30 1 0 -------t 2 5 :> g 20 c 15 .. 10 I 190 220 250 280 310 340 370 Wavelength (nm) This figure shows the absorbance maxima at 220 with a sloping plateau at 250. Absorbance profile slopes toward 0 about 300 run. 36 400

PAGE 48

It is interesting to note how similar the spectra for the peaks appearing at 3. 7 and 14 minutes are to each other when they appear in the chromatogram over ten minutes apart. Spectra from these two peaks may be superimposed almost exactly suggesting the two compounds have a very similar structure. As the retention times differ so greatly, the polarity of these compounds must be very different. The compound appearing at 14 minutes may be a dimer of the compound eluting at 3.7 minutes. If two molecules were linked at positions that did not alter the chromophore, such a spectral similarity may be observed for the two compounds. UV profiles with absorbance above 300 nn indicate extensive conjugation. A relative absorbance maximum around 300 run is characteristic of psoralens whereas for coumarin that maximum is shifted up to 320. Coumarins also exhibit a maximum around 270 run; the psoralen maximum is usually much stronger and appears around 250 run. The structure of coumarin is given below and an UV spectrum of the compound is located in Figure 2.13. 32 5 4 6 7 Coumarin with atoms numbered 37

PAGE 49

Spectra for the peaks eluting at 3. 7 and 14 minutes have closer similarities to the coumarin spectrum than the AMT spectrum (Figure 2.4). Photochemistry surrounding formation of coumarin products from AMT would involve rupture of the furan ring. Data presented in Chapter 2 has given many clues as to the number of products obtained during illumination as well as some of the structural characteristics of those photoproducts. Key to the identification of some of these photoproducts will be the determination of the masses associated with the compounds. Mass spectrometer data will be presell:ted in the next chapter. 38

PAGE 50

Figure 2.11 Spectrum of Peak at 3.7 Minutes UV Spectrum of 3 7 Min Peak 50 -----------... ... . --------------45 40 35 5' 30 c( .. .. 25 .. -e D .. .a c( 20 15 10 5 190 220 250 280 310 340 Wavelength (nm) This figure shows three absorbance maxima: 21 0, 264, and 345 run. 39 370 400

PAGE 51

Figure 2.12 Spectrum of Peak at 14 Minutes 25 _ ___________________________________ ____ _____________ _ _ _ 20 S' 15 l " c -e 0 0 .a < 10 5 UV Spectrum of 14 Min Peak 190 220 250 280 310 340 370 400 Wavelength (nm) This figure shows three absorbance maxima: 212, 260, and 350 nm. 40

PAGE 52

Figure 2.13 Coumarin Spectrum from Literature 0.!!!1< : ";: 0.4 0..:3" 0.2. 0.1 The figure is presented in units of absorbance on the Y -axis and wavelength (run) on the X-axis. In this spectrum absorbance maxima are observed at 270 and 320 run. 41

PAGE 53

3. Analysis of AMT Pbotoproducts by HPLC Mass Spectrometry 3.1 Introduction The number of AMT degradation products was identified in Chapter 2 but only very qualitative statements may be made about the identity of the photoproducts without additional analysis. In mass spectrometry, the mass of an unknown compound may be determined without comparing analytical results with known compounds for positive identification. The mass spectrometer is calibrated for mass independently of the specific sample being analyzed. Associating a mass to each peak identified in the HPLC photodiode array analysis may be a powerful piece of information to determine the structure of the AMT photoproducts generated during this sterilization procedure Structural characteristics gained from the UV profile and mass for each peak can also be combined with additional information to ensure proper peak identification. The photochemistry expected to occur in solution can be extremely powerful in assessing the likelihood of formation for any proposed structure. A review 1 of psoralen reactions most likely to occur in an aqueous environment as well as the bonds parent molecule which would most likely to break under UV exposure has been published in the literature. From such information, a structure as well as a mechanism can be 42

PAGE 54

presented to describe the AMT photochemistry occurring in the blood matrix during illumination. In this chapter, the masses associated with the 4 degradation products identified in Chapter 2 will be presented Structures of the degradation products will be presented as well as reaction mechanisms to support formation of these structures in solution. 3.2 Experimental The HPLC used in this study was a 2690 separations module (Waters) equipped with a 2487 dual absorbance detector (Waters). Following the UV detector a Quattro IT triple Quadrupole Mass Spectrometer (Micromass) was connected. The MassLynx 3 2 (Micromass) operating system was used to control and communicate between the components. Settings for the mass spectrometer were cone voltage 5V, capillary voltage 2 8 kV, and extractor voltage 3V. In addition the desolvation temperature was 250C, the source block temperature was 80C, and the ionization energy was set to 0.7 keV. This is a z-spray spectrometer and the analysis was run in positive ion mode. In positive ion mode the molecules are bombarded with the ionized reagent gas CH5 +. Molecules are protonated under this technique. Usually 1 proton sticks to the molecule to form a charged species {M+ 1) + that can be selected in the quadrupole magnet for detection. 43

PAGE 55

Sample extraction and the HPLC method used for these analyses are outlined in detail in chapter 2. The same sample extracts which were run for HPLC photodiode array analyses were also run for mass spectrometry analysis. 3.3 Results Retention times for the same peaks were shorter by 0.5 minutes in the HPLC mass spectrometry analysis than for the HPLC photodiode array analysis. For reasons of clarity, the retention times for the peaks presented in Chapter 2 will be used for the mass data instead ofthe retention times observed in this analysis. Figures 3.1 and 3.2 contain representative chromatograms for this analysis. 44

PAGE 56

Figure 3.1 Representative Chromatogram of Illuminated Sample 20000753 Scan ES+ 51!: 3.74e 4 .00 5.40 i .. o l __ _ 20000753 Scan ES+ 100., 5.40 258 1 n 9.93e6 i ; \ 6.54 o.L---==,---_ _;. ... 20000753 .... . . . 100l %4 0 . 20000753. 100-, l 2.76 3.85 5.40 I\ \ f \ \ _j' I Scan ES+ 248 2 .17e6 % : 0 18 3 .54. ; 618 881 9 79 1274 1.78 2 .61 ' 4 42 5.76 . 9.27 10.98 12.27 13: 9214.70 0 .. .. /-. --. .\ .......... .: -. .... ...... ... ...._ ___ ... -"-!." .... ----.. -. .._,_ ._ .......... l\. -...... .. .. ---...... .... ...... 20000753 Scan ES+ 1001 4.00 % 1 ; 01 ... 20000753. 100J %1 o.+ .... 20000753. 2 36 3.07 i\ \ ,_ 1 1 h ,, } ,.J \ .: . I . 6.18 ...... "-. 4 .90 n i\ j \ 5.67 1 \, f\ I I J 13.07 254nm An1 10.00e5 This figure shows the mass traces for 515, 258, 241, 248 and 498 m/z. The last signal shown in this chromatogram is the UV trace collected at 254 run. 45

PAGE 57

Figure 3.2 Representative Chromatogram of Illuminated Sample 20000753 100-\ 13.56 Scan ES+ 290 2.31e6 l o/o 1 .06 3 .95 5 .04 6 .18 7.317.67 9.07 9 5 8 11.70 12.99 ; \ \ / \13.7 7 \I r \ \ 14.18 \ .-':" ..... ...... ....... ..... ..... .. \,..V""' ,_ '-' /, ....... --1"-.'\....J 0 20000753 v". 14.80 Scan ES+ 601 2.75e 5 100l! 2-r:.62 13.51 ill 5.14 12.53 1 o 80 1 21 l 7 .36 10.36. 1072 r1 14.60 o/o .. 2.51 1 ; : 3.33 3 .9o fir a.a1 -1 1 11 70/ n 1 1 I '' l I 8 .148.45: I l \11.13 i 14. 8 6 . .. i.,. J ,, l 'lu , ,1!13.9a, "' 9 .27.. ' ; '. , \'IiI [ , ,, '.'\ Vif 1 V : t l 1v t v v N1i.rt ',. .. ... ,,.: !V: i \/\ .. \/.: .. ; ; 254nm An1 10.00e5 100This figure shows additional mass traces for the sample in Figure 3.1 at 290 and 601 m!z. The last signal shown in this chromatogram is the UV trace collected at 254 nm. 46

PAGE 58

T bl 31M a e asses b R t f T" ly_ e en 1on 1me Sample 3.7 4 5.5 7 14 Not Illuminated 241 Illuminated AMT1 248 498, 515 241, 258, 515 498,515 290 AMT2 248 498, 515 241, 258, 515 498,515 290 AMT3 248 498, 515 241, 258, 515 498,515 290 AMT4 248 498, 515 241, 258, 515 498,515 290 This table shows the mass signals associated with the peaks observed in the UV chromatogram. All masses listed have units of rnlz and are an M+ 1 positive ion. The data in this table were obtained from samples prepared on 8/17/00. 47

PAGE 59

T bl 32M a e asses b R t f T. )y e en Ion 1me Sample 4 5.5 7 14 AMT media not 241, 258, 299, 515 illuminated AMT media illuminated 498, 515 241' 258, 515 498,515 290 AMTplasma 498 241, 258, 515 498,515 290 illuminated AMTplasma 498 241, 258, 515 498,515 290 illuminated AMT plasma not 241, 258, 299, 515 illuminated This table shows the mass signals associated with the peaks observed in the UV chromatogram. All masses listed have units of m/z and are an M+ 1 positive ion. The data in this table were obtained from samples prepared on 02/26/01. 48

PAGE 60

Figure 3.3 Chromatogram of Concentrated Plasma Extract Illuminated Sample 7 (3X) Renee Williams (02126/01) 010254 100 13 62 % 0 .. 010254 1\ % f \ Scan ES+ 601 Scan ES+ 29( 1.15e7 100j 13.67 3.33 3.75 j \ '010254 % 3.133. 76 II 254nrn An1 5.61e5 1001 sr 2 37 1 j \ 6.03 6 47 13.43 .. This figure shows the appearance of a mass signal at 601 m/z for the peak eluting at 14 minutes in the concentrated AMT plasma extract. The mass signal at 290 m/z also increases for the same peak in the concentrated extract. The last signal shown in this chromatogram is the lN trace collected at 254 nm. 49

PAGE 61

T bl 33M a e b R etentmn asses 'Y T" fi c 1me or oncentrate dS amples Sample 3.7 4 5.5 7 14 AMT plasma 248,231 498,578 241, 258, 515 515 290,331,601 illuminated AMT plasma 248,231 498,578 241, 258, 515 515 290,331,601 iII uminated AMT plasma not 241, 258, 299, illuminated 515 This table shows the mass signals associated with the peaks observed in the UV chromatogram for the concentrated plasma extracts. Additional mass signals of interest appear for the peak eluting at 14 minutes. All masses listed have units of m/z and are an M+ 1 positive ion. The data in this table were obtained from samples prepared on. 02/26/01. 50

PAGE 62

3.4 Discussion In the control three mass peaks are observed for the AMT peak: 241 (M + 1), 258 (M + 1) and 515 (M+ 1) rnlz. A fourth mass of 299 rnlz appears for the control samples in tables 2 and 3 The molecular weight of AMT is 257. The mass at 258 m/z is the molecular ion peak of AMT. The strong signal at 241 is 17 mass units less than 258 suggesting a loss of an OH group from the molecule. This indicates that the most stable ion form of AMT in the mass spectrometer has one less hydrogen and oxygen on the molecule. The mass at 515 is double the mass at 257 (molecular weight of AMT) and indicates dimer formation within the mass spectrometer The 299 rn!z only appears in the AMT samples when the concentration ofthe molecule is high. This suggests that the molecule is combining with another molecule in the mass spectrometer and is not a true representation ofthe mass of the molecule. Acetonitrile constitutes 25% of the mobile phase and has a molecular weight of 41. Adding 41 mass units to 258 equals 299 explaining the appearance of the 299 mlz peak. Reviewing the illuminated samples, a mass of 515 consistently corresponds to peak eluting at 7 minutes. In chapter 2, it was shown that the UV profile of the peak eluting at 7 minutes matched that of a psoralen homodimer presented in the 51

PAGE 63

literature. 33 As stated above a mass of 515 (M+ 1 ), is the mass of a molecule twice the weight of AMT. This evidence confirms that the peak eluting at 7 minutes is an AMT dimer. By UV profile and the mass data the peak eluting at 7 minutes has been identified as the AMT pyrone : pyrone homodimer. Also presented in chapter 2 was the similarity between the UV profile of the peak eluting at 4 minutes and the furan:pyrone heterodimer. The main mass associated with this peak is 498 m/z. Interestingly, this mass is 17 units less than 515. A loss of 17 mass units was also seen for the major peak of AMT. The pyrone:pyrone homodimer does not have this same phenomenon. The difference between the pyrone:pyrone homodimer and the other two molecules is that the both pyrone ring ) moities are protected through dimer formation. In the furan:pryone dimer only one pyrone moiety is protected. AMT also has an unprotected pyrone moiety. The reduction of the molecular mass in the mass spectrometer must be associated with the rupture of the pyrone ring and subsequent loss ofOH. Using this reasoning the peak eluting at 4 minutes has been identified as the AMT furan : pyrone heterodimer by both UV spectral match and mass. A paper reviewing photooxidation studies on 8-methoxypsoralen characterized a dimer link at the C3 position of the molecule.25 The C3 position of AMT is not 52

PAGE 64

substituted and is a possible site for a dimer linkage. Psoralens are also known to dimerize at the 4'5' double bond of the furan ring. For both dimers, [2+2) cycle addition is the most probable reaction mechanism for cyclobutane ring formation between the AMT molecules. Figures of the dimers are presented below. These structures were adapted to AMT from dimers presented for HMT. 33 0 0 CH:J AMT Pyrone:Pyrone Homodimer 53

PAGE 65

0 0 0 CHa 0 CHa AMT Furan:Pyrone Heterodimer A 248 rn/z is associated with the peak eluting at 3.7 minutes. A structure has been presented to have a mass of247 below. This structure is formed by the scission of the 4' ,5' double bond in the furan ring. Rupture of this ring at the 4' ,5' position has been reported for other psoralens.25 It is thought that this reaction occurs via a dioxetane formation involving singlet oxygen attack to the double bond. Substituting electron rich groups such as methyls on to the double bond enhances this reaction. Both the 4' and 5' positions are substituted with such groups. This molecule would have increased hydrophilic character due to the hydroxyl group and would be expected to elute before AMT. This mass peak appears before the AMT masses and is associated with the peak at 4 minutes in the HPLC chromatogram. The proposed reaction mechanism involving singlet oxygen of AMT is given below. 54

PAGE 66

Figure 3.4 6-Aminoacetyi-7-Hydroxy-4,8-Dimethylcoumarin Formation 0 0 0 Singlet oxygen is presented in this figure attacking the psoralen molecule causing the furan ring to rupture in formation of a coumarin compound. 55

PAGE 67

Mass 248 does not appear in the data presented in table 3 as the 3.7 minute peak is too weak. When samples 6 and 7 are concentrated a mass of 248 as well as 231 appears for the peak. The pyrone ring is unprotected in this molecule as well and may rupture upon entry into the mass spectrometer. The repeatable retention time, mass data and UV profile for this peak observed in chapter 2 demonstrates that the same compound has been generated in both data sets. The peak eluting at 3.7 minutes has been identified as 6-aminoacetyl-7-hydroxy-4,8-dimethylcoumarin. 0 6-Aminoacetyl-7-Hydroxy-4,8-Dimethylcoumarin A mass of290 rnlz is associated with the 14 minute peak. A weaker mass of331 rnlz appears in the data presented in Table 3.5 for the concentrated extracts. A mass of 331 is 41 mass units higher than 290 and is consistent with acetonitrile combining with the 290 molecule as was seen for the AMT molecule. A mass of289 does not correspond with any known psoralen degradation product. The structure of this compound probably has similarities to 6-aminoacetyl-7-hydroxy-4,8dimethylcoumarin as the spectral profile for the two peaks is so similar. The 56

PAGE 68

concentrated samples also show a small peak at 601 m/z for the peak eluting at 14 minutes. This mass is probably closer to the molecular weight ofthe molecule but a structure with the mass of600 atomic weight units (AMU) cannot be determined from 6-aminoacetyl-7 -hydroxy-4,8,-dimethylcoumarin. Additional characterization of this compound will be required for positive identification. With 3 ofthe 4 major degradation products identified, experiments will be reviewed in Chapter 4 that begin to assess the AMT /protein interactions. From the data presented in Chapter 2 a significant portion of the AMT, on the order of 50%, can be expected to interact with the plasma proteins. 57

PAGE 69

4. Analysis of AMT Treated Plasma Proteins 4.1 Introduction With the major AMT photoproducts identified in Chapters 2 and 3, the focus of the research will now turn to the protein component of the samples. A cursory attempt at a mass balance of AMT in plasma post illumination was presented in Chapter 2. From that analysis it was estimated that approximately 50% of the AMT was missing post UV exposure. Since the samples were not inoculated with pathogens and the plasma had been tested for known infectious agents, AMT must have additional targets within plasma besides the DNA. The major component in plasma besides water is proteins. Research studies have been conducted to determine if psoralens will also bind to proteins such as bovine serum albumin (BSA) and human serum albumin (HSA) in addition to DNA.33 36 Strong associations between proteins and psoralens have been documented in the literature which occur without light energy for excitation. 33 3 4 In photochemistry, this non covalent association is often called dark binding. Dark binding of psoralens and proteins has been strong enough to interfere with the determination of covalent binding of psoralens to proteins. High affmity dark binding has been observed between lipoproteins and furocoumarins and has been measured by several analytical 58

PAGE 70

techniques including gel exclusion chromatography, difference spectrometry, equilibrium dialysis of radio-labeled furocoumarins, and fluorescence quenching.33 Results from the fluorescence quenching experiments suggest that the tryptophan residue ofhuman serum albumin is involved in binding with psoralens indicating 1 psoralen binding site per molecule of human serum albumin. Under the influence of light, enough energy is present for psoralens to form covalent bonds with proteins. Photobinding of psoralens is postulated to occur by two separate pathways. Either the excited furocoumarin binds directly to the protein or an excited furocoumarin forms, probably via an excited singlet oxygen mechanism, photoproducts that covalently bind to the protein. This theory was presented after reviewing data where 8-methoxypsoralen in combination with BSA was illuminated in the presence and absence of oxygen with varying molar ratios and illumination times.36 Binding was observed both where 8-Methoxypsoralen was pre-irradiated or irradiated in the presence ofBSA. In an effort to determine the extent of AMT binding to plasma proteins during the sterilization process, an analysis of the plasma proteins in conjunction with radiolabeled AMT was designed and executed. 59

PAGE 71

0 Structure of Tritiated AMT 4.2 Experimental HPLC used was a 2690 separations module (Waters) equipped with a 441 UV detector (Waters). In series behind the UV detector was a flow scintillation analyzer (Packard, 500TR series Ser #423464 ). Flow 1 (Packard, Ver 3 .6.1) software was used for the operation ofthe equipment and data manipulation The chromatography achieved using a size exclusion column (Zorbax GF250, Hewlett Packard, PIN 884973-901). This stationary phase is silica based and has diollinkages The silanol groups are clad with zirconium to reduce any analyte charge interactions with the stationary phase. The significant interaction between the analytes and the stationary phase is through entry into the pores. Therefore retention time is determined by size. Mobile phase consisted of 130 mM sodium chloride, 20 mM chloride, and 50 mM sodium phosphate, at pH 7. The flow rate was 1 mL/min and the column temperature was 30C. The detector wavelength was set to 210 nm. Analysis time was 20 minutes and injection volume was 100 uL. 60

PAGE 72

Samples were prepared as outlined in Chapter 2 with the following exceptions. Two one mL aliquots of tritiated AMT (Cerrus, 0.4 C/mmol) were added to 10 mg of AMT (Sigma, #A4330). At the time of the last analysis in 1995 the activity of the AMT was 0.6 mC/mL. Therefore approximately 1.2 mC of AMT was added to the media mixture. The concentration of AMT in the final plasma samples was 250 4.3 Results Representative chromatograms of this analysis are presented in Figures 4.1 to 4.3. Scintillation counts from the sample chromatograms are listed in Table 4.1. 61

PAGE 73

Figure 4.1 Chromatogram of Plasma Blank 13 (min) 0 10 15 2C This figure contains the chromatogram from the analysis of plasma without AMT present and without lN exposure. The radioactivity counts with respect to time are shown in the top graphic with the lN chromatogram underneath. The lN chromatogram was obtained at 254 nm. The Y scale for the radioactivity counts ranges from 0 to 13 and shows the noise level of the detector. A large protein peak is observed in the lN chromatogram beginning at 1.5 minutes. 62

PAGE 74

Figure 4.2 Chromatogram of AMT in Plasma-Not Illuminated Radiomatic SOO'I'R v3.60/3.60 S/N:423464 User: SUE RunPile: RENEOOOJ Date: 11/10/2000 !O:SIAII Page:! Report Format File: INT_PRT 11.10 ... 3.30 ... (ntin) 0 10 20 1 963 I 0 032 0.491 o (min) 0 10 IS 20 This figure contains the chromatograms from the analysis of plasma with AMT present and without UV exposure The radioactivity counts with respect to time are shown in the top graphic with the UV chromatogram underneath. The UV chromatogram was obtained at 254 nm. The Y scale for the radioactivity counts ranges from 0 to 1970. A large protein peak is again observed in the UV chromatogram beginning at 1.5 minutes. 63

PAGE 75

Figure 4.3 Chromatogram of AMT in Plasma-Illuminated Radiomatic SOOTR v3.60/l.60 S/N:42l464 User: SUE RunFile : RENE0004 Date: 11/10/2000 10:53AI! Page: l Report Format File: INT_PRT IS1l .30 (m.,) 0 10 IS 1.963, 1 4 0 .981 0.491 10 I S 20 This figure contains the chromatograms from the analysis of plasma with AMT and after 2 J/cm2 oflN exposure. The radioactivity counts with respect to time are shown ih the top graphic with the lN chromatogram underneath. The lN chromatogram was obtained at 254 nm. The Y scale for the radioactivity counts ranges from 0 to 1516. A large protein peak is again observed in the lN chromatogram beginning at 1.5 minutes. 64

PAGE 76

T bl 4 1 S n D a e cmt1 at1on etector R I f esu ts rom PI asma p A I rotem nalySIS Counts by peak retention time Sample Description 2.5 min 3 min Smin 11 min Total %Bound counts 1,8 saline blank Nd Nd Nd Nd 2,9 plasma blank Nd Nd Nd Nd 3,10 Not illuminated Nd 2050 1308 42226 45584 7.37 4,11 plasma 1 1639 6801 2169 7373 17982 59.00 5,12 plasma 2 1651 6441 2366 9294 19808 53.08 6,13 plasma V def 1620 6599 1322 7268 16808 56.76 7,14 plasma vm def 2167 5688 1656 6929 16440 57. 85 This table contains the scintillation counts from pertinent peaks in the analytical chromatograms. The sample labeled "saline blank" consisted of salt solution without AMT or plasma and was not illuminated. The sample labeled "plasma blank" did not contain AMT and was not exposed to UV light. The sample labeled "Not illuminated" contained radioactive AMT but did not receive a dose ofUV light. The remaining samples labeled "plasma 1, 2 V def and VID def' all received an UV dose of2 J/cm2 Plasma 1 and 2 were replicate samples while plasma V defand plasma VIII defwere plasma samples deficient in factor V and factor Vill, respectively. The term "Nd" indicates a peak was not detected. The total radioactivity counts decrease by over halfpost-UV exposure. It not clear why this has occurred but may indicate that the tritium atom may be released from the AMT molecule during illumination. 65

PAGE 77

4.4 Discussion Several interesting observations may be noted upon initial review of the data contained in Table 4.1. The peak appearing at 11 minutes is the unbound AMT. This peak has the largest peak area in the chromatogram and is the last to elute. AMT is much smaller than the proteins and unbound AMT should elute later than AMT associated with proteins in the control sample. The observation of two additional peaks besides the unbound AMT peak indicates that dark binding is detectable by this technique. This fact is very surprising as some protein denaturing is expected under these analytical conditions. AMT protein associations have remained intact demonstrating a strong affinity with at least one, maybe two proteins. The radioactive peaks at 2.5, 3 and 5 minutes are believed to be associated with proteins within the plasma. It is difficult to determine exactly which proteins are responsible for the binding unless experiments were conducted with proteins individually to confirm both retention time and AMT association. The UV trace shows one large peak with a long tail beginning about 1.5 minutes. Plasma proteins range in size from 50 ,000 to 900,000 daltons (see Table 4.2). The column used in this analysis is capable of achieving baseline separation between proteins with such a large span in size. The fact that most of the sample elutes at the void volume 66

PAGE 78

indicates that the proteins are not interacting with the stationary phase, as intended. They do not appear to be entering the stationary pores of the column, which would retain the proteins longer and extend the time before they would be detected. 67

PAGE 79

Table 4.2 Prevalent Human Plasma Proteins Protein Size (Daltons) Mean Concentration %of Total (mg/dl) Plasma Protein Albumin 65,000 4200 64 IgM 900,000 1,100 17 Fibrinogen 340,000 350 5 Coagulation Factors 56,000-330,000 16 >1 IgG 160,000 366 5 Transferrin 80,000 280 4 IgA 170,000 240 4 Data presented in this table was obtained from a hematology reference book.37 68

PAGE 80

One reason for such poor separation and low interaction with the column may be protein denaturing. If the proteins become denatured during the analysis they would not enter the size exclusion pores and would have a much lower retention time than anticipated for their size. Using this analytical method, HSA is expected to have a retention time closer to 8 minutes. No protein peak is observed so late in the UV chromatogram. To achieve more definitive results the analytical method should be refined to ensure the tertiary and quaternary integrity of the proteins is maintained throughout the analysis In this manner, proteins could be separated and identified by retention time In addition, radioactive peaks could be aligned with more defined peaks in the UV trace. The information presented in Table 4.1 demonstrates the amount of AMT bound to protein increases from 7% in the control to over 50% in the illuminated samples. Another study published on covalent AMT/protein binding following UV illumination reported only 36% AMT bound?3 Ultrafiltration was used for detection in this study. In this technique the sample is passed through a pore filter prior to the radioactivity assessment. The filter is designed to trap large molecules in the upper chamber while allowing small molecules to pass through. Unbound AMT would pass 69

PAGE 81

through the filter. Since over half of the radioactivity is not observed post illumination, some of the AMT/protein adduct may not been measured by the ultrafiltration technique. Based upon the information presented in this study, the ratio of AMT bound versus unbound to protein appears to be greater than 36% and may have been underestimated by ultrafiltration analysis previously reported. In Chapter 2 it was postulated that at least half of the AMT was not accounted for post illumination. AMT strongly or covalently associated with protein was eliminated from sample during the extraction process. Plasma proteins are too large and hydrophilic to stick to the solid phase extraction cartridge. As a result they are washed away prior to the final elution step which captures all the small hydrophobic molecules in the sample. Data presented by this analysis corroborates information in Chapter 2 accounting for the missing AMT. AMT interactions with plasma protein are very significant accounting for half the initial concentration within the sample. Psoralens have been shown to interact with albumin in previous studies.23 33 36 These studies suggested that 80 to 90% of all psoralen binding to protein is with albumin. It is likely that a significant portion of AMT protein binding in plasma occurs with albumin as well. The peak eluting at 5 minutes accounts for 60% of the bound AMT signal. Due to the reproducibility of the peak areas and retention times in all samples 70

PAGE 82

it is not likely that albumin is present in all three protein peaks It is very reasonable that 40% of the bound AMT is with proteins other that albumin. Albumin accounts for 60% of plasma protein and performs the function as a non specific protein carrier. Albumin is a very versatile protein can interact with a variety of different molecules requiring transport in the blood stream. Due to its function and prevalence in the blood it is likely that AMT associates with this protein in plasma and may become irreversibly bound during the sterilization process. It has been suggested that AMT may have affinity for tryptophan. 33 Human serum albumin (HSA) has one tryptophan residue in its active form and is a potential AMT binding site.38 The purpose for analyzing the factor deficient plasma as to understand if coagulation factors were interacting with AMT. When either factor V and Vill were missing from the plasma the counts in the 5 minute peak appear to decrease. Additional replicate analyses would be required to determine if the decrease is significant. Any assessment of coagulation factor/ AMT interaction in the plasma will be difficult as they constitute less than 1% of plasma protein. If a source of individual coagulation factors could be located, examination of the AMT interaction could be carried out directly with these proteins. Factor V and Vill deficient plasma was chosen as those 71

PAGE 83

factors have a wider therapeutic benefit. This initial data indicates that the AMT may interact with these factors The decrease in therapeutic effectiveness of factor Vill in chapter 4 also suggests and AMT interaction. This data is not definitive however, and direct analysis is recommended for solid confirmation. The data presented in this study demonstrates that half of the AMT added to plasma is bound to protein post illumination. Although possible AMT targets were explored in this exercise no definitive information was supplied to identify which proteins have formed AMT adducts. Prior to the use of such a sterilization process on blood products in the hospital or clinic additional studies are recommended to determine the major AMT bound proteins and the impact those adducts may have on the therapeutic benefit of the plasma product. Protein quality data will be presented in Chapter 5 to assess the impact of AMT and the illumination process upon the coagulation proteins within the plasma product. 72

PAGE 84

5. Analysis of Protein Function of AMT Treated Plasma 5.1 Introduction As plasma has many therapeutic applications, any process intended to enhance the safety of the blood product by reducing disease transmission must also maintain the efficacy of the product for the clinical setting. One of the main benefits plasma supplies to the medical community is in the treatment of clotting disorders such as hemophilia. There are 12 proteins called clotting factors involved in the complex process of clot formation.37 The clinical manifestation of hemophilia can be caused by a deficiency in one of several or a combination of clotting factors. The type of hemophilia a patient may have is dependent upon which proteins do not function properly causing the clotting cascade to break down. For treatment, specific clotting factors from as many as 1000 different units of plasma are combined and concentrated.10 Patients receive a life sustaining dose of the specific factors they need at regular intervals from the concentrated supply. If the sterilization process reduces the functionality of the clotting factors below a clinically acceptable level, implementing the process would not be a practical solution to the disease transmission dilemma. Factor quality of the plasma must be assessed before and after AMT administration and before and after UV exposure. In this 73

PAGE 85

chapter the coagulation protein activity experiments and data are presented. For this study coagulation factors with the most therapeutic benefit were examined individually. This data will not directly determine which proteins may have AMT interactions but will indicate which proteins are most affected by the sterilization process. In addition, the data collectively will assess the overall impact of the process on the plasma. 5.2 Experimental Ten illuminated plasma samples in addition to a control and a blank were prepared as outlined in Chapter 2. These samples were not extracted prior to analysis. The blank consists of plasma without AMT added and without illumination. The "Not illuminated" sample contained AMT but was not exposed to illumination. Samples were irradiated for a dose of3 J/cm2 or 5 J/cm2 Following illumination, 2mL aliquots of each sample were submitted for protein activity assessment. Coagulation analyzer (AMAX Sigma 190 Plus ) was used to determine clotting factor activities within plasma samples. Sample run included analysis of controls to ensure instrument was operating within specifications and that results were accurate. The coagulation cascade is complex and includes many reactions. The proteins analyzed in this experiment include antithrombin, prothrombin, and various factors from 2 to 1 1. 74

PAGE 86

Coagulation factors are assessed one at a time to determine their activity. An aliquot of each plasma sample is placed in a sample cup for the instrument. The instrument is equipped with various reagents depending upon the coagulation factors of interest to be tested. The instrument takes a small portion of the sample and introduces it into a reaction well. Then a reagent plasma, deficient in the factor being tested, is introduced into the reaction well. When the starter reagent is added to the reaction well to initiate the clotting cascade the timer begins counting. There are two different detection methods employed to determine the clotting time. For most factors the optical detection method is used. A beam of light is projected through the reaction well. Once the light transmission through the cell has decreased by 10 to 30 milliequivalence (me) absorbance units clot formation is complete. In the case ofF2 or fibrinogen, a mechanical method of detection is used. At the initiation of the analysis, a metallic ball inside the reaction vessel is in contact with a magnet. As the clot forms the metallic ball is pulled away from the magnet. The end of clot formation is signaled once the metallic ball is no longer in contact with the magnet. For both detection methods, the clotting time for each protein is compared to nominal values and presented as a functionality percentage ofthe nominal value. In the case 75

PAGE 87

of F2 the clotting time is converted to mg/dL through the use of a standard curve. The longer it takes to reach the specified endpoint signal the lower the protein quality in the sample. 5.3 Results Data for this experiment are located in Table 5.1 76

PAGE 88

Table 5.1 Protein Function Data Sample Pro(F2) FV FVII FVIII FX FXI Anti-thrombin (mg/dL) (%) (%) (%) (%) ( 0/o) thrombin (%) (%) blank 98 34 27 26 25 37 33 36 Not 103 35 28 21 22 38 27 34 illuminated 31-1 76 30 20 14 10 26 19 34 31-2 67 31 22 14 11 28 19 33 31-3 82 31 21 14 12 27 19 35 3 J -4 75 30 19 14 11 26 19 33 3J-5 76 30 21 14 11 27 19 34 SJ-1 53 26 16 11 8 21 16 35 5 J-2 78 28 21 14 11 26 19 34 SJ-3 85 32 21 15 11 29 20 37 51-4 78 29 21 14 10 26 19 33 51-5 77 28 21 14 11 27 19 35 Avg31 75 30 21 14 11 27 19 34 %blank 77 89 76 54 44 72 58 94 Avg5 J 74 29 20 14 10 26 19 35 %blank 76 84 74 52 41 70 56 97 This table shows the results of protein quality assessments of coagulation factors from plasma samples with and without AMT as well as with and without lN exposure. The sample labeled "blank" did not contain AMT nor was it exposed to lN light. The sample labeled "Not illuminated" contained AMT but was not exposed to UV light. The samples labeled 31 1-5 contained AMT and received an UV dose of 3 J/cm2 while the samples labeled 51 1-5 also contained AMT and received an lN dose of5 J/cm2 F2 also known as fibrinogen is reported in mg/dL instead of%. F2 is commonly reported with these units and only reflects the concentration of active protein within the sample. Averages of the 31 and 51 samples were calculated and divided by the blank values to determine the impact of the sterilization process on the coagulation factors. Average values were listed as a percentage of the blank values in Table 5.1. 77

PAGE 89

5.4 Discussion A review of Table 5.1 shows some change in activity for all factors tested post illumination. Interestingly, the control also showed some moderate decline in factors VII, VXl and maybe VIII. AMT may combine with factors to reduce protein function or initiate reactions that reduce coagulation performance. Additional samples would be required to determine a statistically significant decline. In the post illumination samples a 50% decrease in protein function was observed in factors VIII, Vll and XI. The effect of illumination upon antithrombin was very small (3%). For the blood banking industry, a reduction in activity of 50% is considered significant and defines the lower acceptable activity limit. All other factors showed a decline of 15 to 25% in activity. Of all the factors examined in this study, a decline in factor VIII would be expected before any of the other proteins as it is the most labile. This factor is also one of the most therapeutically useful as a majority of hemophiliacs have a factor vm deficiency. There does not seem to be a difference between the samples illuminated for 3 and 5 J/cm2 This result suggests the protein damage may have occurred rapidly upon the initiation ofUV exposure. 78

PAGE 90

The data in this study show the sterilization process to have negative impact upon protein quality. The impact of AMT alone has not been definitively shown but any impact is moderate Factors Vll, VIll and XI suffered the most damage and may not be considered of therapeutic use post sterilization. All remaining proteins, except antithrombin, suffered some damage but activity levels were higher than for factors Vll, VIll and XI. Antithrombin did not show a significant decrease in activity level post sterilization. The remainder of the thesis will focus on the toxicity issues associated with AMT. Chapter 6 will review the results of a cytotoxicity screen performed on AMT in media pre and post illumination. 79

PAGE 91

6. Cytotoxicity of AMT 6.1 Introduction To obtain a general understanding of AMT and AMT photoproduct toxicity a direct contact cytotoxicity test was performed. A cytotoxicity test is designed to determine the biological reactivity of mammalian cell cultures following contact with the material under test. Extreme measure of the test is to determine the extent of apoptosis or cell death that may occur as a result of contacting the material. Cytotoxicity is determined by microscopic evidence of malformation degeneration, sloughing or lysis of cells, or a moderate to severe reduction in cell layer density. An immortalized murine cell line is used in this study. A monolayer of cells is uniformly spread across the test plate. The substance to be tested is applied topically to the test plate. This test is called out in the United States Pharmacopeia 24 and is titled Biological Reactivity Tests In Vitro <87>.39 For this test cells are inspected periodically after administration of the test solution. A score greater than 2 constitutes a test failure. Table 6 1 outlines the grading scale used for this test as well as the criteria for determining each grade. 80

PAGE 92

Table 6.1: Reactivity Grades for Direct Contact Test Grade Reactivi!Y_ Conditions of all cultures 0 None Discrete intracytoplasmic granules, no cell lysis 1 Slight Not more than 20% of cells are round, loosely attached and without intracytoplasmic granules; occasional cell lysed cells are present 2 Mild Not more than 50% of the cells are round and devoid of intracytoplasmic granules; extensive cell lysis and empty areas between cells 3 Moderate Not more than 70% ofthe cell layers contain rounded cells and/or are lysed. 4 Severe Nearly complete destruction of the cell layers. 81

PAGE 93

6.2 Experimental Strain L929 mouse fibroblast cells were cultured up to a confluent monolayer on a 35 mm diameter plate and exposed 2 mL of media sample. Observations were made at not less than 24 and 48 hours for toxicity. Extracts were diluted 1 part AMT sample to 2 parts of commercial cell media prior to addition to the cell layers. AMT was prepared in media as outlined in Chapter 2. Concentration of AMT in media was approximately 170 ,....M. One 30 mL aliquot of AMT in media was exposed to 3 J/cm2 Two samples were submitted for cytotoxicity testing: one illuminated sample and one non-illuminated sample. The AMT media samples were mixed 1 part to 2 parts of commercially available cell culture media prior to administration on the cell layers. This was done to ensure cell lysis was not caused by an imbalance between intra and extra-cellular tonicity. 6.3 Results and Discussion The control received a scale of 4 after 24 and 48 hours while the illuminated sample received a score of 0 after both time points. 82

PAGE 94

A UV dose of 3 J/cm2 was chosen to produce complete photodegradation of AMT in the hopes that a worst case concentration ofphotoproducts would be generated. It was not discovered until after the cytotoxicity assessment was performed that the additional J/cm2 of energy destroyed not only the AMT but the photoproducts as well. Data where AMT in media was illuminated for 3 11 cm2 showed lower levels of photoproduct generation when compared to similar samples exposed to only 2 J/cm2 (see Tables 2.1 and 2.4). AMT dimer photoproducts were present 1/5 and 1/10 the concentration when illuminated for the longer dose. Heterodimer constitutes 60% of initial AMT concentration in the 2 J/cm2 samples but in this sample accounted for only 14%. Homodimer accounted for 12% of initial AMT concentration in the samples with lower dose but only present at 1% after 3 J/cm2 of exposure The coumarin compounds were also reduced significantly Since the two coumarin compounds compare with less than 3% ofthe initial AMT concentration their toxicological contribution to the sample would expect to be low in a worst case scenario However, the AMT control still produced a toxic response in the murine cells even after dilution with the commercial cell culture media. Therefore AMT at a 83

PAGE 95

concentration of 57 uM is still considered toxic by the results of this test but a concentration of less than 6 uM does not produce a toxic response in this cell line. AMT has also demonstrated mutagenic behavior in the absence oflN light. A mutagenicity study was performed on platelet suspensions containing AMT.23 Platelet suspensions containing up to 160 f-LM AMT were tested against the Salmonella/Mammalian Microsome Reverse Mutation Assay. Illuminated and non illuminated samples were tested. In samples without UV exposure or where the residual AMT concentration was high, a significant number of mutations were detected above background levels. The number of observed mutations decreased with AMT concentration. Results ofthis AMES test show that AMT is mutagenic whereas the photodecomposition products demonstrate no mutagenic activity. The mutagenesis is. postulated to occur via AMT binding to nucleic acids. Cytotoxicity tests are performed because they are rapid and straightforward with endpoints which can be measured reproducibly and accurately.40 The disadvantage to such tests however, is that the assays are not mechanistically based and do not usually provide information as to how and why the chemicals tested cause irritation. In addition, it is difficult to use cytotoxicity test results as a predictor of in vivo assessments. 84

PAGE 96

AMT was shown to be cytotoxic to the cells examined in this assay. AMT has also been shown to be genotoxic in the literature. Even though AMT induced a toxic response in these cells, such data should be combined with a more strenuous toxicological regiment to understand the mechanism of toxicity. Assays with a more mechanistic design such as a liver assay using cytochrome P450 as a marker would help to determine the potential metabolic interactions of AMT in a living system. Since the concentration of AMT photoproducts in the illuminated were not a worst case representation, a comparison of the toxicity of the sample pre and post UV exposure cannot be made. Additional toxicological assessment of the AMT dimers is especially recommended as they have the greatest peak areas of all the photoproducts in the sterilized blood matrix 85

PAGE 97

7. Toxicity Concerns 7.1 Introduction Psoralen compounds have been examined for use in blood sterilization technologies specifically for their mutagenic and toxic effects on DNA in the presence ofUV light. Finding a compound suitable for inhibition of virus and bacteria within blood is a challenge, as the compounds that are effective have to meet a very rigorous set of requirements. As outlined in Chapter 2, it must have much high DNA affinity, low affinity for proteins and cell membranes and sufficient solubility for application in a blood matrix.1 AMT, in particular, has meet some ofthe criteria as it has sufficient water solubility and shown a high DNA affinity through demonstration of acceptable levels ofviral inactivation .33 Much has been learned about how the type and placement substituents on the psoralen molecule affect its ability to produce lethal affects with UV exposure. Research studies have found that substituents on the psoralen molecule dramatically impact its ability to form DNA adducts. Methylation of psoralen molecule, in general, increases dark binding affinity, the quantum yield of photoaddition and the quantum yield of photo breakdown of the compound. Dark binding affinity describes the ability for the molecule to intercalate in the DNA and the strength of the non-covalent interaction. 86

PAGE 98

Substituents on the C4 position are thought to cause steric interference with the thymidine C5 methyl group as psoralens with C4 substituents form less than 2% pyrone side adduct. Substituents at the 3 and 4 positions appear to decrease the photoactivity of the psoralen as steric effects exist with the methyl group on the 5 position of the thymidine .19 There is evidence that molecules with only bifunctional unsaturated groups at 3,4 and 4',5' positions produce lethal and mutagenic effects when added to bacterial cultures supporting the crosslinking theory. Monoadducts and even diadducts or interstrand crosslinking can occur when psoralen is relatively free from steric hinderence at both the furan and pyrone sides of the molecule Psoralen affinity for DNA will be maintained for molecules that survive illumination and are subsequently infused into a patient. From the experiments presented in this thesis, anywhere from 15 to 60 p.M of AMT is present in plasma post UV exposure and will bind to patient DNA once infused. The impact of dark binding by psoralens on DNA replication and protein expression in humans is not mentioned in the literature. Presumably the effects of dark binding would be minor in humans for this particular application given the milligrams each patient would receive from each 87

PAGE 99

procedure. Due to the toxic nature of psoralens and AMT specifically these impacts should be investigated prior to implementing such a technology in the field. 7.2 Literature Review For much of the toxicological literature reviewed hereafter, studies involving 8methoxypsoralen will be presented. Historically 8-methoxypsoralen has been the drug used to treat the skin conditions mentioned previously and is connected with the side effects observed clinically. Toxicology studies using AMT specifically are not present in the literature due to the lack of medicinal therapies involving its use. For purposes of understanding psoralen toxicity in general, toxicological studies involving all psoralens will be presented and statements relating to AMT will be made where possible. The toxic nature of psoralens in humans has been reported in the literature. As was mentioned in the introduction, patients who had been treated for conditions such as vitiligo and psoriasis with psoralens manifested complications such as erythema (redness ofthe skin due to congestion ofthe capillaries), edema (swelling), . genotoxicity, risk of skin cancer and cataracts. 30 In Chapter 6, data were presented demonstrating both cytotoxic and mutagenic behavior of AMT. Psoralens such as 8methoxypsoralen have also been under investigations for liver toxicity.41 88

PAGE 100

In many toxicology assessments of drugs, impairment or changes of hepatic activity are examined as the liver performs the function of sequestering or breaking down substances for elimination. If a new drug is toxic to the body it often manifests itself in the liver first. The kidneys are also involved in detoxification and may be included in a toxicology assessment as well. The liver and the kidneys constitute the primary ways of eliminating substances from the body. If the substance is not sequestered by either of these mechanisms, the compound may often be "stored" in the fat c e lls or within cell membranes especially for those compounds that are hydrophobic. If the substance is not eliminated, and exposure extended over time long term toxic effects may be exhibited. In the case of psoralens it does appear that breakdown and elimination does occur hepatically. Both furan and pyrone ring opened urinary metabolites of 8methoxypsoralen were detected subsequent to cytochrome P450 (liver enzyme) mediated oxidative attack on the molecule .41 It has been shown that 8methoxypsoralen is metabolically activated by both rat and human liver cytochrome P450 The metabolites ofthis reaction in tum then inactivate P450. Through investigations of coumarin and other psoralen derivatives it has been suggested that a both furan and pyrone ring metabolites contribute to the inactivation of the enzyme. Methylation of the furan ring appears to retard enzyme inhibition and presumably 89

PAGE 101

AMT would not be as reactive toward this enzyme. Investigations of AMT cytochrome P450 have not been presented in the literature. In Chapter 4 the high affinity of AMT for serum proteins was demonstrated. Additional studies have shown that AMT as well as other psoralens have a high affinity for albumin.33 42 Dark binding ofpsoralens may play an important role in delivery of the drug to target organs in the body.30 At least one psoralen binding site mainly the tryptophan residue on human serum albumin has been presented. 33 Other research using a Scatchard analysis proposed at least four binding sites for 5methoxypsoralen. 3 4 The ability of psoralens to interact once circulating in the body is greatly facilitated with the aid of active transport into the cells. Due to the non-polar nature of poralens it is likely there are additional targets besides DNA and proteins. Hydrophobic lipoproteins and cell membranes are other possibilities for psoralen interactions. Studies have shown furocoumarins will form adducts with lipids including those which form cell membranes.43 Many cell membranes are involved in hormone regulation as well as control the transport of ions and molecules in and out of the cell. hnpact upon hormone regulation may have a wide range of effects from metabolism to cell division. 90

PAGE 102

Besides formation of adducts via cycloaddition other interactions of psoralens with fatty acids may include lipid oxidation. Evidence has been presented suggesting the generation of singlet oxygen species in PUV A therapy. 30 Singlet oxygen can lead to lipid oxidation and peroxidation. Lipid peroxidation can lead to degradation ofthe cell membrane and cell death. Cancer, strokes, asthma, senescence associated with aging and other serious health conditions have been associated with lipid peroxidation It is reasonable to state psoralen induced lipid oxidation may be the cause of cytotoxicity other bioeffects of furocoumarins. Peroxidation reactions have greater significance for sterilization of cellular products. Investigations for this thesis were focused on a plasma matrix. Blood sterilization applications will also need to include more the complex platelet and red cell products to further reduce disease transmission from transfusion. In these cases cell membrane damage is of paramount concern as the cellular component provides the therapeutic benefit. Some AMT sterilization studies involving platelet concentrated have been published. In these investigations general assays on platelet quality post-sterilization have been performed. No investigations of lipid membrane peroxidation were reviewed 91

PAGE 103

It has been suggested that the presence of acidic hydrogens on a photosensitizer facilitates hydrogen bonding with the phosphate segment ofphospholipids.44 Such interactions can thermodynamically favor accumulation of the sensitizers within cell membranes. AMT has two such acidic hydrogens on the ammonium side chain at the 4' position. For any AMT sterilization technique involving cells the use of quenchers or some other means may be needed to protect the integrity and viability of cell membranes during exposure. Additional side effects of psoralen-induced lipid peroxidation may include anti proliferative effects on cells. Psoriasis patients treated with 8-methoxypsoralen have shown anti-proliferative effects on the treated skin.15 Protein kinease C helps to regulate enzymes involved in cell metabolism. Studies have also shown that this same enzyme also responds to phosopholipids and such modified lipids may have an effect on the enzyme. Protein kinease C has essential regulatory enzyme roles for cell metabolism. Examinations of psoralen modified phospholipids with respect to cell metabolism and proliferation have not been presented in the literature at this point. Until investigations are reported, the mechanism for the anti-proliferative effects observed clinically will remain unclear. 92

PAGE 104

Data presented as to psoralen chemistry indicates a myriad of potential interactions within the body. Many of the interactions have not been completely investigated. One such area in need for investigation is the ability of furocoumarins generate singlet oxygen species during illumination. Some studies indicate that the photodecomposition products have the ability to initiate oxidation reactions in the dark. 30 Potential reactions such as these may involve the inactivation of proteins or oxidation of fatty acids. With all the literature reviewed nothing has been presented on the reactions of the AMT photoproducts identified in the data presented in this thesis. The furan:pyrone AMT heterodimer and pyrone:pyrone AMT homodimer are of greatest concern they may be the photoproducts generated in the largest concentration within the plasma matrix. It is not clear if these dimers maintain the ability to intercalate in DNA, be transported through the body via albumin, or the affinity of this molecule for lipid membranes. The liver may still metabolize the dimers. If the liver is still involved, there may be metabolite interactions with cytochrome P450. As for the reactions of the coumarin photoproducts, toxicity studies indicate that coumarin is has greater interactions with cytochrome P450 than psoralen.45 There is also data demonstrating toxicity of coumarin in the mouse lung cells as little as 2 days 93

PAGE 105

following treatment with a high dose (200 mg/kg).46 Of the AMT photoproducts identified in this thesis, coumarin photoproducts may constitute only a small fraction of the all products generated during illumination (accounting for less than 5% ofthe initial AMT peak area). However, since they have been identified they should still be investigated for toxic effects in living systems. It is recommended that additional effort be focused on the biological interactions of AMT protein adducts and photoproducts. AMT has been demonstrated to bind effectively to bacterial DNA and achieve 4 and 5 log reduction of titers within a biological matrix. It is also clear through data and other literature presented in this thesis that biological targets for AMT and psoralens other than DNA are significant. Through proper investigations of the lack of selectivity of AMT within a living system it may be discovered that the potential damage of this compound within the human body may outweigh the benefit of the bacterial sterilization process. Such a conclusion would turn research efforts to look for alternative compounds for use in a blood sterilization process that still achieves the desired pathogen kill without providing toxic side effects in the human body. 94

PAGE 106

References 1. Goodrich, R.P ., Platz, M. S. "The Design and Development of Selective Photoactivated Drugs for Sterilization of Blood Products." Drugs ofthe Future. 22: 159-171; 1997 2. Ward, J.W. et al. "The Natural History ofTransfusion-Associated Infection with Human Immunedeficiency Virus." The New England Journal of Medicine. 321: 947-952; 1989 3. Selik, R.M., Ward, J.W., Buehler, J.W., "Trends in Transfusion-Associated Acquired Immune Deficiency in the United States, 1982 through 1991." Transfusion. 33: 890-893; 1993 4. Leveton, L.B., Sox, H.C., Stoto, M.A., HIV and the Blood Supply : An Analysis of Crisis Decision Making." Transfusion. 36: 920-927; 1996 5. Cumming, P.D. et al, "Exposure of Patients to Human Immunedeficiency Virus Through the Transfusion of Blood Components That Test Antibody-Negative." The New England Journal ofMedicine. 321: 941-946; 1989 6. Dodd, R. Y. "The Risk of Transfusion Transmitted Infection." The New England JoumalofMedicine. 327: 419-421; 1992 7. Glynne, S.A. et al. "Trends in Incidence and Prevalence ofMajor Transfusion Transmissible Viral Infections in US Blood Donors, 1991 to 1996." JAMA. 284: 229-235; 2000 8. Tocci, L.J., Napychank, P.A., Cable, R.G., Snyder, E.L. "The Effect of Solvent/DetergentTreated Plasma on Stored Platelet Concentrates." Transfusion. 33: 145-149; 1993 9. Leebeek, F.W.G, Schipperus, M.R., Van Vliet, H.H. "Coagulation Factor Levels in Solvent/Detergent Treated Plasma." Transfusion. 39: 1150-1151; 1999 10. Horowitz, M.S., Rooks, C., Horowitz, B., Hilgartner, M.W. "Virus Safety of Solvent/DetergentTreated Antihaemophillic Factor Concentrate." The Lancet. July 23 : 186-188; 1988 95

PAGE 107

11. Song, P., Tapley, K.J. "Photochemistry and Photobiology ofPsoralens." Photochemistry and Photobiology. 29: 1177-1197; 1979 12. Lin, L., Wiesehahn, G.P., Morel, P.A., Corash, L., "Use of 8Methoxypsoralen and Long-Wavelength Ultraviolet Radiation for Decontamination ofPlatelet Concentrates." Blood. 74: 517-525; 1989 13. Redfield, D., Richmann, D., Oxman, M., Kronenberg, L., "Psoralen Inactivation of Influenza and Herpes Simplex Virus and Virus-Infected Cells." Infection and Immunity. 32: 1216-1226; 1981 14 Hanson, C.V., "Photochemical Inactivation ofViruses with Psoralens: An Overview." Blood Cells 18: 7-25; 1992 15. Pathak, M.A., Fitzpatrick, T.B. "Relationship ofMolecular Configuration to the Activity ofFurocoumarins which increase the cutaneous responses following long wave ultraviolet Radiation." Journal oflnvestigational Dermatology 32: 255-262; 1959 16. Pinkus, H. "Clinical Applications ofPsoralens, and Related Materials. Journal of Investigational Dermatology. 32: 281-284 ; 1959 17. Caffieri, S., Favretto, D. "UV -A Photolysis of Khellin: Products and reaction mechanism." Journal ofOrganic Chemistry 58: 7059-7063; 1993 18. Cimino, G.D., Gamper H., Isaacs, S T Hearst, J.E "Psoralens as Photoactive Probes ofNucleic Acid Structure and Function: Organic Chemistry, Photochemistry and Biochemistry." Annual Review of Biochemistry. 54 : 11511193; 1985 19. Pfluger, C.E., Ostrander, R.L. "The Direct Observation of a PsoralenThymine UV A inducted Solid-State Cycloaddition Reaction Product by Single Crystal XRay Difftactometry." Photochemistry and Photobiology. 49: 375-379; 1989 20. Pathak, M A., Worden L.R. Kaufman, K.D "Effect of Structural Alterations on the Photosensitizing Potency ofFurocoumarins and Related Compounds Journal of Investigative Dermatology. 48: 103118 ; 1967 96

PAGE 108

21. Margolis-Nunna, H., Williams, B., Rywkin, S., Geacintov, N., Horowitz, B. "Virus Sterilization in Platelet Concentrates with Psoralen and Ultraviolet A Light in the Presence ofQuenchers." Transfusion. 32(6): 542-547; 1992 22. Moroff, G., Wagner, S. Benade, L., Dodd, R.Y. "Factors Influencing Virus Inactivation and Retention of Platelet Properties Following Treatment with Aminomethyltrimethylpsoralen with Ultraviolet A Light." Blood Cells. 18:43-56; 1992 23. Wagner, S. et al. "Determination ofResidual4'-Aminomethyl-4,5'8-Trimethylpsoralen and Mutagenicity Testing Following Psoralen Plus UV A Treatment ofPlatelet Suspensions." Photochemistry and Photobiology 57: 819824; 1993 24. Gervais, J De Schryer, F.C. "Photochemistry of Some Furo (3,2-g)-Coumarin and 2,3-Dihydrofuro (3,2-g)-Coumarin Derivatives." Photochemistry and Photobiology. 21,: 71-75; 1975 25. Logani, M.K., Austin, W A., Shah, B., Davies, R.E. "Photooxidation of8-Methoxypsoralen with Singlet Oxygen Photochemistry and Photobiology. 35: 569-573; 1982 26. Kao, J.P., Isaacs S.T., Hearst, J.E. "The Molecular and Stereochemical Structures ofPhotoproducts Generated by UV-Irradiation of 4' -Hydroxymethyl,4,5' ,8trimethylpsoralen in Aqueous Solution." Photochemistry and Photobiology. 51: 273-283; 1990 27. Marley, K.A., Larson, R.A. "A New photoproduct from Furocoumarin Photolysis in Dilute Aqueous Solution: 5-Formyl-6-Hydroxybenzofuran." Photochemistry and Photobiology. 59(5): 503-505; 1994 28. Gasparro, F.P., Chan, G., Edelson, R.L. "Phototherapy and Photopharmacology." Journal ofBiological Medicine 58: 519; 1985 29. Gambro BCT Laboratory Notebook #1706-99-010, pp. 15 30. Gasparro, F.P. Psoralen DNA Photobiology. Volumes I and ll. Boca Raton, FL: CRC Press; 1998. 97

PAGE 109

31. Fowlks, W.L. "The ChemistryofPsoralens." Journal oflnvestigative Dermatology. 2: 249-254; 1959 32. Abu-Eittah, R.F., El-Tawil, R. The Electronic 9-Absorbtion Spectra of Some Coumarins. A molecular orbital treatment." Canadian Journal of Chemistry. 63: 1172-1179; 1985 33. Veronase, F.M., Bevilacqua, R., Schiavon, 0., Rodighiero, G. "The Binding of8-Methoxypsoralen by Human Serum Albumin." II Farmaco Ed. Sci. 33(9): 667675; 1978 34. Veronase, F.M., Bevilacqua, R., Schiavon, 0., Rodighiero, G. "Drug Protein Interaction: Plasma Protein Binding ofFurocoumarins. II Farmaco Ed. Sci. 34(8): 716-725; 1979 35. Veronase, F .M., Schavon, 0., Bevilaqua, R., Bordin, F., Rodighero, G. "The Effect ofPsoralens and Angelicans on Proteins in the Presence ofUV A Irradiation." Photochemistry Photobiology. 34: 351-354; 1981 36. Yoshikawa, N.M., Mori, N., Sakakibara, S., Mizuno, N., Song, P.S. "Photo conjugation of 8-Methoxypsoralen with Proteins." Photochemistry and Photobiology. 29: 1127-1133; 1979 37. Hoffman, R., Benz, E.J., Shattil, S.J., Furie, B., Cohen, H. J., ed. Hematology Basic Principles and Practice. New York, NY: Churchill Livingstone; 1991 38. Dugaiczyk, A., Law, S., Dennison, 0. "Nucleotide Sequence and the Encoded Amino Acids of Human Serum Albumin mRNA." Proceedings of the National Academy of Sciences. 79: 71-73; 1982 39. "Biological Reactivity Tests, fu Vitro." U.S. Pharmacopeia 24. 1831-1832; 2000 40. Gad, S.C. In Vitro Toxicology. New York, NY: Raven Press; 1994 41. Wilkinson, D.J., Fry, J.R. "Rat Liver Cytochrome P450-mediated Metabolic Activation ofMethoxsalen and Structurally Related Compounds and its Relation to Enzyme Inhibition." Journal of Pharmacy and Pharmacology. 47: 79-84; 1995 98

PAGE 110

42. Artuc, M., Stuettgen, G Schalla W Schafer, H Gazith, J "Reversible Binding of 5and 8-Methoxypsoralen to Human Serum Proteins and to Epidermidus in Vitro British Journal of Dermatology. 101(6) : 669; 1979 43. Melo, T. deS., Morliere, P., Goldstein, S., Santus, R., Dubertret, L., Lagrange, D. "Binding of 5-MOP to Human Serum Low Density Lipoproteins." Biochemistry and Biophysics Research Communications. 120(2): 670-676; 1984 44. Goodrich, R.P et. AI. Selective Inactivation ofVirsuses in the Presence of Human Platelets: UV Sensitization with Psoralen Derivates." Proceedings of the National Academy of Sciences USA. 91: 5552-5556; 1994 45. Zhuo, A Gu, J., Zhang Q., Spink, D.C., Kaminsky, L.S., Ding, X. "Biotransformation of Coumarin by Rodent and Human Cytochromes P-450: Metabolic Basis of Tissue-Selective Toxicity in Olfactory Mucosa of Rats and Mice." The Journal of Pharmacology and Experimental Therapeutics. 288 (2): 463-471; 1999 46. Born, S.L. Fix, A.S ., Caudill, D., Lehman-McKeeman, L.D. "Selective Clara Cell Injury in Mouse Lung Following Acute Administration of Coumarin." Toxicology and Applied Pharmacology. 151: 45-56 ; 1998 99