Matthias Kresse - KOPS - Universität Konstanz

Mechanism of liver cell injury
by the cytostatic drug melphalan
DISSERTATION
zur
Erlangung des akademischen Grades eines
Doktors der Naturwissenschaften (Dr. rer. nat.)
des Fachbereiches für Biologie
an der
Universität Konstanz
vorgelegt
von
Matthias Kresse
Tag der mündlichen Prüfung: 11. Februar 2005
1. Referent: Prof. Dr. A. Wendel
2. Referent: Prof. Dr. A. Bürkle
Konstanz, im Februar 2005

Acknowledgement
This thesis was prepared between November 2000 and September 2004 at the Chair of
Biochemical Pharmacology of the University of Konstanz under the instruction and
supervision of Prof. Dr. A. Wendel.
My special thanks go to my supervisor Prof. Wendel. He did not only provide excellent
working facilities and a generous financing of the project, but also enabled this study by
his patience, his encouragement, his criticism and stimulating ideas. Moreover he was
always open to new ideas concerning projects or events and encouraged me to realise
them by his support.
This work benefited greatly from a number of excellent collaborations and discussions
with other groups and within the DFG graduate trainee program “Biomedical Drug
Research” directed by Prof. Dr. A. Wendel and Prof. Dr. K.P. Schäfer (ALTANA Pharma,
Konstanz). This program enabled me to attend conferences, local and international
training courses which provided a basis for fruitful discussions and insights in different
fields of scientific and industrial interests of research. I am grateful for this support and
the grant supplied by the graduate trainee program.
I would like to extend my gratitude to Prof. Dr. R. Schmid for the excellent collaboration
on the NFκB-work, to Dr. H.-M. Riehle for the help with histology and the interesting
discussions, to Dr. N. Van Rooijen for providing Chlodronate Liposomes, to Dr. M.
Menges for providing CD11b Microbeads, to Prof. Dr. K. Pfizenmaier for providing
several mouse strains and to Prof. P. Scheurich for providing modified human fibroblasts.
Very special thanks go to Prof. Dr. M. Schwarz and E. Zabinski for their great support in
the tumor project and to Dr. N. Dikopoulos who has not only turned out to be a perfect
partner in scientific joint-ventures, but also as a very good friend.
Me and this work benefited also greatly from collaborations and discussions with a lot of
members of ALTANA Pharma, namely Prof. Dr. K.-P. Schäfer, PD Dr. C. Schudt, Dr. R.
Beume, Dr. M. Eltze, Dr. Th. Klein, Dr. Ch. Hesslinger, and K. Graf.
Within the chair of Biochemical Pharmacology I am especially grateful to Dr. G. Künstle
and Dr. R. Lucas for their supervision of the project and their help in technical questions,
Dr. Th. Meergans for his advise and support, Dr. C. Hermann for her help concerning
endotoxin and last but not least M. Weiller, Ch. Hart, Dr. M. Latta, Dr. C. Braun, Dr. K.
Eichert and Dr. Markus Müller for their friendship, support and valuable discussions.
Further thanks go to A. Hildebrandt and U. Gebert for their excellent technical
assistance in protein analysis, ELISA-technique and animal experiments and to Gudrun
Kugler for her never-ending patience and support in all kinds of daily problems.
Finally, on a personal note, I would like to thank my family and my friends for their
everlasting support that I’m very grateful for.
Konstanz, September 2004 Matthias Kresse

Table of contents
1 Introduction _____________________________________________________________10
1.1 From battlefield to bedside – History of chemotherapy ____________________________10
1.2 Alkylating agents ___________________________________________________________11
1.3 Mechanism of action and cytotoxicity __________________________________________13
1.3.1 Cyclophosphamide _______________________________________________________________ 15
1.3.2 Melphalan ______________________________________________________________________ 17
1.4 Chemical and pharmacological aspects of melphalan______________________________18
1.5 Nitrogen mustards - Clinical aspects of tumor-therapy ____________________________19
1.5.1 Systemic application of melphalan __________________________________________________ 20
1.5.2 Isolated limb and liver perfusion____________________________________________________ 21
1.6 TNF – heaven and hell in cancer therapy _______________________________________26
1.7 A whole family of cytotoxic cytokines – the TNF superfamily _______________________27
1.7.1 CD95, Fas and APO-1 ____________________________________________________________ 29
1.7.2 TRAIL _________________________________________________________________________ 30
1.8 Combinatorial therapy_______________________________________________________31
1.8.1 TNF – the good, the bad or the ugly in ILP ___________________________________________ 32
1.8.2 Investigation of hyperthermia in ILP ________________________________________________ 32
1.8.3 Pre-clinical results of melphalan in cancer therapy_____________________________________ 34
1.9 Apoptosis __________________________________________________________________35
1.9.1 Signaling of TNF receptors in apoptosis and anti-apoptosis______________________________ 36
1.9.2 The role of TNF receptor 1 ________________________________________________________ 37
1.9.3 The role of TNF receptor 2 ________________________________________________________ 38
1.9.4 Anti-apoptotic signaling ___________________________________________________________ 39
1.9.5 Modulation of apoptosis by changing cellular redox and energy state _____________________ 40
1.9.6 Differences in apoptosis of primary versus transformed cells ____________________________ 42
1.9.7 Apoptosis in cancer_______________________________________________________________ 44
2 Objectives of the thesis _____________________________________________________46
3 Materials and methods _____________________________________________________47
Materials and animals _____________________________________________________________47
3.1 Chemicals _________________________________________________________________47
3.2 Antibodies and recombinant proteins __________________________________________47
3.3 Cell culture materials________________________________________________________48

Table of contents
3.4 Animals ___________________________________________________________________48
Methods _________________________________________________________________________49
3.5 Cell culture ________________________________________________________________49
3.5.1 Cell lines________________________________________________________________________ 49
3.5.2 Isolation and culture of primary cells ________________________________________________ 50
3.6 Isolated liver perfusion ______________________________________________________51
3.6.1 Technical procedure of liver preperation and perfusion_________________________________ 51
3.6.2 Treatment schedules______________________________________________________________ 53
3.6.3 Sampling of material _____________________________________________________________ 54
3.7 Animal experiments _________________________________________________________54
3.7.1 Treatment schedules______________________________________________________________ 54
3.7.2 Sampling of material _____________________________________________________________ 55
3.8 Tumor model_______________________________________________________________56
3.8.1 Tumor induction _________________________________________________________________ 56
3.8.2 Isolated liver perfusion____________________________________________________________ 56
3.8.3 Treatment schedules______________________________________________________________ 56
3.9 Cytokine determination ______________________________________________________57
3.10 Light microscopy ___________________________________________________________58
3.10.1 HE-staining _____________________________________________________________________ 58
3.10.2 Tunnel-staining __________________________________________________________________ 58
3.11 Measurement of enzyme activities _____________________________________________58
3.11.1 Liver enzyme activities in plasma or samples _________________________________________ 58
3.11.2 Lactate dehydrogenase activity _____________________________________________________ 59
3.11.3 Caspase 3,7-like activity ___________________________________________________________ 59
3.11.4 TNF-α-converting enzyme (TACE)-activity___________________________________________ 60
3.12 Determination of ATP _______________________________________________________60
3.13 Determination of glutathione__________________________________________________61
3.14 Cytokine determination ______________________________________________________61
3.15 Cell-based ELISA___________________________________________________________61
3.16 Western blot analysis ________________________________________________________62
3.17 Statistics___________________________________________________________________63
4 Results__________________________________________________________________64
4.1 Characterisation of melphalan-induced cell death in primary murine liver cells _______64
4.2 The role of death receptors in melphalan-induced apoptosis________________________67

Table of contents
4.3 TNF mediates melphalan-induced toxicity ______________________________________69
4.4 Kupffer cells are the source of TNF mediating melphalan-induced toxicity ___________70
4.5 Kupffer cells lead to enhanced TNF secretion after LPS but not after melphalan
stimulation_________________________________________________________________72
4.6 Melphalan inhibits recombinant TACE activity in vitro ___________________________76
4.7 Melphalan is able to block TNF-dependent liver injury in vivo _____________________77
4.7.1 GalN/TNF and GalN/LPS model____________________________________________________ 77
4.7.2 GalN/SEB ______________________________________________________________________ 80
4.7.3 The Con A model. ________________________________________________________________ 81
4.7.4 The CD95 model _________________________________________________________________ 83
4.7.5 Melphalan and melphalan/TNF model _______________________________________________ 84
4.8 Cell-cell contact is necessary between Kupffer cells and hepatocytes _________________84
4.9 Role of TNF receptors on Kupffer cells _________________________________________85
4.10 Melphalan-induced toxicity on Kupffer cells_____________________________________87
4.11 Melphalan-induced hepatotoxicity in the presence and absence of Kupffer cells _______88
4.12 Melphalan induces TNF-mediated hepatotoxicity in situ ___________________________90
4.13 Modulation of intracellular signaling by various inhibitors_________________________94
4.13.1 Role of NFκB____________________________________________________________________ 94
4.13.2 Role of p38 MAPK _______________________________________________________________ 96
4.13.3 Role of JNK _____________________________________________________________________ 97
4.13.4 Role of PARP____________________________________________________________________ 98
4.14 Influence of melphalan on cellular redox and energy state and vice versa ____________100
4.14.1 Melphalan and cellular redox state _________________________________________________ 100
4.14.2 Melphalan and cellular energy status _______________________________________________ 101
4.15 Effect of melphalan on DEN-induced liver tumors in isolated liver perfusion of tumorbearing
mice ______________________________________________________________106
4.15.1 DEN induced massive tumor growth in murine livers following intraperitoneal injection ____ 106
4.15.2 Isolated liver perfusion of tumor-bearing livers with melphalan/TNF ____________________ 108
4.15.3 Activation of caspases in normal and neoplastic tissue _________________________________ 109
4.15.4 Histological examination of perfused livers derived from tumor-bearing mice _____________ 109
5 Discussion______________________________________________________________112
5.1 The role of TNF and TNF receptors in melphalan-mediated hepatotoxicity __________113
5.2 Poseidon’s trident - mechanism of melphalan-induced cytotoxicity _________________115
5.3 Melphalan-induced upregulation of membrane TNF _____________________________118

Table of contents
5.4 Clinical relevance __________________________________________________________119
5.5 Differential role of TNF receptors in the presence and absence of Kupffer cells_______120
5.6 Intracellular signaling of TNF receptors following melphalan incubation____________122
5.6.1 Nuclear factor κ B_______________________________________________________________ 123
5.6.2 p38 mitogen activated protein kinase and c-Jun N-terminal kinase ______________________ 124
5.7 Modification of melphalan-toxicity by altered cellular redox or energy state _________126
5.7.1 Glutathione ____________________________________________________________________ 126
5.7.2 ATP __________________________________________________________________________ 127
5.8 Influence of ATP depletion on melphalan/TNF-mediated organ toxicity in tumor-bearing
mice _____________________________________________________________________129
6 Summary_______________________________________________________________131
7 Zusammenfassung _______________________________________________________133
8 References______________________________________________________________135

αCD95 agonistic anti-CD95-antibody
ActD actinomycin-D
ALT alanin-aminotransferase
AP-1 activator protein 1
Apaf-1 apoptosis activating-factor 1
ATP adenosintriphosphate
bp basepair
BIR baculoviral inhibitory repeat
BSA bovine serum albumin
BSO buthionine-sulfoximine
CARD caspase recruitment domain
CD cluster of differentiation
CD95 Fas, Apo-1
CD95L CD95-ligand
CHX cycloheximide
Con A Concanavalin A
CR complete response
CTL cytotoxic T lymphocyte
Cyt c cytochrome c
DD death domain
DED death effector domain
DEM diethylmalate
DEN diethlnitrosamine
DEVDafc N-acetyl-asp-glu-val-asp-7-amino-4trifluoromethyl
coumarin
DISC death-inducing signaling complex
EDTA ethylendiamintetraacetic acid
ELISA enzyme-linked immunosorbent assay
FADD Fas-associated protein with death
domain
FAS FS-7-associated surface antigen
FCS foetal calf serum
FLICE FADD-like ICE
GalN D-galactosamine
GPT glutamate-pyruvate-transaminase
GSH glutathione (reduced state)
GSSG glutathione disulphide
GSx glutathione (total)
HAI hepatic artery infusion
HC hepatocyte
HCC hepatocellular carcinoma
HEPES N-2hydroxyethylpiperazi-N’-2ethansulfonic
acid
HSA human serum albumin
HSP heat shock protein
hu human
i.p. intraperitoneally
Abbreviations
i.v. intravenously
ICE interleucin-1ß-converting enzyme
IFNγ interferon gamma
IHP isolated hepatic perfusion
IL interleukin
ILP isolated limb perfusion
IPML isolated perfused mouse liver
JNK JunNH2-terminale kinase
LAL Limulus amoebocyte lysate
LDH lactate dehydrogenase
LPS lipopolysaccharide
mAb monoclonal antibody
MOF multi-organe failure
mu murine
MΦ macrophage
NCS new-born calf serum
NFκB nuclear factor kappa B
NIK NFκB-inducing kinase
NPCs non-parenchymal liver cells
ORR overall response rate
p55 Tumor necrosis factor receptor 1
p75 Tumor necrosis factor receptor 2
L-PAM L-phenylalanine mustard
PAGE polyacrylamide gel electrophoresis
PBS phosphate-buffered saline
PCD programmed cell death
PR partial response
RIP Receptor-interacting protein
RAIDD RIP-associated ICE-homologous protein
with death domain
S.E.M. standard error of means
SD standard derivation
SDS sodiumdodeclysulfate
SEC sinusoidal endothelial cells
TLR TOLL-like receptor
TNF tumor necrosis factor
TNFR1 tumor necrosis factor receptor 1
TNFR2 tumor necrosis factor receptor 2
TRADD TNF receptor-associated protein with
death domain
TRAF2 TNF receptor-associated factor 2
TRAIL TNF-related apoptosis inducing ligand
Tris tris-(hydroxymethyl-)aminomethan
vs. versus
w/o without
zVAD-fmk benzyloyxcarbonyl-val-ala-asp-(OMe)fluoromethylketone

1 Introduction
Introduction
1.1 From battlefield to bedside – History of chemotherapy
During the First World War a new kind of warfare agent was developed by German
chemists, the so-called sulphur mustards, also called LOST in remembrance of their
discoverer Lommel and Steinkopf. The first well-documented military uses took place on
July 12 th 1917 in Ieper (Flanders, Belgium) and Abessinia (Italy) 1935-36 (1), but there is
also evidence that a kind of sulphur mustards had been used in the 4 th century by
Chinese commanders on the battlefield blowing sulphur mustards against their enemies
by the use of fans. In Belgium in the 20 th century at least 70 soldiers were killed, 2000
very heavily poisoned and the number of chronic injuries has never been established.
Sulphur mustards were observed to be taken up very easily through the skin
because of their strong lipophilic behaviour. Subsequently these substances cause
severe lung oedema, irritations of eyes and skin, followed by severe leucopoenia, bone
marrow aplasia, dissolution of lymphoid tissues and severe ulceration in the
gastrointestinal tract. Long term consequences in liver and kidney have still been
observed for decades later. These findings, which suggest that the toxic effect of these
substances targets rapidly dividing cells, raised the idea that there could be a
pharmacological implementation in diseases characterised by uncontrolled cell
proliferation, such as cancer.
The first clinical trial in cancer therapy was published in 1931. A Sulphur mustard
solution was applied topically onto or injected directly into human tumors (2). The
procedure was subsequently abandoned due to the high toxicity in the patients, but the
attempts to cure cancer with substances of this type continued. It was however for more
than ten years later in 1946 that Gilman and Philips published their initial results of
successful clinical trials with this new kind of substances (3). This work is referred to as
the beginning of modern cancer therapy.
Since those early works in the field of chemotherapeutic agents, hundreds of
different substances have been developed and used clinically. Most of them target the
DNA by direct chemical reaction (alkylating agents, platinum compounds), by synthesis
inhibition (topoisomerase inhibitors, anti-metabolites) or by interaction in the cellular
machinery during cell division (taxanes, vinca-alkaloids). Consequently, all anti-cancer
drugs display a high activity against proliferating cells or tissues with a large proportion
of dividing cells. Unfortunately, most tumors have proportions of dividing cells very
similar to what is found in regenerating or strongly proliferating healthy tissues, like the
- 10 -

Introduction
bone marrow, hair follicles, germ cells, the gastrointestinal mucosa and the liver,
contributing to a wide range of side-effects mostly observed with the use of these
compounds. In recent years, this topic attracted much attention and many investigations
have been done to develop clinical strategies, in order to prevent general toxicity,
without interfering with the toxic effects on the tumor itself. In addition to chemotherapy,
surgery and radiotherapy are corner-stones in the treatment of malignancies. Each of
these modalities is associated with specific risks and benefits, according to the individual
prognosis of the patient and therefore modern cancer therapy often combines different
treatment strategies with the aim to concentrate the cytotoxic properties at the tumor
site, while diluting the unwanted. In contrast to surgery and radiotherapy, which can be
considered as a local or regional regimen, chemotherapy is mostly given systemically,
offering a treatment strategy that potentially reaches metastatic cells, even if their
existence is far away or unknown. On the other hand, systemic application is often
accompanied with more extensive and systemic toxicity, leading to the severe sideeffects
mentioned above. To solve this conflict, several strategies have been developed,
which combine surgery and chemotherapy in a regional treatment regimen. However,
despite of all advantages these new techniques and strategies offer, no strategy has
been developed to distinguish between the tumor cell and the neighbouring healthy cell.
This thesis focuses on the molecular basis of cell death induction in primary and
transformed liver cells by melphalan, an alkylating agent derived from nitrogen mustard.
The interest in this topic resulted from the expanded knowledge in tumor biology and cell
death during the last decades, especially in the field of induction and modulation of
apoptosis in normal versus transformed tissue.
1.2 Alkylating agents
Alkylating agents are compounds that react with electron-rich atoms in biologic
molecules, to form covalent bonds. Traditionally these compounds have been classified
into two types, on the one hand compounds that react directly with biological molecules
and follow a so-called SN1-mechanism and on the other hand compounds that form a
reactive intermediate, which then reacts with the biologic molecule (SN2-mechanism).
A large number of chemical compounds have been found to alkylate biological
molecules under physiological conditions, but the most frequently used among them are
derivatives of nitrogen mustards. In the last century hundreds of different nitrogen
mustards have been synthesised and tested, but only five of them are commonly used in
- 11 -

Introduction
cancer therapy today. These are mechlorethamine, cyclophosphamide, ifosfamide,
chlorambucil and melphalan (L-PAM), as illustrated in figure 1.
- 12 -

Introduction
1.3 Mechanism of action and cytotoxicity
Nitrogen mustards exert their cytotoxic action through covalent interaction with
intracellular nucleophils, especially DNA, as a result of spontaneous formation of a
reactive cyclic intermediate, the aziridinium ion (figure 2). Bifunctional agents are much
more effective anti-cancer drugs than monofunctional ones, because these compounds
are able to cross-link a DNA strand within a double helix (intrastrand) (figure 2), between
two strands (interstrand) or between DNA and proteins.
Cross-linking of DNA is probably the most important factor for the cytotoxic effect,
resulting in inhibitory effects on DNA replication and transcription, which subsequently
leads to cell death. It has been shown that nitrogen mustards exert their antitumor
activity only when the two-armed mustard is intact (4). The reactivity of the nucleophilic
sites in the DNA relates to electronic and steric properties, as well as to the involvement
of hydrogen bonds within the double helix.
- 13 -

Introduction
Detailed studies have revealed that the N-7 position of guanylic acid is the most
susceptible site for alkylation (5). Brookes and Lawley also suggested that the nitrogen
mustard cross-link was between the N-7 guanine atoms in base-paired G-C sequences
in the DNA (6). Bauer and Povirik could demonstrate that mustards form two basestaggered
interstrand cross-links between the 5’ G and the G opposite C in the 5’ G-G-C
sequence in double stranded DNA of CHO cells in vitro (7). In contrast to melphalan,
mechlorethamine and phosphoramide were also able to form intrastrand crosslinks
between two contiguous Gs in the G-G-C sequence in double stranded DNA. It was
shown that melphalan is only able to cause intrastrand cross-links in single stranded
DNA with a very slow kinetics of second arm alkylation and therefore cross-links
between two DNA strands can only be formed when monoadducted DNA becomes
transiently single stranded during transcription or replication (7).
Another interesting finding is that patients treated with nitrogen mustards had an
increased risk for secondary myeloma (8). Although the exact mechanism of this finding
remains largely unknown, it was found that aromatic nitrogen mustards like melphalan
can also lead to A-T→T-A transversions, apparently resulting from adenine N-3
alkylations (9, 10).
But alterations of the DNA are not the only effect nitrogen mustards cause.
Investigations on the mechanism of primary lesions caused by nitrogen mustards, lead
to the so called “Papirmeister-hypothesis” (11). In response to DNA damage, the cell
cycle is arrested and a class of enzymes called poly-adenosinediphosphate-ribosepolymerase
(PARP) (other names in literature are poly(ADP-ribose) synthetase (PARS)
or poly(ADP-ribose) transferase (pADPRT)) (12), which is abundantly present in the
nucleus (13, 14), is activated. The obligatory triggers of PARP activation are nicks and
breaks in the DNA strand, which can be induced by a broad variety of environmental
stimuli including alkylating agents (12). On the average, one molecule of this enzyme is
present per 1000 bp of DNA in the nucleus. Upon activation, the enzyme uses
nicotinamide-adenine-dinucleotide (NAD) as a cellular substrate to build up
homopolymers of adenine diphosphate ribose units. In case of massive DNA alkylation,
this leads to a massive decrease of cellular NAD-pools which, in turn leads to inhibition
of glycolysis and energy metabolism, to activation of monophosphate shunts, to the
breakdown of the cellular redox status, the release of proteases and finally to cell death
(15, 16).
Evidence for this phenomenon also comes from a totally independent field of
scientific interest. In 1999, Burkart and co-workers published their data on an
experimental model for type 1 diabetes due to the loss of pancreatic beta-cells. The
- 14 -

Introduction
authors demonstrated that PARP -/- mice were completely protected from streptozocininduced
diabetes (17). Streptozocin (also: Streptozotoxin) is able to induce double
strand breaks in islet cell DNA upon uptake via glucose-transporters type 2 GLUT-2 (16,
17). PARP was shown to be activated and cellular NAD-level decreased to a very low
level, resulting in abrogation of sufficient energy generation and finally beta-cell death
(for illustration see also (18). In PARP -/- mice this effect was totally abolished and mice
were completely protected from Streptozocin-induced hyperglycaemia.
Nitrogen mustards and alkylating agents in general can also react spontaneously
with glutathione, but this aspect will be discussed later more detailed. In the following
section the mechanism of two antitumor compounds, cyclophosphamide and melphalan,
will be discussed in detail, because of their greater clinical relevance during the last
decades.
1.3.1 Cyclophosphamide
Cyclophosphamide is not toxic itself, but it undergoes activation in the liver which
finally leads to a toxic product. The initial activation reaction is carried out by the
microsomal oxidation of cytochrome P450 (CYP2B6) producing 4hydroxycyclophosphamide,
which is predominantly present in that form at physiologic
pH, but which is also in equilibrium with aldophosphamide. This equilibrium mixture is
able to diffuse from the hepatocyte into the plasma and is distributed throughout the
body. Since 4-hydroxy-cyclophosphamide is relatively nonpolar, it is able to enter target
cells in the periphery readily by diffusion. Aldophosphamide instead decomposes
spontaneously to produce phosphamide mustard (for illustration see also figure 3),
which in turn is able to react spontaneously with DNA or proteins as described by other
sulphur- or nitrogen-mustards.
- 15 -

Introduction
Phosphamide-mustard is also produced outside the cell, but this compound is
very polar and enters cells poorly. The intracellular toxification of cyclophosphamide was
of great advantage for clinical investigations. It can be administrated in much higher
doses (up to 170 mg/kg as continuously infusion over 4 days or 110 mg/kg over 90
minutes), as compared to melphalan (0.6 mg/kg intravenously or 0.25 mg/kg orally) for
example. But unlike melphalan, patients treated with cyclophosphamide displayed much
less side effects.
- 16 -

1.3.2 Melphalan
Introduction
Melphalan was first synthesised by two British scientists, i.e. Bergel and Stock, in
the early 1950’s. The two scientists believed that, based on previous results with serinealanine-
and phenylalanine derivatives, certain natural amino acids or peptides might
show antitumor activity, possibly by interfering with the nucleic acid or protein
metabolism of malignant cells (19-21). The first publication focused on derivatives of
phenylalanine, in which three forms of bis(2-chloroethyl)aminophenylalanine were
described, the D-, L- and the racemic D-L-form (table 1). Subsequent testing of these
substances in a Walker rat carcinoma model revealed that only the L-form had a high
antitumor activity, while the D- and the racemic DL-form was much less active. The lead
structure of p-L- bis(2-chloroethyl)aminophenylalanine was later named “melphalan”,
derived from mustard –L-phenylalanine. According to this nomenclature, the D-form was
named medphalan and the racemic DL-form merphalan.
- 17 -

Introduction
In parallel to Bergel and Stock, some Russian scientists Larionov and colleagues
worked on a different compound which was named “sarcolysine” (table 1), because of its
remarkable effect on soft tissue sarcomas (22). Later on it was found that sarcolysine
and merphalan were exactly the same substances and therefore had been often referred
to as merphalan. The racemic compound was only used clinically in Eastern Europe for
a period of time, but never reached the same international interest, as compared to the
L-enantiomer. The reasons for this remain unclear. But it could be related to early
observations of Bergel and Stock, who demonstrated a better cellular uptake of the Lform
in vivo.
In 1965 Schmidt and colleagues published a series of in vivo toxicity experiments
conducted at the US Cancer Chemotherapy National Service Center (23-25). Those
studies focused on the toxicity and anti-tumor activity of 20 reference substances, 39
bis(2-chloroethyl)amino-derivatives, 20 aziridines and 35 methanesulfonates in mice and
rats. For L-forms of phenylalanine mustards a clear tendency for increased toxicity was
found compared to D-forms. Therefore, not surprisingly, the L-enantiomers were clearly
favoured.
1.4 Chemical and pharmacological aspects of melphalan
Melphalan is a bifunctional alkylating agent, chemically known as 4-[bis(2chloroethyl)amino]-L-phenylalanine,
with a molecular formula of C13H18Cl2N2O2 and a
molecular weight of 305.20 g/mol. Melphalan is a white odourless powder that is almost
insoluble in water (pKa of ~2.5), but which is subject to rapid hydrolysis in water and
plasma. It is soluble in ethanol, propylene glycol, alkaline solution, 2%
carboxymethylcellulose and diluted mineral acid. The half-life of melphalan in a
physiological buffer at 37°C is approximately 1.5 hours, but the stability in solution is
influenced by pH, temperature and protein content (plasma proteins) of the solvent and
of the concentration of the drug. In general, more dilute solutions of melphalan degrade
slightly quicker (26). For NaCl-solutions it was found that melphalan is 30% more stable
in 150 mM NaCl solution (normal saline), as compared to Dulbecco’s phosphatebuffered
saline. Stored in normal saline at room temperature the drug lost 5% of its
activity (t0.95) in 1.5 hours, at 4°C the t0.95 was, however, 20 hours (26). In addition to that
Flora, Smith and Cradock reported that hydrolysis of melphalan is inhibited by increasing
chloride ion concentrations (27). In an aqueous solution of 28°C an 80% decomposition
was observed after 24 hours.
- 18 -

Introduction
The pharmacokinetics of intravenously and orally administered melphalan has
been extensively studied in humans. For all other species hardly any experimental data
are available. Following injection drug plasma concentrations declined rapidly in a biexponential
manner. With the distribution phase and terminal elimination phase half-lives
of approximately 75 minutes were calculated. Melphalan is eliminated from plasma
primarily by chemical hydrolysis to mono- and dihydroxy derivatives. Other metabolites
have not been observed in humans. Melphalan is able to pass the blood-brain-barrier,
but the penetration into the cerebrospinal fluid is low. Interestingly, there is evidence that
nitrogen mustards enter the cell via an active cellular uptake since Goldenberg et al.
have been demonstrated that there is evidence for a transport carrier of nitrogen
mustard in nitrogen mustard-sensitive and -resistant L5178Y lymphoblasts. This
assumption was based on the finding that the time course for the uptake of labelled
nitrogen mustard was much more rapid in sensitive lymphoblasts and with a higher
extends compared to resistant lymphoblasts. Additionally cellular uptake could be
inhibited by 2,4-dinitrophenol (DNP, 1 mM) and oubain (0.5 mM) (28, 29). The extent of
melphalan binding to plasma protein is high, ranging from 60% to 90%. After injection of
radiolabelled melphalan, radioactivity is excreted both in urine and faeces (30).
1.5 Nitrogen mustards - Clinical aspects of tumor-therapy
Since 1946, when the first clinical trial with sulphur mustards was presented by
Gilman and Philips a new kind of therapy for patients with cancer concerning various
organs has been born. At the beginning of the 1960s melphalan was the first drug that
improved the therapy of multiple myeloma. The median of survival was prolonged from
several months to 3 years. In the following four decades several new drugs were tested
including cisplatin, mitomycin c and 5-fluorouracil. But no other drug brought better
results than melphalan (31). From 1950 till today numerous articles and clinical studies
have been reported, demonstrating enhanced success with different application forms,
clinical protocols and combinatorial therapies of nitrogen mustards, cytokines and
hyperthermia. Table 2 gives an overview of the milestones that will be discussed in the
next sections.
- 19 -

Introduction
1.5.1 Systemic application of melphalan
Therapy with melphalan and prednisone in intermittent course has represented the
“golden standard” for newly diagnosed symptomatic multiple myeloma for many years
(32). Approximately 50% of treated patients responded, but the frequency of complete
remission was rather low (10%) (33). In addition the median remission time was only two
years and the median survival three years (31). A disease-free long term survival could
only be accomplished by allogeneic bone marrow transplantation.
- 20 -

Introduction
Comparative studies have been started to investigate higher response rates by
polychemotherapy, but none of the treatment modalities differed in the survival
parameters (31). The so-called VAD regimen (infusion of vincristin and doxorubicin and
oral dexamethasone) turned out to be an effective treatment schedule for melphalan
non-responders (34), but the overall survival was not enhanced compared to melphalanprednisone
treatment (35). Therefore, the transience of enhanced remission by
polychemotherapy and alternative VAD regimens led to an investigation of dose
intensification. In 1983 McElwain reported results using a melphalan dose of 140 mg/m²
for the treatment of patients with plasma cell leukaemia or multiple myeloma (36). The
first promising results led to further investigations with even more dose intensification up
to 200 mg/m² (37). But despite of higher response rates that could be reached by high
dose drug treatment, the overall survival rates did not change dramatically. More than
14% of patients died very early due to the high substance-related toxicity (38) and the
morbidity was also very high (39).
In view of the high systemic toxicity resulting from orally or intravenous application
of cytostatic drugs, a new strategy for treatment of melanoma patients was under
development since the 1950’s, using a local regimen in form of an isolated limb
perfusion (ILP). Cancer treatment with ILP enabled high response rates, enhanced
survival rates and low systemic toxicity due to the local application of nitrogen mustards.
In the following section this idea of cancer treatment will be reviewed in more detail.
1.5.2 Isolated limb and liver perfusion
The most common application of isolated perfusion has been isolated limb
perfusion for intransit-melanoma and as a limb-salvage technique for non-resectable or
locally recurrent extremity sarcoma (40, 41). The perfusion circuit allows delivering
hyperthermia, biological agents and/or chemotherapy via a closed recirculation usually
for one or two hours (42). At the end of the treatment, the perfused organ is flushed with
saline or colloid solution to remove any remaining traces of the therapeutics, before the
vascular continuity is restored. This technical procedure is more or less common to all
settings of isolated perfused organs. Especially for isolated liver perfusion, two different
approaches have been published in literature, the isolated hepatic perfusion (IHP) and
the hepatic artery infusion (HAI).
Starting at the late 1950’s Ausman (43) and Healey (44) studied a new technique
for an isolated hepatic perfusion in animals. The initial clinical study performed by
- 21 -

Introduction
Ausman in 1961 was limited to five patients and the treatment was restricted to one hour
(43). But against all hopes, high morbidity and mortality due to this technically
demanding procedure was observed and only a weak acceptance of this technique was
gained. Today their work can nevertheless be considered as a milestone in clinical
cancer therapy.
The technique was intended, however, to improve the efficacy of high-dose
chemotherapy within the tumor site, whilst keeping systemic drug levels low to reduce
systemic toxicity. Subsequently a number of articles were published, reporting different
strategies for isolated hepatic perfusion (IHP), concerning the mode of perfusion
(transarterial vs. intravenously). The idea of hepatic artery infusion (HAI) resulted from
the observation that liver metastases with a diameter smaller than 5 mm are nourished
mainly by the hepatic artery (45). In addition to that some drugs are more effectively
extracted by the liver during their first pass through its arterial circulation and many
drugs have a steep dose-response curve, so that large dose given regionally will lead to
a better response (46). IHP has several theoretical advantages over HAI. First, the
concentration of the perfused drug can be higher (only the organ toxicity is limiting).
Second, the drug must not necessarily have a high hepatic extraction rate on the first
pass. Third, conditions acting synergistically with chemotherapy might be added, like
hypoxia or hyperthermia. Last but not least additional biotherapy using TNF and IFNγ
might be more efficient and safe (46-49). Studies performed in rats by Marinelli et al.
favour this rationale by the finding that administration of mustards in isolated rat liver
produced higher tumor and tissue concentrations as compared to HAI (50, 51).
In 1983 and 1986 Skibba and co-workers published their results using a four hour
IHP under hyperthermic conditions without application of alkylating agents (52, 53).
Unfortunately, within those publications no classical response criteria were mentioned.
In 1994 Hafström and colleagues published data from their two clinical trials. In the first
paper, 29 patients suffering from liver metastases originating from various primary
lesions (breast cancer, colorectal cancer, midgut carcinoids and miscellaneous
primaries) were treated with cisplatin and melphalan under hyperthermic conditions. The
authors stated that four patients died within 30 days of multiple organ failure and partial
tumor regression was only observed in 20% of the patients. All surviving participants
developed reversal liver and renal toxicity (54). In their second paper a clinical study
including 32 patients with cancer confined to the liver was referred. Patients were
treated with 5-fluorouracil in a local regimen. Patients undergoing IHP displayed a
median survival of 17 months, while two patients died due to toxicity from the wrong
dose of 5-FU and the wrong route of administration (55). In 1996, Marinelli and coworkers
reported an overall response rate of 25% and a median survival of 17 months
- 22 -

Introduction
using mitomycin c (56). In the same year Leff et al. reported a study of 20 patients with
metastatic colon carcinoma treated with 180 mg/m² melphalan intravenously and
subsequent autologous bone marrow transplantation. This study reached an overall
response rate of 45% and a median survival of 198 days (57). Two years later, Aigner
and colleagues reported first results of IHP using 300 mg and 450 mg 5-fluorouracil. In
both patient groups several metastases had disappeared and the remaining tumor tissue
showed extended necrotic areas.
In 1998, De Vries and colleagues published a paper stating results of pre-clinical
use of melphalan combined with exogenous TNF in isolated livers of guinea pigs and an
additional phase I clinical study including nine patients. A partial response rate was
reported for five patients, one patient developed stable disease and three patients died
due to operation. The design of the clinical protocol included the cannulation of the
artery and the portal vein for IHP with melphalan and TNF being used simultaneously for
cancer treatment (58). Thought this trial was very successful on the one hand showing
an overall response rate of nearly 100 %, new problems became obvious on the other
hand. The combination of melphalan and recombinant TNF for therapy was often
descibed to be associated with higher liver toxicity, fatal coagulative disturbances and
severe systemic toxicity based on operative leakage of the perfusate.
The problem of capillary leak during isolated hepatic perfusion was also discussed
in a paper by Alexander et al. published in 1998. The authors presented a clinical study
evaluating TNF as a potential mediator of capillary leak occurring during IHP. In that
study 27 patients underwent a 60 min hyperthermic IHP with 1,5 mg melphalan, either
combined with (n=7) or without (n=20) 1.0 mg of recombinant TNF. The study revealed
that no significance could be calculated between both groups concerning leakage of
radiolabelled albumin. Moreover TNF did not seem to be responsible for melphalan
uptake in tumor cells. Thus leading the authors to the conclusion that the cytokine TNF
has to be critically evaluated for clinical use in cancer treatment (59).
More recently, being inspired by the promising report published for combinatorial
melphalan and TNF in the isolated limb perfusion by Lienard et al. in 1992 (60), several
institutions have been investigated combinatorial TNF/nitrogen mustard-based therapy.
In their initial reports the authors observed a 90% complete response rate and an overall
response rate of 100% in isolated limb perfusion using a combinatorial therapy with
TNF, IFNγ and melphalan for patients with intransit melanoma or non-resectable
extremity sarcoma (60, 61). This was considerably better than what had been reported
previously by Alexander et al., using melphalan alone (62). The effects of these initial
reports could be largely confirmed by others (reviewed in (42, 46) and complete
- 23 -

Introduction
response rates of greater than 75% for in-transit melanoma and greater than 80% for
non-resectable extremity sarcoma were published (62-64).
To complete this overview, it has to be mentioned that several other articles were
published using mitomycin C, cisplatin and 5-fluorouracil in clinical protocols (54, 56,
65), but the results obtained by these cytostatic drugs added no further advantages to
the ones observed with melphalan.
Table 3 summarizes recent publications of clinical investigations using melphalan
and other chemotherapeutic drugs or combinatorial treatment regimens for cancer
therapy. The role of TNF, its antitumor capacity and its clinical use alone or in
combinatorial regimens will be discussed in the next chapter.
- 24 -

Introduction
- 25 -

Introduction
1.6 TNF – heaven and hell in cancer therapy
Tumor Necrosis Factor (TNF) was discovered by Carswell et al. in 1975. However,
the trace of this molecule goes back through history far further. According to the writings
of the Ebers Papyrus (c 1500 BC), the Egyptian physician Imhotep proposed ca. 2600
BC to induce a local infection caused by a poultice and a small incision, as the
recommended treatment for “swellings”, as tumors were described in those days. In
Italy, the priest Peregrine Laziosi spontaneously healed from a huge malignant tumor
after the tumor broke through the skin and became severely infected (reviewed in (66)).
In Germany around 1740, the clinician Dr. Schwenke noticed that some patients with
various malignancies could benefit from certain infectious disease progressions. The
American surgeon W.B. Coley observed a complete remission of an egg-sized sarcoma
after the surgical wound became infected with Streptococcus pyrogenes. Subsequently,
he infected ten more cancer patients with these bacteria, resulting either in strong and
finally fatal infection with a strong regression of malignancies, or in less to no
inflammatory reaction and without tumor regression in the patients. Because of this
unpredictability, Coley developed a treatment strategy using a bacterial “vaccine” of the
killed bacteria S. Pyrogenes and Serratia marcescens, which was called “Coley’s
toxins”. The success of these toxins was more or less limited to the tumor type
(reviewed in (66, 67)).
In contrast to clinical interest in the curative use of toxins, a wide range of microorganisms
and microbial products have been studied as antitumor agents, including
bacteria, fungi, plasmodia and trypanosomes. One of the most prominent bacteria was
the mycobacterium Bacillus Calmette Guérin (BCG), which was shown to induce severe
haemorrhagic necrosis of certain mouse tumors upon injection (68). The systemic
antitumor effect of BCG (-extract) injection was thought to be due to a general increase
in immunological reactivity, but the mechanism or the molecule of action was still
unknown. The discovery of TNF by Carswell et al. provided a clue as to how these
diverse reactions in response to microbial infections might be linked (68) (reviewed in
(69)).
After the initial publication a number of investigations turned to the study of TNF,
although the enormous toxic side-effects in the patients discredited its systemic use for
cancer patients. The intensive investigations on this molecule led to at least a complete
description of the biological factor and were closely linked to the identification of
bacterial structures responsible for the inflammatory response. Shear and colleagues
have been the first who succeeded in purifiing a microbial protein present in the bacterial
wall, which was termed bacterial polysaccharide, later on lipopolysaccharide (LPS) and
seemed to be very effective in inducing tumor necrosis (reviewed in (70)).
- 26 -

Introduction
TNF is initially synthesised as a pro-hormone containing an aminoterminal peptide
of varying length, depending on species. The propeptide segment of the molecule is
highly conserved, showing an 86% homology between mouse and human. TNF consists
of 157 amino acids with a molecular weight of 17 kD. In humans the molecule is nonglycosylated,
in contrast to the murine form. In its biologically active form it consists as a
non-covalently linked homotrimer. TNF can be found in an active membrane-bound
precursor form and in a soluble form after cleavage from the membrane through a matrix
metalloproteinase, the so-called TNF alpha converting enzyme (TACE) (71, 72). TNF
was found to be produced mainly by monocytes and macrophages (73) upon activation
by various stimuli, but also by neutrophils (74-77), basophils/mast cells (78, 79) and
natural killer cells (80, 81). Further more, some other cell types have also been found to
produce and secrete soluble TNF including astrocytes and glia cells in the nervous
system and synovial cells in rheumatiod arthritis (82).
Now, more than 30 years after its discovery, a whole spectrum of pleiotropic effects
of the molecule is known, including physiological and pathophysiological actions. After
the molecule has been cloned in the 1980’s independently by the groups of Fiers,
Pennica and Shirai (83-86), TNF was shown to be a key mediator in inflammation and
sepsis, immunological reactions, apoptosis, rheumatiod arthritis, liver regeneration and
so on. (for further literature see (69, 70, 86-90). Between 1985 and 1988 recombinant
TNF was made available to medical oncology and the hope for curing cancer raised.
Unfortunately, systemic application in advanced cancer patients showed a very low
maximally tolerated dose of 300 mg/m² (91) and a tumor response was hardly seen at
that dose. Therefore, systemic application was more or less abandoned in 1988.
1.7 A whole family of cytotoxic cytokines – the TNF superfamily
Beginning with TNF, several other factors were identified to be able to induce
death in multiple cell-types and organs in humans and other species. They all belong to
a huge family, the TNF superfamily. The most prominent among them are TNF-ß (92),
CD95L (FAS-L, Apo1L) (93) and TNF-related apoptosis-inducing ligand (TRAIL) (94).
Interestingly, also the receptors of these molecules are closely related and belong all to
a huge receptor family, the so-called TNF receptor superfamily. At present, more than
19 members of the TNF superfamily are known that signal through 29 receptors (95).
The identification of these cytokines and receptors was mainly driven by two
scientific fields of interest, tumor-therapy and apoptosis. A detailed description of those
- 27 -

Introduction
ligands and receptors would be beyond the scope of this introduction. Therefore, only a
few pairs of receptors and ligands will be highlighted which have been discussed
intensively for tumor treatment.
In 1968 Williams and Granger described a protein produced by lymphocytes
capable of inducing death of tumor cells (92). After purification and determination of the
amino acid sequence, the relationship to TNF was revealed (84, 96-100). The protein
was first named TNF-ß, in view of the large homology to TNF, later lymphotoxin α
(reviewed in (95)). By direct cloning strategies a number of further ligands closely related
to TNF were identified (93, 99, 101). One among them was CD95L (Fas-L, Apo1L). A
third ligand also being able to induce death of other cells was found independently by
two different groups. Wiley and colleagues published a new type II membrane protein of
281 and 291 aa which they named TRAIL and which in between has been shown to
induce cell death in a broad variety of transformed cell lines (94). One year later in 1996,
Pitti and co-workers published a protein of 281 amino acids found by expressed
sequence tag, which seems closely related to Apo-1L and therefore named Apo-2L
(102). By comparison of both proteins and their DNA-sequence it became clear that
TRAIL and Apo-2L are identical.
In general, all members of this TNF superfamily exert physiological and
pathophysiological actions. Bharat B. Aggarwal therefore named their signaling
pathways a double-edged sword (95). There is no doubt that the bright side of the TNF
superfamily is their anticancer potential. This and the problems associated with this
molecule have already been discussed.
In the next sections two other members of this family will be highlighted, namely
CD95 and TRAIL, which have been shown to be key players in physiology and
pathophysiology.
- 28 -

1.7.1 CD95, Fas and APO-1
Introduction
In 1989 two independent groups reported an unexpected induction of apoptosis by
an antibody generated against tumor antigens (101, 103). Three years later the
corresponding cytokine receptor could be cloned and was termed Apo-1 (104) by the
Krammer group and Fas by the Nagata group (93). Finally, in 1994, the natural ligand
CD95 was purified and cloned (105, 106).
CD95 (Apo-1, Fas) was found to be expressed in many cell types, including heart,
liver, lung and thymus (107). On the contrary, CD95L (Apo-1L, FasL) seems to be most
likely expressed in the immune system, especially produced by T-cells and natural killer
cells (105). In the following years, CD95/CD95L was found to play a major role in liver
apoptosis (108-111), but similar to the other members of the TNF superfamily also
physiologic functions of CD95 have been found, including removal of tumor cells (112),
transducing growth promoting signals in proliferating T-cells (113, 114) and fibroblasts
(115) and implication in liver regeneration after partial hepatectomy (116). Up-regulation
of CD95 has also been demonstrated in active viral hepatitis (117). Since CD95L was
found to be able to kill tumor cells, hopes were raised again for having a new anti-cancer
drug. But this sword also turned out to be double-edged. Recombinant human CD95
proved to be insufficient for general cancer therapy, because of severe liver toxicity
observed in mice (118). In addition, Xiao-Zhong Wang and co-workers found that CD95
and CD95-L were present in the majority of specimen collected from 50 HCC
(hepatocellular carcinoma) patients. CD95 and CD95L were detectable in large numbers
on carcinoma cells, indicating that these cells strongly express both (119). Similar
findings have also been made by the group of Krammer et al. before (120). The question
why these ligands are expressed raised a number of hypotheses. First, co-expression of
receptor and ligand might lead to a so called fratricide death of HCC cells, induced not
only by lymphocytes but also by their own ligands in an autocrine or paracrine manner.
Such a mechanism might be involved in drug-induced HC apoptosis during tumor
therapy. Evidence for this hypothesis is raised by the finding that bleomycin is able to
upregulate CD95/CD95L in HepG2 cells (121). Second, CD95L might be important
during the infiltration of tissue as well as during the dissemination into the liver (122), as
demonstrated in hepatic metastasis of colon cancer cells (123). Finally, a third possibility
is termed “tumor-counterattack”. This hypothesis implies a self-defence strategy of the
tumor (cells). By expressing FasL on their surface, tumor cells will be able to induce
apoptosis to antitumor immune cells instead of being killed by them and therefore
survive. This was hypothesized by Stand et al. underlying the observation that HepG2
cells, expressing CD95L after treatment with cytostatic drug, were able to kill CD95positive
Jurcat t-cells. Furthermore, high prevalence of expression of CD95L in diverse
- 29 -

Introduction
tumor types (including colon cancer, liver cancer, melanoma and lung cancer) suggests
that CD95L might be a general, perhaps an essential factor in the inhibition of anti-tumor
responses (124). Since this hypothesis was formed, a hot debate has been started (for
further literature see (125-127)]. Nevertheless, if this idea is true or wrong, CD95L or the
agonistic antiCD95 antibody failed for cancer therapy.
1.7.2 TRAIL
Based on the homology of its extracellular domain to CD95L (28% identical), TNF
(23% identical) and LTα (23% identical) (94) a new member of the TNF superfamily,
TRAIL, was isolated in 1995 by Wiley and co-workers. Similar to others, TRAIL was
found to be a type II transmembrane protein which is able to induce apoptosis in various
transformed cell lines after binding to its specific receptor (94). Interestingly this factor
did not seem to be toxic to normal cells in vitro. The identification of a number of
receptors binding TRAIL, including DR4 (TRAIL-R1) (128), DR5 (TRAIL-R2, TRICK2)
(128-130), DcR1 (TRID, TRAIL-R3) (128, 129, 131) and TRAIL-R4 (DcR2) (132, 133)
(for review see also (134)) was confusing. Whereas DR4 and DR5 contain an
intracellular death domain and are therefore able to induce apoptosis via DISC (death
inducing signaling complex)-formation and subsequent caspase activation, DcR1 and
DcR2 seem to function as decoy (Dc) receptors with no intracellular domains. Yet the
function of these two receptors is still under debate. Corresponding transcripts of these
receptors have been found in many human tissues, including foetal liver, adult testis, but
not in most cancer cell lines examined (133, 134) (for review see (135)). Interestingly, it
was found that TRAIL-Rs 1 and 2 seem to be able to induce two different types of cell
death, apoptosis after intracellular recruitment of FADD (Fas associated death domain)
followed by caspase-8 recruitment and activation (122, 136-138), or necrosis after
recruitment of RIP (receptor-interacting protein) (139). In vivo studies performed by
Walczak and co-workers revealed that neither murine TRAIL nor human TRAIL lead to
general or organ specific toxicity or alterations in mice. In contrast, the authors could
demonstrate that intravenous application of TRAIL led to a lowered tumor burden in
SCID mice and the survival of tumor-injected mice was prolonged (140). Since it is
known that TRAIL also induced apoptosis in human hepatocellular carcinoma and
cholangiocarcinoma cells (141, 142), Gores and Kaufmann reviewed the question
whether TRAIL is toxic to the liver from the view of hepatology and oncology (143).
In their summary they state that current data suggest TRAIL to be the most
promising cytokine for cancer therapy. But only less information is accumulated and
- 30 -

Introduction
reports stating that TRAIL might be toxic to normal hepatocytes should not be ignored
(144, 145). All in all, TRAIL needs much more attention before early clinical trials might
be started. As the authors argue, it is not known if pathophysiological perturbations or
the influence of (cytostatic) drugs might render hepatocytes susceptible to TRAIL (for
further discussion see also (145-155)). Therefore, more pre-clinical studies have to be
undertaken.
1.8 Combinatorial therapy
Metastatic or primary non-resectable cancers confined to the liver, mostly resulting
from metastatic colon carcinoma, still represent a significant clinical problem. Of 140.000
new patients diagnosed with colorectal cancer in the US each year, approximately 20%
to 30% will die of progressive metastatic disease confined to the liver (156-158). For
Europe the prognosis is not better. In the UK colorectal cancer is the second most
common cause of cancer death, with more than 40% of these patients destined to die of
the disease despite current clinical management (159). More than 70% of all patients
with colorectal cancer will develop liver metastasis during progression of the disease
and autopsy studies indicate that in at least half of these cases the liver is the sole
metastatic site (157). For those patients, a complete surgical resection has remained the
only curative treatment strategy, with a five year survival rate of 40%. Only 10% to 20%
of the patients are possible candidates for such surgery (46) and the vast majority of
colorectal metastases confined to the liver are considered to be non-resectable (160,
161). For the remaining 80%, prognosis is grim if the disease is confined to the liver: a
mean survival rate of ten months (162) and less than six months with concomitant
extrahepatic metastasis or regional recurrence (163, 164). Standard application of
chemotherapeutic drugs prolongs this mean survival up to ten to fourteen months (165).
Therefore, new methods, technologies or treatment strategies have been taken into
account. But also other approaches like arterial embolization, chemoembolization,
ethanol injection, radio therapy, cryosurgery and radiofrequency ablation did not result in
prolonged survival (46). The most promising idea was an isolated setting with increased
drug concentration at the tumor site using melphalan and TNF together.
- 31 -

Introduction
1.8.1 TNF – the good, the bad or the ugly in ILP
Initial results using TNF and melphalan in isolated limb perfusion (ILP) achieved a
complete tumor regression in 90% of the cases for patients with melanoma lesions (61).
These and other preliminary results in the field of ILP with melphalan and TNF led to
further clinical investigations on the effect of cytostatic drugs combined with cytokines
also in isolated hepatic perfusion. However, the first results were disappointing. In a
phase I clinical trial in 1994 using melphalan and cisplatin, tumor regression was
registered in only 20% of treated patients (54). Another clinical study in the same year
revealed also positive results for the use of TNF (166). In this article TNF was discussed
to be mainly responsible for the breakdown of tumor vascularity. Therefore, TNF was
also made responsible for the capillary leak which was often observed during IHP.
In contrast, Alexander and co-workers could demonstrate that this augmented
capillary leak during IHP occur via TNF-independent mechanisms (167). In this paper, a
study of 27 patients was presented showing no leakage of hepatic perfusate,
independent of TNF application. Furthermore, TNF did not affect melphalan tumor
concentrations after IHP, too. In 1997 Borel Rinkes et al. published a study performed in
pigs using TNF in the presence and absence of melphalan (168). In this study, which is
one of the few pre-clinical trials, they found that IHP is technically feasible without
systemic toxicity, mild transient hepatotoxicity, minimal systemic drug exposure and
minor transient disturbances of liver biochemistry and histology.
1.8.2 Investigation of hyperthermia in ILP
In 1998, a number of clinical reports were published using melphalan and TNF
concomitantly. De Vries et al. reported a partial response (PR) in five of six cases using
1 mg/kg melphalan and 0.4 mg TNF with simultaneous hyperthermia of >41°C (58).
However, also a mortality of 33% was stated. Hyperthermia in general has been shown
to induce apoptosis in human gastric carcinoma cell lines in a p53-dependent manner
(169). For Hela cells it could be shown that an intact p53 gene is essential for heatinduced
apoptosis of cancer cells (170) and studies on human colorectal carcinoma cell
lines revealed that p53-inactivated cells did not undergo apoptosis in response to
hyperthermia, whereas normal cancer cells died (171).
The use of hyperthermia in clinical protocols has been published by several
authors. Hafström et al. reported a PR in three of eleven cases with a mortality of 18%
- 32 -

Introduction
(172) and Oldhafer et al. reported one complete response and 26 PR with 0% mortality
using 60-140 mg melphalan and 200-300 µg of TNF (173). Last but not least, Alexander
et al. published a phase II trial with an overall response rate of 75% (1 CR, 26 PR) using
hyperthermia, 1.5 mg/kg melphalan and 1.0 mg of TNF (174).
On the contrary to these promising results Lindner et al. reported a clinical trial in
1999 in which eleven patients with non-resectable liver malignancies were treated with
0.5 mg/kg melphalan and 30-200 µg TNF. In this study patients with metastases
resulting from colon carcinoma displayed no response and only three patients with
metastases from malignant melanoma showed partial response. In conclusion the
authors stated that this regimen is a method with high toxicity (two patients died within
the postoperative month). But it is not possible to distinguish between toxicity resulting
from TNF and toxicity due to the operative procedure (65). On the other hand,
melphalan/TNF was still successful in isolated limb perfusion as published by Plaat et al.
1999 for the treatment of soft tissue sarcoma (175).
Finally, in 2001 Alexander et al. published a study including 51 patients with nonresectable
malignancies confined to the liver. Thirty-two of them were treated with
melphalan/TNF, nineteen with melphalan alone. Despite one perioperative death (2%)
patients displayed an overall objective radiographic response rate of 76% (38 of 50
assessable patients) with a median duration of 10.5 months (range, 2 to 21 months).
The median survival was 16 months, respectively. The authors could observe 18 PRs in
26 patients (69%) whose prior therapy had failed.
To sum this efforts up Weinreich and Alexander wrote in their review on
transarterial perfusion of liver metastases in 2002 (42) that it is possible to perform safe
IHP in patients with non-resectable cancers confined to the liver, with a low morbidity
and mortality. It is possible to mediate clinically meaningful regression of refractory or
advanced cancers, but they also remark that further research is needed and that the role
of TNF during IHP still remains unclear.
- 33 -

Introduction
1.8.3 Pre-clinical results of melphalan in cancer therapy
Since melphalan was first used in clinical trials beginning in the 1940s and most
progress has been made in clinical trials using various nitrogen mustards and cytokines
alone or in combination, very few data concerning pre-clinical in vitro studies on cells or
in vivo studies in animals are available. Among present publications, most interest has
been focused on the role of glutathione. In 1991, Skapek and co-workers observed an
enhanced melphalan-mediated toxicity in nude mice after buthionine sulfoximine (BSO)
pre-treatment (176). BSO is a potent inhibitor of glutathione γ-glutamylcysteine
synthetase, resulting in intrahepatic GSH depletion (88%). The toxicity included
reversible gastrointestinal toxicity, myelosuppression and histologic evidence for
nephrotoxicity, but no evidence for hepatic failure. In 1992, Laskowitz et al. reported
enhanced melphalan-induced gastrointestinal toxicity after regional hyperthermia and
BSO-mediated GSH depletion (177). In 1999, Smith and co-workers investigated the
effect of BSO on melphalan-induced toxicity in mice. The authors found that a low dose
intravenously BSO administration did not alter melphalan toxicity (5mg/kg used),
whereas a high dose application lead to potentiated myelotoxicity, bone marrow toxicity
and nephrotoxicity (178). Interestingly, no liver toxicity was assessed, although Kramer
et al. demonstrated enhanced liver toxicity with BSO and alkylating agents in Fischer
rats before (179). In 1999 Vahrmeijer and colleagues investigated the effect of GSH
depletion on inhibition of cell cycle progression and the induction of apoptosis by
melphalan in human colorectal cancer cells (180). In this manuscript, the scientists
reported that GSH depletion below 20% of normal cellular GSH level enhanced
melphalan-induced cytotoxicity nearly 3-fold and DNA fragmentation was 4-fold higher
than compared to the non-GSH depleted state, indicating enhanced apoptosis. Further
analysis revealed that mitochondrial membrane potential was also enhanced both in
melphalan and melphalan/BSO treatment.
In summary, numerous cellular studies had been undertaken to reveal GSX
dependent changes in melphalan-induced cytotoxicity resulting in enhanced cellular
toxicity. This might be due to the fact that melphalan is able to react spontaneously with
GSH independent of the activity or presence of glutathione-S-transferase (181). Studies
performed in man and rat by Vahrmeijer and colleagues confirmed these results
indicating that hepatic GSH conjugation plays a very minor (if any) role in the elimination
of melphalan (182). It might be possible, especially in low dose therapy that the more
GSH is present in the cell, the more alkylating mustard is eliminated. On the contrary,
when GSX is depleted more molecules of alkylating agents might be able to crosslink
cellular DNA and subsequently toxicity is enhanced.
- 34 -

1.9 Apoptosis
Introduction
Today’s knowledge in the field of apoptosis mainly originates from research
activities in the second half of the 20 th century, although the German pathologists
Virchow (183) and Weigert have been interested in cell death nearly one century before.
They termed cell death “necrosis” or “coagulative necrosis” (Weigert 1877), derived from
the Greek expression for death.
Beginning with the second half of the 20 th century, developmental biologists
recognised that there must be different forms of cell death (184), in view of
morphological changes and the shape of death. Additionally a defined pattern seems to
regulate these forms of cell death, since only defined cell populations have been
destined to die. Investigation of the development of the worm Caenorhabditis elegans
revealed that exactly 131 of the 1,090 cells die during development (185, 186). Because
this cell death seems to be part of a developmental plan or program, the term
“programmed cell death” was used for this shape of death. Finally Kerr, Willy and Currie
defined this kind of death as “apoptosis”, based on the morphologic characteristics of
this type of death (187). It could be clearly distinguished from necrosis which represents
a pathological state of a cell (organ), including loss of the integrity of the plasma
membrane, disruption of the membrane potential and a spill-out of cellular contents in
the environment (188, 189) finally resulting in inflammation.
Apoptosis is evolutionary conserved and can be found in many organisms ranging
from plants and insects to birds and mammals (for review see (190). Apoptosis is
necessary for foetal and embryonal development, responsible for a proper function of
the immune system and for tissue homeostasis. But there is also recent evidence that
dysregulation of apoptosis is involved in a number of human diseases, including AIDS
(acquired immunodeficiency syndrom), viral infections, neurodegenerative disorders
(Alzheimer’s disease, Parkinson’s disease), autoimmune diseases, ischemic disorders
(myocardial infarction, stroke, reperfusions injury), toxin-induced liver diseases and
cancer (191). In view of the huge information material according to the knowledge about
induction and regulation of apoptosis, the biochemical processes and the involvement in
physiological and pathophysiological processes, it is not possible to discuss this topic in
depth here, because this would be beyond the scope of this introduction. Many of these
topics have been covered by a number of excellent reviews, which I would like to refer to
(189, 192-210). Only the role of apoptosis in cancer, the differences of apoptosis in
primary versus transformed cells and signaling of TNF receptors in apoptosis will be
discussed very shortly because of its importance for the current topic.
- 35 -

Introduction
1.9.1 Signaling of TNF receptors in apoptosis and anti-apoptosis
Since TNF was introduced as a molecule with a whole plethora of actions, with
great importance in physiology and pathology, it is also important for this thesis to give a
short overview about the signaling of TNF and TNF receptors, in order to understand
how these pleiotropic effects are transmitted to the cell.
TNF mediates its cellular effects mainly by two distinct surface receptors, namely
the 55 kD TNFR1 and the 75 kD TNFR2 (211, 212). Interaction between TNF and
TNFR1 activates numerous signal transduction pathways in multiple subcellular
compartments including the plasma membrane, cytosol, endosomes, mitochondria and
the nucleus (211). Among these diverse cellular effects of this pleiotropic cytokine, two
effects have gained the most attention: the initiation of cell death and the activation of
transcription factors, leading to the expression of genes, some of which promote cellular
survival, some of which are involved in cell growth, development, oncogenesis and
some of which are important in immune, inflammatory and stress responses. The
balance of these pathways at least determines the cellular fate.
To date, most of the players in the TNF pathway have been validated. However,
many questions remain poorly understood. For example, TNF is known to induce
necrotic and apoptotic cell death (213). Which signaling molecules or cellular conditions
decide which of these morphologically separated forms of cell death are induced?
Furthermore, the crosstalk between apoptosis and the pathways leading to cellular
proliferation including NFκB and JNK signaling pathways is still not clear and has to be
further investigated.
- 36 -

1.9.2 The role of TNF receptor 1
Introduction
A common feature of each pathway is the TNF-induced formation of an intracellular
signaling complex at the cell membrane, in case of cell death this is called deathinducing
signaling complex (DISC). The demise of a cell can occur by two distinct
mechanisms, apoptosis and necrosis and TNF can elicit both of them. In most cell types,
induction of apoptosis by TNF cannot occur without inhibiting the expression of new
genes (214) indicating that TNF might also induce cellular signals independent of cell
death via binding to their specific receptors. Indeed, TNF has been shown to activate
survival pathways inhibiting activation of death pathways (215). The pathway leading to
the activation of caspases (216) and finally to apoptotic cell death is primarily mediated
by TNFR1 (217), because the cytoplasmatic domain of TNFR1 contains a region called
death domain which has been shown to be essential for TNF-induced cytotoxicity (218).
In the basal state without bound TNF the death domain (DD) is associated to a silencer
(SODD) which prevents signal transduction (219). Upon binding of trimeric TNF to the
extracellular domain of TNFR1 the inhibitory protein SODD is released from the
intracellular death domain (219, 220) and aggregation of TNFR1 death domains is
enabled. The aggregated receptor is recognised by the adapter protein TNF receptorassociated
death domain (TRADD) which exhibits high affinity for the aggregated death
domains (221, 222) and forms a stable complex with the intracellular domains. The
death domain of TRADD further interacts with the DD of FADD (Fas-associated death
domain), forming also a stable complex (223, 224). To stabilise this huge protein
complex, further proteins are required (222). FADD contains two major domains, the DD
and a N-terminal domain termed death effector domain (DED) (223, 224). The
interaction of TRADD and FADD leads to an altered conformation of the protein with an
exposition of the death effector domain (225) which has an important function. It is
analogous to a so-called caspase- activating domain (CARD) found in the prodomain of
various upstream caspases including procaspase-8 (226). Upon recruitment to the DISC
procaspase-8 becomes activated autoproteolytically (227-229). Next to procaspase-8
procaspase-10 is also be recruited and activated at the DISC (230). Activated caspase-8
on the one hand is able to activate the effector proteases caspase-3, -6 and -7, which in
turn cleave multiple cellular proteins, resulting in cell death (230).
On the other hand, activation of effector caspases is amplified by a mitochondrial
pathway (231). A slight caspase-8 activity is sufficient to activate Bid, a Bcl-2 family
member which then translocates to the mitochondria,. Bax, another Bcl-2 family
member, is activated by this pathway to localise to the mitochondrial membrane and
increases its permeability resulting in mitochondrial dysfunction (reviewed in (199)) and
release of cytochrome c and other mitochondrial proteins (232-236) Cytochrome c binds
- 37 -

Introduction
to Apaf-1 and pro-caspase-9 connected via a CARD-domain and forms a protein
complex, the so-called apoptosome, of which dATP is an integral part and thus regulates
the apoptotic signaling via subsequent autocatalytic activation of pro-caspase-9 (231,
237, 238). Active caspase-9 in turn is able to activate further downstream caspases
(199, 239, 240). In addition, another factor called apoptosis inducing factor (AIF) is also
released from the mitochondria and translocates to the nucleus to effect degradation of
the DNA into large fragments (241).
1.9.3 The role of TNF receptor 2
TNF binding to TNF-receptor 1 has been shown to be the major ligand-receptor
interaction involved in apoptotic signal transduction. In contrary to TNFR1, TNFR2 does
not posses an intracellular death domain and hence, is unable to recruit death adapter
molecules like TRADD or FADD. Under certain circumstances, however, TNFR2 may
also be involved in inducing apoptosis (242-245). For example, TNFR2 seems to be the
major player in apoptosis of mature CD8 + T-cells (246) and antagonistic antibodies
against TNFR2 are able to inhibit, at least partially, TNF-induced cytotoxicity of tumor
cells (247). Evidence for involvement of TNFR2 in liver injury comes from studies of Con
A-induced liver toxicity in mice, since mice lacking TNFR2 are fully protected from Con
A-induced liver failure (248, 249). This suggests that both receptors upon stimulation
might contribute to cellular reactions leading to cell death. However, the use of
antagonistic antibodies in high concentrations and molecular receptor analysis revealed
that only TNFR1 is responsible for cytotoxic activity (242, 250-254). How can this be
explained? Two roles for TNFR2 in apoptotic signaling have been postulated. First,
TNFR2 might increase the local concentration of TNF so as to enhance death signaling
from TNFR1. Evidence for this hypothesis comes from the fact that the dissociation rate
of TNF from TNFR2 (t1/2= 10 min) is significantly greater than that for TNFR1 (t1/2> 3 h)
whereas TNFR2 has a higher affinity to TNF than TNFR1 (Kd 100pM vs. 500pM) (255).
Therefore, TNFR2 preferentially binds TNF at low concentrations (256). Hence, TNFR2
can perform a ligand passing function and binds many molecules of TNF, which are
rapidly passed to TNFR1 (255, 257, 258). Whether this passing is simply a result of
TNFR2-dependent increase in local ligand concentration or whether the mechanism
requires formation of transient heterocomplex is not clear yet. Second, activation of
TNFR2 might induce ubiquitinylation and subsequent degradation of TRAF2, which
inhibits the formation of the DISC by TNFR1 and which activates anti-apoptotic
processes induced by TNF (259, Duckett, 1997 #1280, 260). Evidence for this fact
comes also from observations of Fotin-Mleczek et al. demonstrating that TNFR2 induces
- 38 -

Introduction
depletion of TRAF2 and inhibitor of apoptosis proteins (IAPs) and therefore accelerates
TNFR1-dependent activation of caspase-8 (261). To get a better understanding of these
mechanisms induced by TNF, the signaling of TNF receptors will be shortly discussed in
the next chapter.
1.9.4 Anti-apoptotic signaling
The activation of transcription factors is a major cellular effect induced by TNF
(reviewed in (211)). In general intracellular pathways specific for TNFR1 or specific for
TNFR2 have been described, as well as signaling pathways shared by both receptors.
NFκB activation is associated with TNF-induced cellular protection. Both TNF
receptors are able to activate NFκB independently (221, 262, 263). For TNFR1 the
protein RIP is essential as scaffold for TRAF1 and TRAF2 heterodimers on the contrary
to TNFR2, which can directly associate with TRAF1-TRAF2 heterodimers (262). TRAF2
has been shown to interact with NFκB-inducing kinase (NIK) (264-266) which is part of
the high molecular weight complex IκB-kinase (IKK also referred to as signalosome)
(266-270). Activation of NFκB relies on phosphorylation-dependent ubiquitinylation and
degradation of inhibitor of κB (IκB) proteins, which normally retain NFκB within the
cytoplasm (271, 272). The phosphorylation of IκB is mediated by IκB-kinase (273). IκB is
phosphorylated and subsequently degrades from NFκB. NFκB translocates to the
nucleus and activates transcription. Alternatively, NFκB can be activated specific via
TNFR1 upon recruitment of IRAK (interleukin-1 receptor-associated kinase) and
subsequent formation of a complex between IRAK and NIK (274-276).
c-Jun NH2-terminal kinase (JNK) activation is also associated with TRAF2 and
both TNF receptors have been shown to be initiate JNK activation (262, 277). Briefly,
TRAF2 interacts with three proteins; TRAF-associated NF-kB activator (TANK) activates
GCK(R) which in turn activates mitogen-activated protein kinase/extracellular signalregulated
kinase kinase kinase 1 (MEKK1). MEKK1 has been shown to be a potent
activator of the JNK pathway (277-279). Active JNK can phosphorylate and activate
various transcription factors including c-Jun, ATF-2, Ets-like protein 1 (Elk-1) and cAMPresponsive
element binding protein (CREB) (277, 279, 280). Alternatively, JNK can also
be activated specifically by TNFR1 involving the activation of phosphatidylinositol 3
(PI3)-kinase via the downstream mediator Rac and cytosolic phospholipase A2.
Additionally, there is also evidence that TRAF2 is initiating the recruitment of cellular
inhibitor of apoptosis proteins (cIAP) (281). IAPs are capable to inhibit caspase activity
- 39 -

Introduction
directly (282-284). These inhibitory proteins contain two domains, baculoviral inhibitory
repeat (BIR)-1 and BIR-2, which are able to block activity of the effector caspases-3, -6
and -7, whereas BIR-3, a third domain of cIAP, selectively inhibits caspase-9 activity
(284).
A third pathway onset by both TNF receptors is p38 kinase activation. Following
TRAF-2 recruitment apoptosis signal-regulating kinase 1 (ASK1) is known to bind TRAF-
2 (285). ASK1 itself can directly activate three MAPKKs (MKK2, -3, -6) (286-288).
Substrates for p38 kinase include cytosolic phospholipase A2 and activating transcription
factor 2 (ATF2), both substrates assists in activator protein 1 (AP-1) transcription factor
formation (289, 290).
To complete this discussion, three more pathways have been shown to be
activated by TNFR1: the ERK kinase activating pathways, neutral sphingomyelinase
pathway and acid sphingomyelinase pathway (reviewed in (211)). The last one links
TNF to lysosomal protease cathepsin D, thereby contributing to TNF-induced cell death
(291). This is of greater interest since this family of proteases has been discussed to be
involved in caspase-independent apoptosis, a phenomenon that attracts more and more
attention.
1.9.5 Modulation of apoptosis by changing cellular redox and energy state
Glutathione
The tripeptide glutathione (GSH, γ-glutamyl-cysteinyl-glycine) represents the most
abundant non-protein thiol in the cell, serving as the major antioxidant and providing
defence against xenobiotics as a phase II conjugation substrate (292-294). Glutathione
affects numerous central cellular functions, like DNA and protein synthesis, gene
transcription, transport, catalysis, cell growth and apoptosis (203, 295, 296).
Interestingly, modulation of intracellular hepatic glutathione has been shown to
protect mice from death receptor-induced liver injury mediated by CD95 (297). In this
study the underlying mechanism for this apoptosis protection by glutathione depletion
was investigated and caspase-3 activity was identified as the major site being
dependent on the presence of GSH, while GSSG attenuated the proteolytic activity
(297). Furthermore, it could be demonstrated that not only apoptosis but also necrosis
occurred following death receptor activation by TNF in the D-galactosamine (GalN) /TNF
- 40 -

Introduction
and GalN/LPS-model, as well as in LPS-shock and ConA-induced liver damage. Cell
death induced by acetaminophen was enhanced under glutathione-depleting conditions
(298), which might be due to generation of the toxic metabolite N-acetyl-pbenzoquinoneimine,
leading itself to depletion of intracellular glutathione, alteration of
redox potential and ultimately, cellular necrosis (299).
All these studies have been undertaken in healthy animals and the in vivo effects
could not be demonstrated in vitro in isolated primary hepatocytes, therefore implications
for cancer have been not investigated. For transformed cells it could be demonstrated
that glutathione depletion mediated by IPTG or L-buthionine-(300)-sulfoximine (BSO), a
specific inhibitor of GSH synthetase) induced severe apoptosis of Ha-ras-transformed
NIH3T3 cells, mostly mediated through generation of reactive oxygen species (301)
whereas for immortalised human keratinocytes a protection from UVA-induced apoptosis
could be shown by BSO-mediated intracellular glutathione depletion (302). For
cholangiocytes Celli et al. could demonstrate that glutathione depletion by BSO is
associated with decreased (87%) levels of anti-apoptotic protein Bcl-2 and subsequently
enhanced apoptosis (303). Last but not least, arachidonic acid-induced apoptosis was
mostly mediated by generation of reactive oxygen intermediates and subsequent lipid
peroxidation (304).
All in all, glutathione depletion could be shown to protect mice from death receptormediated
apoptosis whereas less information is available how glutathione depletion
affects apoptosis in cell lines or cancer cells. The few data available mostly demonstrate
enhanced apoptosis which is mostly due to enhanced radical formation subsequently
inducing cell death.
ATP
Next to the redox state of the cell, apoptosis has also been shown to be influenced
by cellular ATP levels. Investigations on death receptor-mediated apoptosis under ATPdepleting
conditions revealed that reduced TNF-mediated apoptosis in contrast to
enhanced CD95-mediated apoptosis (305). These effects could also be demonstrated in
animal experiments (306) while the important ATP-dependent step in the apoptotic
cascade is still not known. Infusions of D-fructose as a source of energy supply during
parenteral nutrition have been used for many years and especially in patients with
chronic or acute liver diseases fructose has been claimed to have several advantages
over glucose (307, 308). The beneficial action of fructose had been assumed from the
- 41 -

Introduction
biochemical findings that this hexose is metabolised more rapidly than glucose (309).
Fructose metabolism has been shown to be independent of insulin (307) and in liver
diseases uptake of fructose by the liver is less affected than uptake of glucose.
Interestingly, since the ATP-depleting capacity is closely related to the presence and
activity of the rate-limiting enzyme fructokinase, investigations of Adelman, Morris and
Weinhouse on the activity of fructokinase, triokinase and aldolases in liver tumors of the
rat revealed that in these tumors hardly no activity of fructokinase could be detected
(310). Indeed, most cell lines failed for ATP depletion by carbohydrates. On the other
hand, isolated primary cells incubated with ATP-depleting sugars as fructose or tagatose
and isolated livers of rat (309) or man (311) perfused with fructose can be ATP-depleted
without toxic effects on the cells or the isolated organ. This discrepancy between primary
versus transformed tumor cells is of further interest and might be useful also in
chemotherapy. But currently no data on ATP depletion in clinical investigations of
cancers confined to the liver are present. Despite this alteration in primary versus
transformed hepatocytes a number of differences have been found in various cell types.
These alterations will be shortly discussed in the following chapter.
1.9.6 Differences in apoptosis of primary versus transformed cells
As mentioned above, apoptosis is a generally controlled mechanism of cell death
which is common to various cell types and organs and is evolutionary conserved
throughout multicellular organisms. Tumor cells or immortalised cell lines instead display
a number of differences in this control panel, leading to an altered response to death
stimuli, including sensitivity to death signals and altered apoptotic pathways. Especially
the death-inducing members of the TNF receptor superfamily have been studied
intensively leading to the elucidation of their role in tumor evasion from the immune
system (reviewed in (193)).
One of the most prominent members of this family is TRAIL, which kills tumor cells
very effectively but has no effect on healthy primary cells (94). Since at least five
different receptors for TRAIL have been identified, it has been suggested that the
differential expression of non-death receptors for TRAIL (TRAIL-R3) might be
responsible for the different sensitivity to TRAIL observed for normal and transformed
cells (129, 312). On the contrary, Leverkus et al. could demonstrate that TRAIL is able to
induce apoptosis not only in transformed keratinocytes but also in primary keratinocytes
in a dose-dependent manner, displaying a marked shift in sensitivity (313). This shift
could be explained by intracellular modulation of adapter proteins and seems not to be
- 42 -

Introduction
related to the expression profile of TRAIL-receptors on the surface of the cell. Especially
FADD and TRADD have been shown to modulate intracellular apoptotic signaling since
expression of dominant negative forms of those proteins abrogate TRAIL-mediated
apoptosis (130, 228). Further more, cFLIPL, a protein known to interfere with the
proximal death receptor-mediated cascade, has been shown to suppress CD95-, TNFand
TRAIL-mediated apoptosis (314-316) and was shown to be strongly expressed in
primary keratinocytes, while it was undetectable in transformed cells (313). These
findings might also be in line with previous publications on FADD since FLIP might
compete with FADD for binding to pro-caspase-8 (317).
Another publication identified the PI-3-kinase/AKT-pathway responsible for the
major differences between primary and transformed keratinocytes in UV-induced
apoptosis (123). On the DNA-level changes were most frequently observed in the tumorsuppressor
p53-gene. Studies on apoptotic signaling in normal and neoplastic germinal
center revealed that a number of intracellular proteins modelling apoptosis are
dysregulated including Bcl-2 upregulation via chromosomal translocation of the bcl-2
gene (reviewed in (318)). API2/MLT and Bcl-10 were found to be enhanced via
chromosomal translocation in mucosa-associated lymphoid tissue (MALT) lymphoma
resulting in a fusion protein of API 2 and MLT leading to NFκB-mediated inhibition of
apoptosis (319, 320). Finally, Bax was found to be downregulated because of mutations
leading to inactivation of the bax gene resulting in the loss of pro-apoptotic Bax-protein
identified in a Burkitt lymphoma being resistant to CD95-mediated apoptosis (321).
Independent of gene alterations, reduced levels of Bax were also found in several
lymphoma malignancies (322, 323).
On the receptor level mutations of CD95 were found to affect the intracellular death
domain resulting in loss of DISC-assembling (324) and subsequently inhibition of
docking and activation of pro-caspase-8 (325-328). Last but not least, downstream of
receptor signaling, caspase-10 was found to be a common site of mutations in about
15% of B-cell non Hodgkin’s lymphomas. Mutations in the death effector domain (DED)
of the protein led to a loss of function of this domain which is necessary for interaction
with FADD at the receptor site (329, 330).
In summary, mutations resulting in altered apoptotic machinery could be identified
at nearly all levels of intracellular apoptotic signaling making tumors less susceptible for
killing by extracellular triggering. On the other hand the knowledge of these differences
between primary and transformed cells represent new potential targets for molecular
therapy or drug development raising hopes to find the tumors Achilles’ heel.
- 43 -

1.9.7 Apoptosis in cancer
Introduction
Apoptosis plays a crucial role in maintenance of tissue homeostasis and serves as
a safeguard system to prevent cell transformation through elimination of aberrant cells
with dysregulated gene expression caused by mutation or amplification of DNA
damages resulting from various physical, chemical or biological noxa which exceed the
limit of repair (331). In this case, cells incurring an oncogenic mutation or integration of
oncoviral genomes might be eliminated through the onset of the apoptotic program,
either triggered by intrinsic mediators or by engagement of death receptors via their
specific ligands presented by immune cells (332). Especially in hepatic carcinogenesis
and HCV or HBV infections large clonal expansion of cytotoxic T cells is induced, which
are able to induce CD95-mediated death in transformed or infected hepatocytes.
Recent studies have shown that various anti-cancer drugs such as cisplatin,
etoposide and 5-fluorouracil (5-FU) are able to induce apoptosis in cancer cells
(reviewed in (333)), but a functional p53 gene is required (334, 335). Therefore,
mutations in p53 have been shown to be concerned in no less than 50-60% of cancers
in various types and therefore enabling cells to acquire resistance to apoptosis (331,
336). Another gene isolated in rodents is an intronless member of the myc family called
s-myc (337, 338). The resulting gene product has been show to induce apoptosis in a
p53-independent manner when transfected in rat and human glioma cells (339-341). If
the assertion that anti-cancer drugs are able to induce apoptosis in transformed cells is
correct and due to the fact that apoptosis is a highly regulated biochemical process
(reviewed in (196)) biochemical alterations might exist that make cells more or less
susceptible or resistant to apoptosis. This might contribute to cellular or molecular
resistance to a wide variety of drugs (342). Because of these potential implications for
mechanisms of resistance, there is considerable interest in determining whether
anticancer drug initiate the apoptotic process through death receptors or intrinsically via
the mitochondrial pathway.
One hypothesis which is currently discussed is the activation of death via CD95
(Fas, Apo-1) triggered by the expression of CD95L on immune cells by anticancer drugs
of various types including doxorubicin (DNA intercalating), etoposide (topoisomerase II
inhibitor), cisplatin (DNA-crosslinking) and bleomycin (DNA-fragmentation) (121, 343-
348). Nearly all of them have also been shown to upregulate CD95L mRNA (for review
see (342)) due to drug induced activation of the transcription factors AP-1 and NFkB
(101). Unfortunately, CD95L is not only upregulated in cytotoxic B and T cells, but also
in target cells (cancer cells) as discussed earlier (124, 349, 350). Responsible for the
expression of CD95L mRNA might not only be the activation of transcription factors, but
- 44 -

Introduction
also reactive oxygen intermediates which might contribute to CD95L upregulation as
suggested by Hug et al. (351). In addition to CD95, other members of the TNF
superfamily have been shown to be upregulated by anticancer drugs, indicating that
these proteins might also play a role in apoptosis induction during cancer development
(352).
But does this really hold true? In their review on induction of apoptosis by cancer
chemotherapy Kaufmann and Earnshaw discussed these facts very critical (342).
Following their argumentation drug concentrations used to demonstrate CD95L
upregulation in cancer cells are very high exceeding doses used in clinical approaches.
Only for one drug (5-FU) it was demonstrated that CD95L is upregulated by drug
concentrations having similar time courses and dose response curves for CD95L
induction and inhibition of clonogenic survival (353, 354). Additionally, manipulations
inhibiting CD95 and CD95L interactions have been shown to have opposite effects in
different laboratories demonstrating no crucial role for CD95 (for review see (342)). Last
but not least CD95-negative cells have been shown to react normally upon anti-cancer
drug treatment (332, 355-357).
Due to these conflicting results, it is of great importance to shed more light on the
role of apoptosis in cancer, especially since mechanisms of drug resistance are not well
understood, while displaying an enormous clinical problem. Finally, it is important to
keep in mind that apoptosis induction via anti-cancer agents is not restricted to
transformed cells, but also to normal healthy cells which should be prevented from
apoptosis induction. Any knowledge that leads to differences in death induction between
healthy and transformed cells might be extremely important and therefore can contribute
to further advances in tumor therapy.
- 45 -

2 Objectives of the thesis
Objectives
Liver metastases resulting from colon carcinoma or other types of cancer still represent
a major clinical problem. Surgery is often not possible due to numerous metastases
spread over the whole organ. Radiation results in massive necrosis leading to
inflammatory liver failure and chemotherapy, even when applied locally, is either
insufficient or leads to massive cell death also involving healthy tissue. The
understanding of mechanisms leading to liver-specific tissue destruction caused by
chemotherapeutic drugs, the role of cytokines and the interaction of different cell types
being part of this scenario can therefore provide new rationale for therapeutic
intervention strategies in cancer therapy. The cellular ATP content has been shown to
be a central metabolic parameter modifying cell death in different cell types. A possibility
for selective ATP depletion in hepatocytes is exposure to high concentrations of fructose
or other sugars that rapidly trap intracellular phosphate. This approach has the
advantage that sufficient residual ATP (>15% of control) remains in the depleted
hepatocytes, thereby avoiding induction of necrosis due to total ATP loss. However, this
metabolic ATP trapping has been shown to fail for most cancer cell lines providing an
interesting tool for cancer therapy. At present the role of ATP depletion in cancer therapy
has been poorly investigated and pre-clinical studies involving cell culture and animal
models are hardly available.
In the present study the mechanism of melphalan-induced hepatocyte toxicity has been
investigated in primary liver cells and in isolated liver perfusion. The effect of ATP
depletion in melphalan-induced liver cell-death was studied in liver cells and organs of
healthy mice and in livers isolated from tumor-bearing mice.
In particular, the following questions were addressed in the present thesis:
1. What is the mechanism of melphalan-induced hepatotoxicity in
isolated primary murine liver cells?
2. Can this cell death be blocked by modification of the apoptotic
machinery?
3. How does ATP depletion affect melphalan-induced hepatotoxicity?
4. Is there a difference in hepatic apoptosis induced by melphalan
alone or in presence of exogenous soluble TNF?
5. Is the isolated perfused mouse liver a sufficient model to reflect
chemotherapy-induced liver failure in the murine system?
6. How does ATP depletion act on isolated perfused livers of tumorbearing
mice treated with melphalan and exogenous TNF?
- 46 -

3 Materials and methods
Materials and animals
3.1 Chemicals
Materials & Methods
N-acetyl-Asp-Glu-Val-Asp-7-amino-4-trifluoromethylcoumarin (DEVD-afc) and
Pefabloc® were obtained from Biomol (Hamburg, Germany). The caspase inhibitor
benzoyloxycarbonyl-val-ala-asp-fluoro-methylketone (zVAD-fmk) was purchased from
Alexis Biochemicals (Lausanne, Switzerland), as well as the recombinant Apo-2L Killer
TRAIL. D-galactosamine (GalN) and Hepes was purchased from Roth (Karlsruhe,
Germany) and LPS (Salmonella abortus equi) from Metalon (Wusterhausen, Germany).
Melphalan was obtained from Glaxo Wellcome GmbH &Co (Bad Oldesloe, Germany)
and Percoll was obtained from Amersham Biosciences (Freiburg, Germany),
respectively. AlamarBlue was obtained from BioSource (Solingen, Germany).
Cytotoxicity detection kit (LDH) was used from Roche Diagnostics GmbH (Mannheim,
Germany). Concanavalin A (Con A), staphylococcal enterotoxin B (SEB), ketohexoses
such as fructose, tagatose, mannose, mannitol, actinomycin D, DMSO and all other
reagents and recombinant enzymes not further specified were obtained from Sigma
(Deisenhofen, Germany). Pentobarbital (Narcoren®) was purchased from Sanofi
Withrop (München, Germany).
3.2 Antibodies and recombinant proteins
Activating anti-CD95 antibody (Jo2) and polyclonal IgG-horseradish peroxidasecoupled
secondary antibody (goat anti-mouse) were purchased from PharMingen (San
Diego, CA, USA). Recombinant mouse TNF was obtained from Innogenetics (Ghent,
Belgium). Antibody pairs (specific rat anti-murine mAb) for cytokine determinations were
purchased from Pharmingen (San Diego, CA, USA). Atrial natriuretic factor (ANP) was
delivered by Calbiochem (Schwalbach/Ts, Germany). CD11b-coated magnetic
MicroBeads as well as FITC-labelled CD11b mAB were obtained from Miltenyi Biotec
(Bergisch Gladbach, Germany). Other recombinant enzymes not further specified were
purchased from Boehringer Mannheim (Mannheim, Germany) or Sigma (Deisenhofen,
Germany).
- 47 -

3.3 Cell culture materials
Materials & Methods
Cell culture plates (24 and 96 well), petri dishes and other plastic materials were
purchased from Greiner (Frickenhausen, Germany). Cell culture medium RPMI 1640
was purchased from BioWhittaker (Verviers, Belgium), DMEM HIGH GLUCOSE medium
was from PAA laboratories (Pasching, Austria) and collagen was obtained from Serva
(Heidelberg, Germany). Penicillin, streptomycin and FCS were bought from Gibco BRL
Life Technologies (Eggenstein, Germany).
3.4 Animals
Specific pathogen-free male BALB/c, C57Bl/6 wild type mice, TNFR1 k.o. mice,
TNFR2 k.o. mice and TNFR1R2 k.o. mice (approximately 25 g), originally provided by
Dr. Horst Bluethmann, Basel, were from the in-house breeding stock at the University of
Konstanz. LPR mice were from Harlan (Borchen, Germany). Inducible NFkB p65 k.o.
mice (MXcrep65 k.o. and MXcrep65 wt) were a gift from Prof. Roland Schmid, München.
Animals were held at 22°C and 55% humidity, given a constant day-night cycle of 12 h
and fed a standard laboratory chow (Altromin 1310, Lage, Germany). All steps of animal
handling were carried out according to the Guidelines of the National Institute of Health
(NIH), the European Council (directive 86/609/EEC) and the national German authorities
and followed the directives of the University of Konstanz Ethical Committee. Mice were
starved overnight before the onset of experiments, which generally commenced at 8
a.m.
- 48 -

Methods
3.5 Cell culture
3.5.1 Cell lines
HepG2 cells
Materials & Methods
HepG2 cells were maintained in RPMI 1640 medium containing 10% FCS. Cells
were grown in 175 cm² flasks in humidified atmosphere at 37°C, 5% CO2, 21% O2 and
74% N2. The day before the experiments were carried out, human hepatoma cells were
harvested with trypsin/EDTA, centrifuged (200 × g, 4 min), resuspended in medium
containing 10% FCS and plated in 24-well plates (10 5 cells/well).
Hepa1-6 cells
Hepa1-6 cells were maintained in DMEM HIGH GLUCOSE medium containing
10% FCS. Cells were grown in 175 cm² flasks in humidified atmosphere at 37°C, 5%
CO2, 21% O2 and 74% N2. The day before the experiments were carried out, murine
hepatoma cells were harvested with trypsin/EDTA, centrifuged (200 × g, 4 min),
resuspendet in medium containing 10% FCS and plated in 24-well plates (10 5 cells/well).
AML-12 cells
AML-12 cells were maintained in HAM’s F-12 medium containing 10% FCS. Cells
were grown in 175 cm² flasks in humidified atmosphere at 37°C, 5% CO2, 21% O2 and
74% N2. The day before the experiments were carried out, cells were harvested with
trypsin/EDTA, centrifuged (200 × g, 4 min), resuspended in medium containing 10%
FCS and plated in 24-well plates (10 5 cells/well).
- 49 -

RAW cells
Materials & Methods
RAW cells were maintained in RPMI 1640 medium containing 10% FCS. Cells
were grown in 175 cm² flasks in humidified atmosphere at 37°C, 5% CO2, 21% O2 and
74% N2. The day before the experiments were carried out, cells were harvested with
trypsin/EDTA, resuspended in medium containing 10% FCS and plated in 75 cm² flasks.
3.5.2 Isolation and culture of primary cells
Primary murine hepatocytes
Isolation of hepatocytes from 8 weeks old mice was performed by the two-step
collagenase perfusion method of Seglen (358) as modified by Klaunig (359, 360) and
Leist (361). For some experiments, cells were additionally purified by centrifugation
using a Percoll gradient modified from (362). To separate hepatocytes from remaining
non-parenchymal cells, the pellet of the second centrifugation step (50 × g, 2.5 min) was
resuspended in 16 ml Hanks´ Balanced Salt Solution (HBSS) and mixed with 32 ml of an
isotonic Percoll solution (27.8 ml of 100% Percoll, 4.2 ml of 10-fold concentrated HBSS)
followed by centrifugation at 800 × g for 5 min at room temperature. To remove
remaining Percoll, the pellet was washed with HBSS by an additional centrifugation step
(50 × g, 2.5 min). Hepatocytes were plated in 200 µl RPMI 1640 medium including 10%
FCS in collagen-coated 24-well plates at a density of 8×10 4 cells per well. Cells were
allowed to adhere to the plate for at least 4 hours before the medium was changed to
RPMI 1640 without FCS. Incubation of murine hepatocytes with melphalan started 30
min after medium exchange alone or in the presence of other mediators mentioned in
the text. For some experiments Kupffer cells were depleted in vivo by intravenous
injection of 150 µl liposome-encapsulated Clodronate one and two days before liver cell
preparation. Clodronate-liposomes were prepared and injected as described earlier
(363).Incubations were carried out at 37°C in an atmosphere of 40% O2, 5% CO2, 55%
N2 and 100% humidity.
- 50 -

Non-parenchymal liver cells
Materials & Methods
Non-parenchymal cells (NPCs) were plated in 100 µl RPMI 1640 / 10% FCS in 96
well plates at a number of 10 5 per well. For standard experiments cells (mainly Kupffer
cells) were allowed to adhere for 20 minutes before renewal of medium. Incubations
were performed on the following day. For some experiments Kupffer cells were
additionally purified as described by Doolittle et al. (364). After isolation cells were three
times washed with PBS and incubated with CD11b-coated MicroBeads ® (10µl beads per
10 7 total cells in a total volume of 100µl, 10 min, 4°C). After incubation, cells were
washed by adding 900µl buffer with subsequent centrifugation at 300xg for 10 min, 4°C.
After washing cells were resuspendet in buffer (10 8 total cells per 500 µl) and proceeded
to magnetic separation (MS column type). Column was prepared according to manual
instruction and cell suspension was allowed to pass magnetic field by gravity. Column
was washed three times with 500 µl buffer before positive cells have been flushed out in
a total volume of 1ml. After an additional washing step cells have been plated as usual
in final concentration of 10 5 per well.
For transwell experiments Kupffer cells were left to adhere for 24 hours on cell
culture inserts with 1 µm pore size (Becton Dickinson, USA) and were placed in wells
containing freshly isolated and adhered purified hepatocytes.
3.6 Isolated liver perfusion
3.6.1 Technical procedure of liver preperation and perfusion
Upon a lethal intravenous injection with 150 mg/kg pentobarbital-natrium and 0.8
mg/kg heparin, the vena portae and the vena cava inferior of the mouse liver were
cannulated and ligated. After cannulation the organ was perfused blood-free before
circulation has been closed, to guarantee a blood-free perfusion. Perfusion was carried
out submarine (to avoid oedema formation) using a modified Krebs-Henseleit buffer [147
mM NaCl; 5.36 mM KCl, 0.34 mM Na2HPO4; 0.44 mM KH2PO4; 0.77 mM MgSO4; 10
mM D-Glucose; 9 mM HEPES; pH: 7.3; 325-330 mOsm] with a total volume of 25 ml
buffer in a closed circulation mode under constant pressure conditions of 21.33 mm Hg.
Osmolality of the perfusate fluid was set to 225-330 mOsm and checked before each
individual experiment. The temperature of the perfusate was kept constant at 37°C,
guaranteed by a separate thermo-circulation and oxygenation with pure oxygen at a
pressure of 500 mbar was performed. During perfusion, the perfusate flow through the
- 51 -

Materials & Methods
liver as well as the pressure were constantly measured and recorded. Samples for
metabolite and enzyme measurements were taken from the perfusate at different time
points, as indicated in the text. All technical equipment and tubes for performing isolated
mouse liver perfusion was delivered by Hugo Sachs Electronik (March-Hugstetten,
Germany). For illustration see also figure 4.
- 52 -

3.6.2 Treatment schedules
Materials & Methods
After lethal anaesthesia mice livers were immediately prepared and perfused with
freshly oxygenated buffer to avoid ischemia. Total preparation time was below 5
minutes, otherwise the experiment was cancelled. After closing the recirculation, organs
were perfused for about 20 minutes to control leakage, perfusion parameters and quality
of preparation. As soon as perfusate flow has stabilised, liver injury-inducing compounds
were added and sampling of material started. Following injury-inducing models have
been used:
• Liver injury induced by anti-CD95: hepatic apoptosis mediated via CD95 was
induced by application of agonistic anti-CD95 antibodies in concentrations of 100
ng/ml given to the perfusate in a volume of 100 µl 0.1% HSA/saline at the
beginning of experiment (t=0).
• GalN/TNF-induced liver injury: GalN was given i.p. in doses of 500 mg/kg 30 min
before lethal anaesthesia. TNF was given to the perfusate at a final concentration
of 100 ng/ml in 100 µl 0.1% HSA/saline at the beginning of experiment (t=0).
• GalN/LPS-induced liver injury: GalN was given i.p. in doses of 500 mg/kg 30 min
before lethal anaesthesia. LPS was given in a final concentration of 1 µg/ml to the
perfusate in a total volume of 300 µl sterile and pyrogen-free saline at the
beginning of the experiment (t=0).
• Melphalan-induced liver injury: Melphalan was prepared and added immediately
to the perfusate in dose of 150 mg/kg at the beginning of the experiment (t=0)
• Further compounds: carbohydrates and phorone were injected i.p. in doses
indicated in the text/graph 30 min before lethal anaesthesia of animals. zVAD-fmk
was added at concentrations indicated directly to the perfusate in a total volume
of 100 µl 0.1% HSA/saline.
• For depletion of Kupffer cells mice were injected two times three days and one
day before the experiment with Chlodronate liposomes with a total volume of 300
µl of provided liposome suspension according to Dr. Nico van Rooijen ((363) and
personal communication) and Schümann et al. (365).
- 53 -

3.6.3 Sampling of material
Materials & Methods
At the time points indicated (usually in 30 min interval) 100 µl of perfusate samples
were taken, stored at 4°C and analysed immediately after the experiment (ALT and LDH
measurement). In case of samples for TNF-measurement, samples were taken in
shorter intervals as indicated and immediately frozen at -80°C. After the experiment
livers were perfused for 10 s with cold perfusion buffer (PB, 50 mM phosphate buffer pH
7.4, 120 mM NaCl, 10 mM EDTA), immediately excised and processed as follows:
• Slices of the large anterior lobe were frozen in liquid nitrogen and stored at -80°C
until the measurement of caspase-3,7-like activity.
• For liver histology, liver specimen were immediately cut into 1 mm thick slices and
fixed in phosphate-buffered neutral 4% formalin solution for light microscopy or
frozen immediately on dried ice and stored at -80°C for cryoslices.
3.7 Animal experiments
3.7.1 Treatment schedules
After treatment with liver injury-inducing compounds, animals were sacrificed by
lethal anaesthesia to obtain samples at different times as described. Alternatively, the
survival of animal was monitored over a period of at least three months.
• Liver injury induced by anti-CD95: hepatic apoptosis mediated via CD95 was
induced by application of agonistic anti-CD95 antibodies in doses of 2 µg/mouse
given i.v. in a volume of 300 µl 0.1% HSA/saline.
• GalN/TNF-induced liver injury: TNF was given i.v. in a dose of 2 µg/kg in 300 µl
0.1% HSA/saline and the aminosugar GalN (500 mg/kg, given in 300 µl saline,
i.p.) was administered 30 min before TNF to block hepatic transcription (8 hour
model).
• GalN/LPS-induced liver injury: after sensitisation with GalN as mentioned above,
LPS was administered i.p. in a volume of 300 µl sterile saline in a dose of 1.5
µg/kg (8 hour model).
- 54 -

Materials & Methods
• GalN/SEB-induced liver injury: after sensitisation with GalN as mentioned above,
SEB was administered i.p. in a volume of 300 µl sterile saline in a dose of 2
mg/kg (8 hour model).
• Con A-induced liver injury: T cell-dependent liver injury was induced by Con A
according to Tiegs et al. (366). Con A was injected i.v. into naive mice in a volume
of 300 µl pyrogen-free saline at a dose of 25 mg/kg (8 hour model).
• Further compounds: Melphalan was given in variable doses of 25, 50, 75 and 100
µg/kg at different time-point before or after challenge, in a volume of 300 µl
solvent.
3.7.2 Sampling of material
At the time-points indicated, mice were euthanized by i.v. injection of 150 mg/kg
pentobarbital plus 0.8 mg/kg heparin and blood samples were obtained:
• To assess the extent of liver damage, blood was withdrawn by cardiac puncture
and subsequently centrifuged (5 min, 14,000 g, 4°C). ALT enzyme activity was
measured in the plasma as described below.
Blood samples for cytokine determinations were obtained either from the tail
veins using heparinized syringes, or alternatively by cardiac puncture as
described above, subsequently centrifuged (5 min, 14,000 x g, 4°C) and stored at
-80°C.
After blood withdrawal, livers were perfused for 10 s with cold perfusion buffer (PB, 50
mM phosphate buffer pH 7.4, 120 mM NaCl, 10 mM EDTA), immediately excised and
processed as follows:
• Slices of the large anterior lobe were frozen in liquid nitrogen and stored at -80°C
until the measurement of caspase-3-like activity.
• For liver histology, liver specimen were immediately cut into 1 mm thick slices and
fixed in phosphate-buffered neutral 4% formalin solution for light microscopy or
frozen immediately on dried ice and stored at -80°C for cryoslices.
- 55 -

3.8 Tumor model
3.8.1 Tumor induction
Materials & Methods
12 to 15 days old C3H/HeN mice (Charles River, Sulzfeld, Germany) were injected
intraperitoneally with 10µg/kg N-Nitrosodiethylamine- (Diethylnitrosamine; DEN) solution
(Serva, Heidelberg, Germany). Animals were held at 22°C and 55% humidity under
specific pathogen-free conditions, given a constant day-night cycle of 12 h and fed a
standard laboratory chow (Altromin 1310, Lage, Germany). All steps of animal handling
were carried out according to the Guidelines of the National Institute of Health (NIH), the
European Council (directive 86/609/EEC) and the national German authorities and
followed the directives of the University of Tübingen Ethical Committee and University of
Konstanz Ethical Committee. Starting 20 weeks after tumor induction mice were killed by
a lethal intravenous injection with 150 mg/kg pentobarbital-natrium and 0.8 mg/kg
heparin and prepared for liver perfusion. Mice were starved overnight before the onset of
experiments, which generally commenced at 8 a.m.
3.8.2 Isolated liver perfusion
Liver perfusion was carried out as described above. Prior to liver isolation tumor
burden was documented by photographs. After cannulation of vena cava inferior and
vena portae and subsequent perfusion with buffer to remove hepatic blood, a second
picture was taken to document the perfusion of the tumors. Further perfusion procedure
was carried out as indicated above. After the experiment, livers were excised out of
mice, wet weight was measured and size of tumors and livers were documented by a
picture taken on paper ruled in millimetre squares. Subsequently liver and tumor
samples were taken and immediately frozen in liquid nitrogen for caspase-determination,
frozen on dried ice for cryo-slices, or stored in freshly prepared 4% paraformaldehydesolution
for histology.
3.8.3 Treatment schedules
After lethal anaesthesia mice livers were immediately prepared and perfused with
freshly oxygenated buffer as soon as possible to avoid ischemia. After closing the
recirculation, organs were perfused for about 20 minutes to control leakage, perfusion
- 56 -

Materials & Methods
parameters and quality of preparation. As soon as perfusate flow has stabilised, liver
injury-inducing compounds were added and sampling of material started. Following
treatment schedules have been investigated:
• TNF: TNF was given to the perfusate at a final concentration of 100 ng/ml in 100
µl 0.1% HSA/saline at the beginning of experiment (t=0).
• Melphalan: Melphalan was prepared and added immediately to the perfusate in
dose of 150 mg/kg at the beginning of the experiment (t=0)
• Melphalan/TNF: Melphalan was prepared and added immediately to the perfusate
in dose of 150 mg/kg at the beginning of the experiment (t=0), TNF was added to
the perfusate at a final concentration of 100 ng/ml in 100 µl 0.1% HSA/saline 30
min later
• Control: 100 µl 0.1% HSA/saline was added at the beginning of the experiment
(t=0)
• Further compounds: carbohydrates were injected i.p. in doses indicated in the
text/graph 30 min before lethal anaesthesia of animals.
3.9 Cytokine determination
Antibody pairs (specific rat anti-murine mAb) were purchased from Pharmingen
(San Diego, CA, USA). Streptavidin-peroxidase was obtained from Jackson Immuno
Research (West Grove, PA, USA) and the TMB liquid substrate system from Sigma
(Deisenhofen, Germany). TNF and IL-10 were determined using a commercially
available ELISA kit (OptEIA, Pharmingen, San Diego, CA. USA). The detection limits of
the assays were 15 pg/ml for TNF and IL-10, 10 pg/ml for IL-4, IL-6 and IFN-γ.
- 57 -

3.10 Light microscopy
3.10.1 HE-staining
Materials & Methods
For light microscopy, liver samples were fixed in 4% buffered formalin and
embedded in paraffin. Five-micrometer sections were cut (Biocut 2030, Reichert Jung,
Germany) and stained with hematoxilin and eosin (Merck Biosciences, Darmstadt,
Germany). Representative sections are shown.
3.10.2 Tunnel-staining
For light microscopy, liver samples were fixed in 4% buffered formalin and
embedded in paraffin. Five-micrometer sections were cut (Biocut 2030, Reichert Jung,
Germany) and stained with sheep-Digoxygenin-POD (Roche, Penzberg, Germany) for
60 min at 37°C. After flushing with buffer, anti sheep-Peroxidase-conjugated antibody
(Dianova, Hamburg, Germany) was added and slices were incubated for 60 min at 37°C.
After flushing with buffer for 5 min, 3-amino-9ethyl-carbazole (Sigma, Deisenhofen,
Germany) was added for 25 min before the slices were embedded in gelatine.
Representative sections are shown.
3.11 Measurement of enzyme activities
3.11.1 Liver enzyme activities in plasma or samples
The extent of liver damage was assessed by measuring plasma alanine
aminotransferase (ALT) activity with an EPOS 5060 analyzer (Netheler & Hinz,
Hamburg, Germany) according to the method of Bergmeyer (367-369) and calculated as
U/l plasma.
- 58 -

3.11.2 Lactate dehydrogenase activity
Materials & Methods
Lactate dehydrogenase (LDH) was determined in hepatocyte cell culture
supernatants (S) and in the remaining cell monolayer (C) after lysis with 0.1% Triton X-
100 according to Bergmeyer (367-369). The percentage of lactate dehydrogenase
release was calculated from the ratio of S/(S+C). Alternatively LDH was measured using
the cytotoxicity detection kit (Roche Diagnostics GmbH, Mannheim, Germany). The
percentage of lactate dehydrogenase release was calculated equal to the calculation
mentioned above.
3.11.3 Caspase 3,7-like activity
Cytosolic extracts from liver tissue were prepared by Dounce homogenization in
hypotonic extraction buffer (25 mM HEPES, pH 7.5, 5 mM MgCl2, 1 mM EGTA, 1 mM
Pefabloc® and pepstatin, leupeptin and aprotinin, 1 µg/ml each), subsequently
centrifuged (15 min, 14,000 x g, 4°C) and stored at -80°C. To determine caspase-3-like
protease activity, the fluorometric DEVD-afc cleavage assay was carried out according
to the method originally described by Thornberry (370). Cytosolic extracts (10 µl,
approximately 1 mg/ml protein as estimated with the Pierce-Assay (Pierce, IL, USA)
were diluted 1:10 with substrate buffer (55 µM fluorogenic substrate DEVD-afc in 50 mM
HEPES, pH 7.4, 1% sucrose, 0.1% CHAPS, 10 mM DTT). Generation of free 7-amino-4trifluoromethylcoumarin
(afc) at 37°C was kinetically determined by fluorescence
measurement (excitation: 385 nm; emission: 505 nm) using the fluorometer plate reader
Victor2 (Wallac Instruments, Turku, Finland). The activity was calculated using serially
diluted standards (0-5 µM afc). Control experiments confirmed that the activity was linear
with time and with protein concentration under the conditions described above.
Alternatively, to determine caspase-3-like and caspase 8 activity in vitro primary
hepatocytes were lysed (freeze-thaw in lysis buffer: 25 mM HEPES pH 7.5, 5 mM
MgCl2, 1 mM EGTA, 1 mM Pefablock and pepstatin, leupeptin and aprotinin, 1 µg/ml
each, 0.1% Triton X-100), subsequent centrifuged (15 min, 14,000 g, 4°C), the
supernatant diluted 1:10 in substrate buffer including DEVD-afc (caspase3) or IETD-afc
(caspase-8) as substrate and generation of free afc determined using the fluorometer
plate reader Victor 2 .
- 59 -

Materials & Methods
3.11.4 TNF-α-converting enzyme (TACE)-activity
Recombinant TACE enzyme (Merck Biosciences, Darmstadt, Germany) was
reconstituted in 1 ml of sterile water and aliquots were frozen at -20° C until usage as
indicated in the data sheet. For photometric analysis, TACE was solved in a buffer
containing 50 mM Tris-HCl, pH 7.4 50 mM NaCl and 4% glycerol. For monitoring TACE
activity, an internally quenched fluorogenic substrate (Substrate IV, Merck Biosciences,
Darmstadt, Germany) was used with an excitation maximum of ~320 nm and an
emission maximum of ~420 nm (371). TACE inhibition experiments were carried out at a
temperature of 37°C in a total volume of 1.0 ml containing 200 ng/ml of the recombinant
protein, 5 µM of TACE substrate IV and various concentrations of a known inhibitor for
TACE, TAPI-1 (Merck Biosciences, Darmstadt, Germany), or melphalan. Spectra were
generated using a Luminescence Photometer LS 50B (Perkin Elmer, Rodgau-
Jügesheim, Germany) with a slot of 5 nm.
3.12 Determination of ATP
Livers were perfused for 2 sec with ice-cold Somatic Cell ATP Releasing Reagent
(Sigma, Germany) and homogenised by Dounce homogenisation. The 10% homogenate
(in ATP Releasing Reagent) was centrifuged at 13,000 x g for 5 min at 4°C.
Immediately, the supernatants were diluted and luminescence was measured in 96-well
plates using an automated procedure (Victor² fluorometer plate reader (Perkin Elmer Life
Sciences, Turku, Finland). Data were compared to calibration solutions and ATP
contents are expressed in relation to protein content as percent of untreated control
livers.
For cellular ATP-determination of isolated hepatocytes, medium was replaced by
150 µl ALP-lysis buffer (ATP Bioluminescence Kit HS II, Boehringer Mannheim,
Germany) per well (24well format) and immediately frozen at -80°C. After thawing up,
samples were diluted 1:10 in dilution buffer according to the standard protocol of the kit
and subsequent measured in 96-well plates as mentioned above. Data were compared
to calibration solutions and ATP contents are expressed in relation to protein content as
percent of untreated control hepatocytes.
- 60 -

3.13 Determination of glutathione
Materials & Methods
The amount of total glutathione (GSx = GSH + 2 x GSSG) in liver cells was
quantified according to the enzymatic cycling method originally described by Tietze et al.
(372) Isolated hepatocytes were incubated as described above. At different time-points
indicated in the text, supernatant was removed and 100µl lysis-buffer containing 100µM
HCl and 10mM EDTA were added and immediately frozen at -80°C. After thawing up,
cellular material was separated from supernatant by centrifugation (5 min, 1,000 g, 4°C).
150µl of the supernatant was removed and total glutathione was quantified with an ACP
5040 analyser (Eppendorf, Hamburg, Germany).
Alternatively the Glutathione Assay Kit from Cayman (Ann Arbor, MI, USA) was used.
The assay was performed according to the manual instructions.
3.14 Cytokine determination
Antibody pairs (specific rat anti-murine mAb) were purchased from Pharmingen
(San Diego, CA, USA). Streptavidin-peroxidase was obtained from Jackson Immuno
Research (West Grove, PA, USA) and the TMB liquid substrate system from Sigma
(Deisenhofen, Germany). TNF and IFN-γ were determined using a commercially
available ELISA kit (OptEIA, Pharmingen, San Diego, CA. USA). The detection limits of
the assays were 15 pg/ml for TNF and 10 pg/ml for IFN-γ.
3.15 Cell-based ELISA
This assay was performed as described previously (373). In order to assess the
expression of membrane-bound TNF, Kupffer cells (10 5 cells/well) were left to adhere in
96-well plates for 24 hours and were then either left untreated or were treated with LPS
(10 ng/ml) or melphalan (200 µg/ml) for various times of incubation. Subsequently, cells
were washed with Hanks Balanced Salt Solution (HBSS, BioWitthaker, Verviers,
Belgium), fixed for 45 min at room temperature with HBSS + 1% paraformaldehyde,
washed again, after which non-specific binding was blocked by means of incubating the
wells with HBSS + 5% bovine serum albumin (Serva, Heidelberg, Germany).
Subsequently, cells were incubated for 60 min at room temperature under mild shaking
with either 30 µl/well of HBSS + 5% BSA, with 10 µg/ml of the 1F3F3 neutralising rat-
- 61 -

Materials & Methods
anti-mouse TNF mAb (374), with 10 µg/ml of the Sylvio neutralising sheep-anti-mouse
TNF mAb, or with 10 µg/ml of an isotype-matched control rat IgM (Innogenetics, Ghent,
Belgium), both diluted in HBSS + 5% BSA. After 2 washing steps, cells were incubated
for 45 min under mild shaking with 3 mg/ml of a goat-anti-rat IgG-alkaline phosphatase
conjugate (when first antibody 1F3F3) or rabbit-anti-sheep IgG-alkaline phosphatase
conjugate (when first antibody Sylvio) and washed 2 times with HBSS + 5% BSA and 1
time with 2.5 M diethanolamine, pH 9.5. Finally, the substrate solution consisting of 0.58
mg/ml of the fluorescent substrate Attophos (Promega, Madison, WI, US) and of 2.4
mg/ml of the endogenous phosphatase activity blocking agent levamisole (Sigma,
Heidelberg, Germany) diluted in diethanolamine buffer was added. After 5 min,
fluorescence was measured at excitation wavelength 485 nm and emission wavelength
530 nm in the Victor² fluorometer plate reader (Perkin Elmer Life Sciences, Turku,
Finland). Background values due to the unspecific control IgM antibody were subtracted
from the anti-mTNF IgM settings.
3.16 Western blot analysis
Samples resulting from cellular extracts were loaded on a 12% polyacrylamide gel,
separated under reducing conditions and subsequently blotted on a Hybond
nitrocellulose membrane (Amersham-Buchler, Brauenschweig, Germany). In a Bio-Rad
semi-dry blotter at 0.8 mA/cm² for 45 min. Homogenous transfer to the membrane was
controlled by Ponceau red staining. Blots were blocked for at least 45 min with 5% nonfat
dry milk in TBS/T containing 0.05% Tween-20 and subsequently incubated with antimouse
TNF monoclonal antibody (Sylvio) (1:5,000 in TBS/Tween, 5% non-fat dry milk).
After washing and incubation with a polyclonal IgG-horseradish peroxidase coupled
secondary antibody (1:10,000 in TBS/Tween, 5% non-fat dry milk), the blots were
developed by chemiluminescence method (ECL, Amersham-Buchler, Brauenschweig,
Germany) following the manufacturer’s protocol.
- 62 -

3.17 Statistics
Materials & Methods
All data are generally given as means ± SD or S.E.M. as indicated. Statistical
differences were determined by one-way analysis of variance (ANOVA) followed by
Dunnett multiple comparison test of the control vs. other groups. Statistical analysis that
included all vs. all comparisons was done by the Tukey multiple comparison test. The
statistic program InStat ® (GraphPad software, USA) was used for statistics and a p
value

4 Results
Results
4.1 Characterisation of melphalan-induced cell death in primary murine
liver cells
Melphalan induced a concentration-dependent cytotoxicity in primary murine liver
cells, with an EC50 of around 100 µg/ml (figure 5A), and lead to a steady increase of
LDH release over control incubations 9 hours after drug exposure (figure 5B).
The mode of cell death induced by this substance was identified as apoptosis, as
characterised by typical nuclear alterations i.e. chromatin condensation and cellular
alterations as rounding off, membrane blebbing and finally secondary lysis of apoptotic
vesicles (figure 6A).
- 64 -

Results
- 65 -

Results
Interestingly, at high melphalan concentrations a slightly different shape of cell death
could be observed beginning 9 hours after melphalan exposure (figure 7).
This shape of death was previously described in primary hepatocytes incubated
with the plant lektin Con A and characterised as a kind of feathery structure with
membrane obtrusions (361). This distinctive morphology could be observed till about 14
hours following melphalan incubation before the cells ended up in apoptosis. In addition
to the morphological findings an increase of caspase-3-like protease activity measured
in liver cell lysates could be obtained. In lysates of cells incubated with 200 µg/ml
melphalan the peak of caspase activity was observed between 12 and 18 hours (figure
6B). Incubations with higher melphalan concentrations (300 and 400 µg/ml) revealed
that these concentrations lead to a switch from apoptosis to necrosis as indicated by a
different morphology and a lack of caspase-activity (data not shown). These results were
also supported by the finding that the pan-caspase inhibitor zVAD was able to protect
cells from melphalan-induced apoptosis at 200 µg/ml (figure 6C+D) but not at 400 µg/ml
(data not shown). Due to these initial findings most of the experiments were performed
with concentrations ranging from 50 to 200 µg/ml.
- 66 -

Results
4.2 The role of death receptors in melphalan-induced apoptosis
Apoptosis is known be induced via an intrinsic (mitochondrial) and an extrinsic
receptor-mediated pathway. Therefore, we have been investigated the effect of death
receptors and their role in melphalan-mediated toxicity. Melphalan is known to induce
interstrand and intrastrand crosslinks in the DNA which might be sufficient to induce
apoptosis via an intrinsic apoptotic pathway. Since TNF, together with CD95 ligand,
represents one of the most important known apoptosis-inducing cytokines in
hepatocytes, we subsequently have been investigated the potential implication of the
death receptors CD95 and TNFR1 in melphalan-induced apoptosis. To find out whether
the death receptor CD95 has any influence on melphalan-mediated cell death in
hepatocytes, we made use of mice expressing a non-functional CD95 (375). As such we
did not detect any difference in sensitivity towards melphalan-induced apoptosis in cells
derived from normal C57Bl/6 mice, as compared to cells derived from lpr-mice, lacking a
functional CD95 receptor on the surface of immune cells and several other cell types
including hepatocytes (figure 8). Furthermore, addition of an agonistic anti-CD95
monoclonal antibody (Jo-2) had no enhanced effect on cell death when co-incubated
with 200 µg/ml of melphalan (data not shown).
- 67 -

Results
TNF mediates its cellular effects mainly via two TNF receptors of which only
TNFR1 is known to mediate apoptosis through an intracellular death domain whereas
TNFR2 seems to fulfil other functions. Interestingly, it turned out that cells derived from
TNFR1 knockout mice were fully protected from melphalan-induced cell death in
concentrations up to 400 µg/ml (figure 9A). To verify if melphalan-induced toxicity is
altered when TNFR2 is missing, cells were derived from TNFR2 knockout mice and
incubated with melphalan. Unexpectedly, these cells were also protected from
melphalan-induced toxicity (figure 9B), indicating a role for both TNF receptors. The
observation that cells derived from TNFR1R2 double knockout mice are also protected
is in line with previous observations (figure 9C).
- 68 -

Results
Last but not least, a third group of death receptors also belonging to the TNF
receptor superfamily with its most prominent members TRAIL receptors 1 and 2 (DR4
and DR5) has been shown to be able to induce death in hepatocytes. We therefore
incubated cells derived from wild type mice with different concentrations of kiTRAIL
alone or in combination with melphalan, but no additional effects could be observed by
this agonist. Indeed, we were not able to induce death in primary liver cells by this ligand
alone or in combination with transcriptional or translational inhibitors, like actinomycin D
or cycloheximide. On the contrary, cell lines like Hep G2 cells were sensitive upon ligand
binding with an EC50 of about 3 ng/ml after sensitisation with 100 µM cycloheximide
(data not shown). These results contribute to recent publications demonstrating that
DR4 and DR5 are not present on murine liver cells and stress the fact that the cellular
effects observed after melphalan exposure are linked to a specific reaction involving
TNF receptors. Therefore, we attempted to investigate the role of TNF in this setup.
4.3 TNF mediates melphalan-induced toxicity
Since we could demonstrate that both receptors for TNF play a crucial role in
melphalan-mediated apoptosis, we subsequently have been investigated the effect of a
neutralising monoclonal rat anti-mouse TNF antibody 1F3F3 on melphalan-induced
apoptosis. This antibody has been demonstrated to neutralise both soluble and
membrane bound TNF (374, 376). Neutralisation of TNF abrogated melphalanincreased
LDH release completely to the level of control incubations (figure 9D)
confirming the causal role of TNF and its receptors as demonstrated before.
Complementary results could be obtained by using another polyclonal sheep-anti mouse
TNF neutralising antibody (Sylvio, inhouse preparation). Assuming that TNF is the major
mediator of melphalan toxicity, additional TNF should be able to further enhance the
observed toxicity. But this was not the case. However, co-incubations of melphalan and
soluble TNF did not enhance the observed toxicity compared to incubations with
melphalan alone (figure 10).
- 69 -

Results
4.4 Kupffer cells are the source of TNF mediating melphalan-induced
toxicity
Since TNF seemed to be responsible for melphalan-induced apoptosis induction
via interaction with TNF receptors and since no report has been shown that hepatocytes
are able to produce TNF, which might lead to autocrine stimulation, the question
remained what the source of TNF has been. Previous experiments showed that cells
isolated according to the method of Seglen et al. are not a pure fraction of hepatocytes
but are contaminated with up to 10% of non-parenchymal cells (NPCs), mainly Kupffer
cells. These liver-specific macrophages have been shown to produce huge amounts of
soluble TNF upon stimulation with e.g. endotoxin (figure 11) and therefore might also be
able to produce sufficient amounts of TNF when contaminating primary hepatocyte
cultures.
- 70 -

Results
To investigate this problem, we prepared highly purified primary hepatocytes
upon depletion of Kupffer cells, achieved by intravenous injection of clodronate
liposomes and additional selection of hepatocytes as described briefly in materials and
methods. In incubations of these purified hepatocytes with melphalan no more toxicity
was detected (figure 12A, white bars). However, hepatocyte apoptosis could be restored
by the addition of exogenous TNF to melphalan (figure 12A, black bars). Moreover,
hepatocyte apoptosis induced by 200 µg/ml of melphalan could also be restored by
addition of increasing numbers of previously purified NPCs (mainly Kupffer cells) as
shown in figure 12B. These results strongly indicate that this cell type is the source of
TNF. To exclude that TNF production of Kupffer cells was due to LPS contamination of
melphalan, we investigated LPS contamination in a LAL-test revealing that melphalan
had only a weak contamination with LPS (11,5 ± 5,78 pg/ml). In addition we incubated
melphalan with primary murine hepatocytes derived either from C3H/HeN-mice or
C3H/HeJ-mice. The latter mouse strain is containing a mutation in tlr-4 leading to loss of
function of this receptor. Lacking cytokine production and LPS-related signaling in
response to endotoxin has been demonstrated for immune cells derived from these
mice. Data obtained with liver cells of these mice are in line with the results from the
LAL-test, showing no LPS effect influencing melphalan-induced toxicity (figure 12C).
- 71 -

Results
Notably, TNF as such failed to directly induce apoptosis in hepatocytes, unless
transcriptional or translational inhibitors are also present. These findings thus indicate
that in absence of Kupffer cells the simultaneous presence of melphalan (which is a
transcriptional inhibitor) and TNF is a necessary condition to initiate hepatocyte
apoptosis.
4.5 Kupffer cells lead to enhanced TNF secretion after LPS but not after
melphalan stimulation
To investigate whether melphalan is able to induce TNF production in isolated
Kupffer cells, NPCs of wild type mice were prepared and incubated with different
concentrations of LPS or melphalan and secretion of soluble TNF was determined in the
supernatants. As shown in figure 13A, secreted TNF was not detected in supernatants
of Kupffer cells incubated with various concentrations of melphalan LPS (10 ng/ml)
instead increased TNF production and secretion significantly. Interestingly, when coincubated
with melphalan LPS-induced TNF production was totally abolished. Moreover,
this LPS-induced TNF production could be inhibited by pre-treatment with
- 72 -

Results
pentoxyphylline, an unspecific inhibitor of phosphodiesterases (figure 13B), showing that
the observed increase of soluble TNF release is related to LPS stimulation. Notably, this
TNF release was also significantly reduced when LPS was co-incubated with melphalan
on MH-S cells, a lung macrophage cell-line (ctrl:163 ± 2,3 pg/ml; LPS: 5,879 ± 574
pg/ml; LPS/TAPI-1: 3,125 ±156 pg/ml; LPS/melphalan: 2,623 ± 125 pg/ml; melphalan:
408,5 ± 48 pg/ml) (data not shown).
- 73 -

Results
To verify this result also in the whole organ we prepared livers from wild type mice
and perfused them in presence of LPS or melphalan. TNF was determined in perfusate
samples at different time points. As such, a weak induction of soluble TNF with a
maximum between 90 and 120 min after addition of LPS was detected. When perfused
with 150 mg/kg of melphalan only amounts of TNF below levels of untreated control
livers could be detected (figure 13C). Moreover, mice injected with either LPS or
melphalan or both displayed release of TNF in the circulation when challenged with
melphalan or melphalan combined with LPS, whereas mice challenged with LPS alone
displayed high amounts of serum TNF 90 min after LPS injection (figure 13D).
TNF is known to be produced first as an active membrane-bound protein before
being cleaved off via an also membrane associated matrix-metalloproteinase called
TACE (TNF-alpha converting enzyme) and released as soluble pleiotropic factor. We
investigated if the amount of TNF presentation was modified on the surface of Kupffer
cells after melphalan stimulation. Therefore, Kupffer cells were prepared and incubated
with either melphalan or LPS. Membrane-associated TNF was determined subsequently
using a cell-based ELISA. In line with previous observations, treatment of purified
Kupffer cells with 200 µg/ml melphalan led to a significantly increased expression or
presentation of membrane-bound (figure 14B) but not of secreted TNF (below the
detection limit of the murine TNF-ELISA of 20 pg/ml for all investigated time points).
Such increased expression or presentation of membrane-bound TNF was measured
from 30 min on after melphalan incubation (figure 14B). In contrast, LPS (10ng/ml)
increased both the production of membrane-bound (figure 14A) and secreted TNF (data
not shown) with clearly different kinetics from melphalan.
- 74 -

Results
Since no positive and negative controls are available for this cell-based ELISA, we
performed Western blots of RAW-cells, a murine Kupffer cell line, stimulated with
melphalan or LPS. In that set of experiments a maximum of membrane TNF (about 26
kDa) was observed after 2 hours of incubation in lysates of RAW cells stimulated with
25, 50 or 100 µg/ml of melphalan whereas untreated control cells did not (figure 15A).
Less membrane TNF could be detected after 1 hour and 3 hours of incubation. When
cells were stimulated with concentrations ranging from 12.5 µg/ml to 200 µg/ml,
maximum expression of membrane TNF could be observed at 100 µg/ml (figure 15B).
When cells were incubated with LPS (0.1 µg/ml; 1.0 µg/ml and 10 µg/ml) a band of 26
kDa could also be detected which was not present in untreated control cells (figure 15B).
- 75 -

Results
4.6 Melphalan inhibits recombinant TACE activity in vitro
Since melphalan has been shown to be able to abrogate LPS-induced TNF
secretion in primary macrophages, while leading to significantly increased levels of
membrane associated TNF on the surface of Kupffer cells, we were interested if
melphalan might be able to exert any inhibitory effect on TACE enzyme activity.
Therefore, we set up a fluorimetric enzyme activity test for TACE using the recombinant
protein and a specific chromophore-coupled substrate as described in the materials &
methods section. To test the quality of the enzymatic assay, we measured TACE-activity
in the presence of a known inhibitor for TACE, TAPI-1, with an IC50 of about 10 nM. As
such, we found an IC50 of about 20 nM in our enzymatic assay for this inhibitor in three
independent sets of experiments (figure 16A), demonstrating the quality of this assay.
Incubations done in the presence of various concentrations of melphalan revealed that
the study drug was indeed able to inhibit TACE-activity in a concentration range that fit
in very well to results obtained in vitro before (figure 16B), stressing the hypothesis that
TACE was also inhibited in our cellular assays.
- 76 -

Results
4.7 Melphalan is able to block TNF-dependent liver injury in vivo
To investigate the role of melphalan-mediated inhibition of TACE enzyme activity in
the living animal, we performed in vivo experiments, using known mouse models of
cytokine-mediated liver injury. If melphalan is also able to block TACE activity and the
release of soluble TNF in vivo, then these mice should at least be partially protected
from cytokine-mediated liver injury.
4.7.1 GalN/TNF and GalN/LPS model
TNF has been shown to be hepatotoxic in man and mice (377). Mice can be
dramatically sensitised towards TNF by viral or bacterial infection or chemicals such as
galactosamine (GalN), resulting in an up to 10 4 -fold increased lethality (378) and
hepatotoxicity (379). The amino sugar GalN selectively depletes uridine in hepatocytes
(380) and therefore causes a selective transcriptional inhibition in the murine liver. The
GalN/TNF model uses exactly the liver-specific inhibitory capacity of GalN to induce a
liver specific toxicity in mice by activating apoptosis in the liver through exogenously
applicated TNF. The GalN/LPS model is a more complex scenario for GalN/TNF
mediated liver failure. The liver harbours the largest pool of resident macrophages in the
body, i.e. Kupffer cells and thus it is a potent TNF-producing organ (381). Kupffer cells
within the liver may produce TNF after direct stimulation, e.g. with LPS or after
interaction with activated lymphocytes in a more complicated immunological setting
(382) leading to a pathological situation, which finally leads to liver failure in GalNsensitised
mice. Passive immunisation of mice against TNF prevented apoptosis (383)
as well as the release of transaminases being a late hallmark of liver damage (383,
384).
Mice challenged with GalN (500 mg/kg) and LPS (Salmonella abortus equi 1.5
µg/kg) displayed significantly enhanced liver-toxicity over controls (serum ALT-activity:
19.6 ± 10 U/L; data not shown), eight hours after challenge. Toxicity was characterised
by a significantly enhanced ALT activity in serum, as compared to controls obtained
eight hours after challenge (figure 17A). In addition, mice displayed a significantly
enhanced TNF release 90 min after challenge. Animals pre-treated with 100 mg/kg
melphalan instead displayed significantly decreased levels of ALT in serum, as
compared to mice challenged with GalN/LPS alone. Also TNF levels were significantly
reduced in these pre-treated mice (figure 17B).
- 77 -

Results
Interestingly, this effect could also be observed when GalN/LPS-challenged mice
were pre-treated with 75 mg/kg or 50 mg/kg melphalan (figure 17C+D). This inhibitory
effect of melphalan could also be observed when GalN/LPS-challenged mice were
injected with melphalan (50 mg/kg – 100 mg/kg) at the same time as GalN/LPS or up to
30 min after GalN/LPS-challenge (figure 17E+F). When melphalan was given one hour
after GalN/LPS-challenge, ALT release was still reduced, but TNF release was shown to
be enhanced (figure 17E+F). This finding is of further interest, because we proposed
- 78 -

Results
that the protective effect of melphalan is only due to its ability to abrogate TNF-release.
But in this case soluble TNF was present in the circulation while ALT-release was still
reduced.
To investigate this effect, we injected mice with GalN (500 mg/kg) /TNF (10 µg/kg)
with and without pre-treatment of melphalan (100 mg/kg). Interestingly, these mice
displayed also a significantly decreased plasma level of ALT activity 8 hours after
challenge (figure 18A). Since this result was not expected we were further interested if
other parameters in this well-characterised model are also altered. Therefore we
determined caspase-activity in liver samples collected after the experiment and found
significantly reduced levels of downstream caspases in melphalan pre-treated mice
(figure 18B). Finally, we determined plasma activity of IFNγ eight hours after challenge,
but no significant difference between GalN/TNF-treated mice and mice additionally pretreated
with 100 µg/mg of melphalan was found (figure 18C).
- 79 -

4.7.2 GalN/SEB
Results
The superantigen SEB exerts its toxicity through activation of a subset of T cells
and subsequent production of TNF (385-389). Essential for the induction of apoptotic
liver damage by SEB is the presence of transcriptional (386) or translational (365, 390)
inhibitors. Dendritic cells are considered essential for clonal expansion of Vß8-specific T
cells (391) and local expression of cytokine mRNA within the spleen (392).
Macrophages, especially Kupffer cells, have also been shown to contribute essentially to
this scenario being important for TNF production (365).
Animals injected with GalN (700 mg/kg) and SEB (2 mg/kg) showed a severe liver
toxicity eight hours after challenge, as shown by serum ALT activity (figure 19A). In
addition, TNF could be detected in serum collected 90 min after challenge (figure 19B)
and enhanced levels of IFN-γ could be detected eight hours after challenge (figure 19C).
In contrast to these results, animals pre-treated with 100 mg/kg of melphalan one hour
before SEB injection displayed significantly reduced levels of ALT (figure 19A), TNF
(figure 19B) and IFN-γ (figure 19C). However, the observed spleen hypertrophy could
not be blocked by melphalan (data not shown).
- 80 -

4.7.3 The Con A model.
Results
Concanavalin A (Con A) is a plant lectin that activates T cells in vitro and in vivo.
When injected into mice, Con A triggers a selective liver injury (366) depending on the
release of the cytokines TNF (393) starting 90 min after challenge, IFN-γ (382) and IL-4
(394). Further, the Con A model involves an early damage of sinusoidal endothelial cells
(298, 395) and requires a cross talk between macrophages and T cells (365, 382). In the
Con A model, both necrotic and apoptotic hepatocyte demise with or without a
contribution of caspases have been described (297, 366, 393, 396, 397). Interestingly,
hepatotoxicity induced by Con A does not require any kind of additional sensitisation and
involves both TNF receptors, TNFR1 and TNFR2, as shown by studies using tnfr1 or
tnfr2 gene deficient mice. Moreover, in contrast to all other TNF-dependent models of
liver injury requiring sensitisation by transcriptional inhibitors, Con A-induced liver injury
is independent of caspase-3-like protease activity, as shown by studies using the pancaspase
inhibitor benzoyloxycarbonyl-val-ala-asp-fluoromethylketone (zVADfmk) (297).
Animals injected with 25 mg/kg of ConA displayed fulminant liver toxicity eight
hours after the treatment, as assessed by ALT-release in the serum (figure 20A). TNF
release (90 min after challenge) (figure 20B) and IFNγ-release (8 hours after challenge)
(figure 20C) in the circulation could be also detected in these mice. Melphalan pretreatment
(100 mg/kg) instead, totally abrogated release of cytokines (TNF and IFN-γ)
(figure 20B+C) and hepatotoxicity was significantly reduced (figure 20A). Since this
model involves a cross talk between macrophages and T cells, we were interested if the
well-characterised cytokine profile was altered upon melphalan pre-treatment.
Therefore, we injected mice with 25 mg/kg of ConA with and without pre-treatment of
100 mg/kg of melphalan for 1 hour and determined IL-4, IL-10, IL-1ß, IL-2 and IFNγ in
blood samples collected after 90 min, 2 hours, 3 hours and 8 hours, respectively. As
expected melphalan pre-treatment altered the whole cytokine profile of this model with
TNF, IL-1ß and IFNγ being significantly reduced and IL-10 being significantly enhanced
as compared to Con A treatment alone.
- 81 -

Results
- 82 -

4.7.4 The CD95 model
Results
This model has been used because liver failure induced by injection of agonistic
antiCD95 antibody has been shown to be totally independent of TNF and its
corresponding TNF receptors. The application of the activating anti-CD95 antibody Jo-2
in naive mice leads to lethal liver destruction within hours, due to massive caspasemediated
apoptosis of hepatocytes (398, 399). Importantly, this effect does not require a
sensitisation of the animals and is seen when instead of anti CD95, the CD95L is used
or endogenously generated (400, 401). Other organs than the liver have also been
reported to be affected by anti CD95 (402, 403), but this depends largely on the
antibodies and mouse strains used and therfore is of minor importance when Jo-2 was
used (404). CD95 is also highly expressed in sinusoidal endothelial cells (405) and a
prominent damage of these cells precedes the onset of liver damage in this model (406).
Mice injected with agonistic αCD95 antibody (2 µg/mouse) displayed a significantly
enhanced hepatotoxicity eight hours after challenge (figure 21A). As expected,
melphalan pre-treatment (100 mg/kg, 60 min) was not able to block CD-95 mediated cell
death, but was able to reduce serum IFN-γ significantly detected 8 hours after αCD-95
challenge (figure 21B).
- 83 -

Results
4.7.5 Melphalan and melphalan/TNF model
Since melphalan has been found to induce apoptosis in primary liver cells and in
isolated perfused organs, we were interested, if melphalan is also able to induce liver
toxicity in vivo. Therefore, we injected mice i.p. with various doses of melphalan, ranging
from 25 mg/kg up to 150 mg/kg. Unfortunately, we were not able to detect any liver
toxicity eight hours after challenge or in animals died earlier (data not shown). Mice
injected with high doses of melphalan displayed renal alterations associated with bloody
urine and black coloured bladders, intestinal toxicity associated with midgut changes
and diarrhoea. Lowered body temperature as compared to untreated control animals
and impaired balance was also observed. Therefore, we investigated if mice challenged
with lower doses of melphalan (25 mg/kg – 75 mg/kg) and TNF (10 µg/kg) display any
hepatic toxicity. But these experiments also failed (data not shown). Since no extended
knowledge could have been expected from further in vivo experiments compared to data
obtained in vitro no further investigations have been undertaken, also with respect to
ethical guidelines for animal experiments.
4.8 Cell-cell contact is necessary between Kupffer cells and hepatocytes
Previous results indicated that melphalan was able to induce enhanced
presentation of TNF on macrophages and inhibited the cleavage off the membraneanchor
through TACE. On the other hand Kupffer cells or at least the presentation of
TNF on the surface of Kupffer cells was shown to be essential for apoptosis induction by
melphalan. So we were further interested if membrane-bound TNF indeed was
responsible for mediating melphalan toxicity or if at least very small amounts of soluble
TNF, undetecable in ELISA, were sufficient to stimulate TNF receptors in order to induce
apoptosis in hepatocytes. Therefore, we performed transwell experiments, in which the
Kupffer cells were cultivated in cell culture inserts that were spatially separated from the
hepatocytes cultivated in the bottom wells. As shown in table 4, melphalan alone was
not sufficient to induced cytotoxicity in hepatocytes when Kupffer cells were present but
separated in cell culture inserts. On the other hand, cytotoxicity occurred when cell-cell
contact between both cell types was enabled indicating that this contact between both
cell types was crucial for apoptosis induction.
- 84 -

Results
4.9 Role of TNF receptors on Kupffer cells
Since both TNF receptors have been shown to play a crucial role in mediating
melphalan-induced apoptosis of hepatocytes, and Kupffer cells have been demonstrated
to be necessary for melphalan-induced TNF-mediated cell death, the question was
raised, whether both receptor types are necessary on both cell types. Since TNFR2 is
lacking an intracellular death domain and therefore does not play a decisive role in
receptor-mediated caspase activation, we have been interested if this receptor anyhow
is only important in Kupffer cell signaling. TNFR1 instead, containing an intracellular
death domain, would have been sufficient for transmitting the death signal given by TNFpresenting
macrophages in this scenario. This alternative hypothesis would have also
included a role for both TNF receptors, but on separate cell types; TNFR1 on the
hepatocyte and TNFR2 on the Kupffer cell.
- 85 -

Results
To investigate this theory we prepared hepatocytes and Kupffer cells from different
mouse strains, purified them and performed co-culture experiments in all possible kinds
of variations using wild type, TNFR1, TNFR2 and TNFR1R2 knockout mice. The results
obtained in these experiments are summarized in figure 22. In detail, hepatocytes
isolated from wild type mice and co-cultured with Kupffer cells derived either from
TNFR1 or TNFR2 or TNFR1R1 knockout mice died in response to melphalan,
independent of the expression of TNF receptors. On the contrary, hepatocytes derived
from TNFR1R2 knockout mice died not when co-cultured with different Kupffer cells and
melphalan (figure 22 B).
In conclusion these results demonstrated that both TNF receptors contribute to the
toxicity on the surface of hepatocytes, while the expression of TNF receptors on the
surface of Kupffer cells did not play a decisive role in mediating melphalan induced
toxicity.
- 86 -

Results
4.10 Melphalan-induced toxicity on Kupffer cells
Since we have been demonstrated that Kupffer cells play a crucial role in
mediating melphalan-induced toxicity via both TNF receptors, the question remained
how purified Kupffer cells respond to various melphalan concentrations.
To investigate this problem, we isolated primary murine liver cells and purified
Kupffer cells as described above. Cells were incubated with different concentrations of
melphalan, varying from 6.25 µg/ml to 400 µg/ml of melphalan and cell viability was
determined by Alamar blue assay, as well as by LDH release using the cytotoxicity
detection kit (Roche diagnostics). As we could demonstrate, isolated Kupffer cells died
in response to melphalan with a sensitivity comparable to hepatocytes (EC50 = 130
µg/ml) (figure 23A). Interestingly, Kupffer cells isolated from mice deficient for both TNF
receptors, displayed a similar concentration response curve, with an EC50 of 75 µg/ml
(figure 23B), indicating that this cell death was independent of TNF receptor
engagement. In kinetic experiments with Kupffer cells derived from wild type mice we
could demonstrate that cells have been intact in the first nine hours following melphalan
treatment since no LDH could be detected in the supernatant before that time point (data
not shown). Having a closer look on the mode of cell death we incubated cells from wild
type mice with melphalan in the presence and absence of 1 µM zVAD. Interestingly, no
beneficial effect of zVAD co-incubation was observed (figure 23C), and corresponding to
this no caspase-3-like protease activity could be determined (data not shown).
- 87 -

Results
These data stress the fact that Kupffer cell toxicity was independent of caspases.
Nevertheless, since we have been measured onset of caspase activity already at about
nine hours in primary liver cell cultures with a caspase-specific activity of about 120
pmol/mgxmin after 12 hours, this Kupffer cell death was not decisive for the toxic effects
on hepatocytes observed in primary liver cell culture. In addition, in presence of anti-TNF
antibodies Kupffer cell death also occured, but hepatocytes were protected from induced
toxicity in primary liver cell culture (data not shown).
4.11 Melphalan-induced hepatotoxicity in the presence and absence of
Kupffer cells
In previous experiments we have demonstrated that the presence of Kupffer cells
is a necessary condition to induce hepatocyte apoptosis by melphalan. Under Kupffer
cell depleted conditions no hepatocyte death could be observed, but toxicity could be
restored by adding small amounts of soluble TNF (figure 12A). This observation raised a
number of questions concerning the kinetics of apoptosis induction, the kinetics of
caspase activation and the crucial involvement of TNF receptors. To address these
questions we prepared hepatocytes from Kupffer cell-depleted mice, additionally purified
them as described in materials and methods and incubated them as indicated.
Interestingly, it turned out that apoptosis induced by melphalan and soluble TNF
under Kupffer cell depleted conditions showed clearly different kinetics, as compared to
apoptosis induced in the presence of Kupffer cells with melphalan alone (figure 24A+B).
Also the onset and the maximum of caspase-3-like activity differed in both settings
(figure 24C). In the presence of Kupffer cells and the absence of exogenous TNF a
maximum of caspase-3-like activity was found 18 hours after melphalan stimulation with
a specific activity of about 200 pg/mlxmin, whereas in the absence of Kupffer cells and
the presence of exogenous TNF the highest caspases-3-like protease activity was
detected seven hours after melphalan/TNF stimulation with an activity of about 300
pg/mlxmin (figure 24C). LDH release was found to be comparable in both settings
showing a maximum between 18 and 24 hours following melphalan or melphalan/TNF
stimulation (data not shown). The amount of cell death assessed was also comparable
in both settings (about 80%) (figure 24A+B). Interestingly, the calculation of effective
melphalan concentration revealed that in the presence of Kupffer cells and the absence
of exogenous TNF EC50 was nearly doubled, as compared to the setting under Kupffer
cell-depleted conditions (figure 24A+B). In addition, also the IC50 values for inhibition of
cell-death by the pan-caspase inhibitor zVAD were completely different, ranging from 6
- 88 -

Results
nM in the situation of Kupffer cell absence to 100 nM in the situation of Kupffer cell
presence (figure 24D).
To address the question how TNF receptors were involved in melphalan/TNFinduced
cell death, hepatocytes from different mouse strains depleted of Kupffer cells
were prepared. As shown in figure 24E, cells lacking TNFR1 were still protected from
melphalan/TNF-induced cell death, while cells derived from tnfr2 gene deficient mice
were susceptible to melphalan/TNF-induced cell death (figure 24F). Notably, cells
lacking both TNF receptors were also protected from apoptosis in this setting (data not
shown).
- 89 -

Results
It is also worth noting that cells derived from macrophage-depleted wild type or
TNFR2 gene deficient mice were also susceptible to apoptosis induction by melphalan
and human TNF which is known to interact only with murine TNFR1 and not with murine
TNFR2 (data not shown). These results were in line with previous findings, but again
raised the question how TNFR2 was involved in this scenario when Kupffer cells were
present.
4.12 Melphalan induces TNF-mediated hepatotoxicity in situ
In previous experiments we could demonstrate that melphalan-induced hepatocyte
death in vitro was mediated by membrane TNF and both TNF receptors in the presence
of Kupffer cells. It could also be demonstrated that due to the inhibition of TACE activity
by melphalan no soluble TNF was released or at least in amounts being below the
detection limit of the ELISA. In animal experiments melphalan pre-treatment prevented
TNF-mediated liver failure in mice, but melphalan alone or in combination with
exogenous TNF did not induce enhanced liver apoptosis or necrosis.
In order to investigate if melphalan induced TNF-mediated hepatocyte apoptosis
was only occurring under in vitro conditions, we decided to confirm the results obtained
in vitro also under conditions of isolated liver perfusion. Isolated liver perfusions with
melphalan and TNF represent a model being closely related to the situation in patients
undergoing IHP for tumor therapy.
In this model, control organs did not undergo a significant hepatotoxicity for up to
540 min as indicated by the absence of ALT (figure 25A ▲) and LDH release (data not
shown) in the perfusate. In addition, no caspase 3-like-activity was measured in liver
samples collected after the experiment (data not shown) and liver architecture was
found to be normal as assessed in H.E. staining and TUNNEL staining of liver slices
collected from the perfused organs. In contrast, perfusion with 150 mg/kg of melphalan
induced a significant hepatotoxicity, as evidenced by the release of ALT in organs
derived from wild type (figure 25A ■) but not from knockout mice lacking TNFR1 (figure
25B ×), TNFR2 (figure 25B ○), or both TNF receptors (data not shown). Moreover, the
causal role of TNF action and crucial role of Kupffer cells was confirmed in livers derived
from Kupffer cell depleted wild type (figure 25C ■), TNFR1, TNFR2 and TNFR1R2
knockout mice (data not shown). In none of these settings melphalan induced significant
liver failure as indicated by lacking release of ALT in the perfusate and the absence
caspase-3-like activity in liver homogenates of perfused livers (figure 25 D).
- 90 -

Results
In line with these biochemical observations liver architecture was found to be
normal up to 8 hours with some light necrotic areas occasionally seen in the periportal
fields, but not in the central vein areas (figure 26). Analogous to these findings, organs
isolated from wild type mice displayed a significant disturbance of liver architecture after
360 min of perfusion with melphalan (figure 26), which was characterised by endothelial
damage, oedema formation and apoptotic chromatin condensation in a subset of
hepatocytes.
- 91 -

Results
- 92 -

Results
Isolated liver perfusion with melphalan and TNF in the absence of Kupffer cells
was also sufficient to reproduce results obtained in vitro. The onset of toxicity as well as
the course of perfusion and the maximum of ALT release in the perfusate was
comparable in both settings, perfusion with melphalan in the presence of Kupffer cells
and perfusion with melphalan and exogenous TNF in the absence of Kupffer cells (figure
25 C). In case of Kupffer cell depletion a maximum of about 2,800 U/L was reached after
6 hours of perfusion, whereas in livers containing Kupffer cells a maximum of ALTrelease
was observed after 6.5 hours with a total amount of about 1,500 U/L (figure 25
C). Additionally, we could also demonstrate that livers derived from TNFR1 (figure 25 C
♦) or TNFR1R2 knockout mice (data not shown) were fully protected up to 9 hours of
perfusion. On the contrary, livers derived from TNFR2 knockout mice perfused with
melphalan and TNF displayed an enhanced ALT release in the perfusate starting from
330 min on and ending up with a total release of about 2,700 U/L after 420 min, as
compared to the release of less than 100 U/L by untreated control livers,
melphalan/TNF-treated TNFR1-deficient livers (figure 25 C▼) or TNF-treated TNFR2deficient
livers (data not shown). In addition to these results, caspase-3-like activity
measured in liver homogenates of tissue samples collected after the individual
experiments was only found in livers perfused with melphalan and exogenous TNF, not
in livers perfused with melphalan alone. Also no caspase activity could be obtained in
livers derived from mice lacking TNFR1 either perfused with melphalan alone or with
melphalan and TNF together (figure 25 D).
These experiments demonstrated that the in vitro results were reproducible in the
whole organ, a model being closely related to the clinical situation of the local
therapeutic liver perfusion with melphalan. The crucial role of TNFR2 in melphalan
induced liver toxicity could be also demonstrated in the whole organ.
Since very less models of liver injury or pathologic situations have been described
demonstrating a crucial role for TNFR2 in liver apoptosis, we were further interested how
TNFR2 might contribute to this scenario in melphalan induced hepatocyte apoptosis in
vitro.
- 93 -

Results
4.13 Modulation of intracellular signaling by various inhibitors
As has been shown in the initial section of this chapter, melphalan/TNF-induced
toxicity was strongly dependent on the activation of caspases, mainly driven by the
activation of TNFR1. Since it is known that apoptosis via TNFR1, in contrary to CD95,
has been only induced in the presence of transcriptional or translational inhibitors,
melphalan seemed to be able to act in a bifunctional manner, blocking cellular
transcription and/or translation on the one hand and triggering TNF-presentation on the
other hand, which in turn lead to the activation of hepatocyte apoptosis. It is also known
that TNF receptors mediate a variety of cellular stimuli upon binding to their specific
ligand, including several biochemical control centres like NFκB, p38 MAPK or JNK.
Because these molecules have been shown to be involved in many cellular processes, a
number of inhibitors have been generated and made commercially available for selective
inhibition of cellular pathways, in order to prove their crucial role in various cellular
processes.
4.13.1 Role of NFκB
One of the most interesting molecules that has been shown to be activated upon
TNF stimulation is NFκB. In order to investigate the role of NFκB, we used an
ammonium salt pyrrolidinedithiocarbamate (PDTC) which has been demonstrated to
prevent NFκB translocation as well as induction of nitric oxide synthase by inhibiting
translocation of NOS mRNA (407). To test the influence of PDTC, we incubated primary
murine liver cells with concentrations from 0.1 µM up to 100 µM of PDTC alone or in
combination with melphalan in concentrations from 12.5 µg/ml to 400 µg/ml. But as
shown in figure 27A (data shown for 10 µM) PDTC had at least no inhibitory influence on
melphalan-induced toxicity. On the contrary, inhibition of NFκB enhanced melphalan
toxicity at low concentrations.
- 94 -

Results
In order to verify these results in a separate experimental system, we made use of
a special transgenic mouse strain, an inducible NFκB p65 (MXcrep65) knockout mouse.
These mice have been shown to loose NFκB p65 on day 4 till day 6 after intravenous
application of polyIC (250 µg/mouse) (Nektarios Dikopoulos, personal communication).
Interestingly, we found out that MXcrep65 -/- mice or liver cells and isolated perfused livers
derived from those mice were sensitive to TNF alone without pre-incubation of
transcriptional inhibitors such as ActD or GalN when treated on day 4 or day 5 following
polyIC injection. This could be shown by toxicity assays and caspase-3-like protease
activity (data not shown). On the contrary, when mice or liver cells derived from these
mice were not injected with polyIC and stimulated with TNF no apoptosis occured. Also
no apoptosis was observed when mice were injected with polyIC and cells or organs
- 95 -

Results
were treated on day 1 or day 2 after the injection. MXcrep65 +/+ mice, carrying a nonfunctional
cre-lox-system, or cells and organs derived from those mice were not
sensitive to TNF alone only in combination with transcriptional or translational inhibitors
(data not shown).
Primary liver cells derived from MXcrep65 -/- mice and incubated with melphalan as
usual, showed no significant difference in apoptosis induction as compared to cells
derived from MXcrep65 +/+ mice or normal C57Bl/6 mice (figure 27B). Unfortunately, due
to a limited number of mice we got for a co-laborational project we were not able to
reproduce these effects in isolated perfused mouse livers Nevertheless, the results
obtained in primary cell culture are in line with our working hypothesis. To investigate if
NFκB translocates to the nucleus upon melphalan stimulation, we performed NFκBbandshift
assays in primary murine liver cells derived from normal C57Bl/6 mice. Data
obtained by this assay indicated that NFκB p65 shows a normal nuclear translocation
after melphalan stimulation compared to unstimulated and TNF-stimulated controls (data
not shown). Analogous results have been found when NFκB translocation was
measured using the Cellomics HitKit NFκB Activation Assay (Pittsburgh, PA, USA)
(figure 27C).
To summarize these data, NFκB seemed to translocate to the nucleus following
melphalan treatment, but inhibition of NFκB p65 translocation into the nucleus had
neither enhanced toxic effects only at very low concentrations of melphalan, nor any
beneficial influence on the cellular outcome.
4.13.2 Role of p38 MAPK
p38 MAPK is another signaling molecule which has been shown to be involved in
various signaling cascades. P38 MAPK among others is also activated following TNFR1
stimulation with TNF. It has been shown to activate ATF which is able to translocate to
the nucleus. Pre-treatment of primary liver cells with 1 µM of a specific inhibitor for p38
MAPK and incubated with melphalan did also not influence the toxicity induced by
melphalan as determined by cell viability assay (figure 28 A). Measurement of caspase
activity (figure 28 B) confirmed this result, indicating that melphalan-induced toxicity was
also independent of p38 MAPK signaling.
- 96 -

4.13.3 Role of JNK
Results
A third important cellular signaling pathway is related to a kinase called Junk
kinase (JNK). JNK has been shown to be activated though MADD and MEK 4/7 and
leads to nuclear translocation of c-Jun upon activation. C-Jun in turn is a transcriptional
activator that led to transcription of TNF, IL-1ß and CD95L mRNA. Initial experiments
using JNK Inhibitor II from Calbiochem (Darmstadt, Germany) revealed that this inhibitor
was toxic itself in higher concentrations (7-10 µM) and that the optimum range was
between 4 µM and 6 µM which was in line with other data obtained in our laboratory
(Georg Dünstl, personal communication). Incubations performed on primary murine
hepatocytes with melphalan and JNK-Inhibitor II pre-treatment (30 min, 5 µM), displayed
that cell death induced by melphalan was significantly reduced (p < 0.001), but not
totally blocked (figure 29 A). Median values calculated for that specific sets of
experiments revealed an EC50 for melphalan of about 90 µg/ml, whereas an EC50 of
about 175 µg/ml was calculated for melphalan in cells pre-treated with JNK-Inhibitor II
(30 min, 5 µM) (figure 29). The beneficial effect of JNK inhibition in melphalan-induced
cytotoxicity was also observed by light microscopy, showing an intact cellular shape of
hepatocytes up to 24 hours, with some apoptotic cells also observed in between roughly
estimated to 15 - 25%. To investigate if this inhibitory effect could also observed in
- 97 -

Results
incubations performed with melphalan and TNF in the absence of Kupffer cells, we
prepared purified hepatocytes and incubated these cells with melphalan and TNF with
and without JNK-Inhibitor II pre-treatment (30 min, 5 µM). Interestingly, the protective
effect of JNK inhibition was also be observed. EC50 values were calculated at about 75
µg/ml for melphalan/TNF alone versus 220 µg/ml under presence of JNK-Inhibitor II
(figure 29), indicating that the beneficial effect of JNK inhibition was due to effects on the
hepatocyte and not related to Kupffer cell actions.
4.13.4 Role of PARP
The idea of the so-called “Papirmeister-Hypothesis” is that a huge activation of
PARP resulting from massive nicks and breaks in the DNA might lead to a drop in
cellular NAD + -pool and therefore might contribute to apoptosis induction. Since
melphalan was shown to induce single- and double strand breaks in the DNA, we were
also interested, if this hypothesis might contribute to melphalan-induced hepatocyte
death.
Therefore, we investigated, if PARP-activation could be measured in isolated
primary hepatocytes or Kupffer cells upon stimulation with genotoxic agents. To address
this problem, we prepared purified Kupffer cells and hepatocytes, stimulated them with
H2O2 or MNNG in presence or absence of the specific PARP inhibitor PJ-34 and
- 98 -

Results
measured PARP activity using the Cellomics Array Scan (Pittsburgh, PA, USA). PARPactivation
could be detected in isolated Kupffer cells (data not shown) as well as in
hepatocytes (figure 30A) following H2O2 or MNNG treatment. Activation of PARP could
be blocked by co-incubation with PARP inhibitor PJ-34, indicating a specific
measurement of PARP. Interestingly, we found only a weak activation of PARP in
melphalan-treated cells, which could be also blocked by PJ-34 (data not shown). To
address this problem in a different setting, we incubated primary murine liver cells in
presence and absence of various concentrations of PJ-34 (ranging from 1.5 µM up to
100 µM) with melphalan. But also in this setting no influence on melphalan-induced
toxicity was measured, stressing previous results (figure 30B, data only shown for 100
µM). EC50 values were calculated in that specific set of experiments at 60 µg/ml for
melphalan versus 70 µg/ml for melphalan co-incubated with PJ-34 (figure 30B).
In summary, a weak activation of PARP was detected following melphalan
incubation, but this slight activity played no decisive role in melphalan-induced
hepatocyte death.
- 99 -

Results
4.14 Influence of melphalan on cellular redox and energy state and vice
versa
In previous projects in our laboratories the influence of cellular redox and energy
state on hepatocyte apoptosis was investigated. Briefly, it could be demonstrated that
depletion of glutathione inhibited receptor-mediated liver apoptosis in vivo (298). Under
ATP-depleted conditions it could be demonstrated that TNFR-mediated apoptosis is
blocked in vitro and in vivo in contrary to CD95-mediated apoptosis (305). Since
previous results indicated that melphalan-induced toxicity is strongly dependent on TNF,
TNFR1 and TNFR2 we were interested if modulation of cellular redox or energy state
might influence cellular outcome in response to stimulation with melphalan.
4.14.1 Melphalan and cellular redox state
Melphalan is known to react spontaneously with glutathione also without the
activity of glutathione-S-transferase. Since it was previously shown that melphalanrelated
toxicity was enhanced in patients following intracellular GSH-depletion and the
cellular redox state influences the cellular decision to die or not to die and how to die
(apoptotic or necrotic) we investigated if cellular redox state is relevant for melphalaninduced
cell death or if the pool of cellular glutathione is influenced by the drug.
To address these questions, we isolated primary hepatocytes of tnfr1r2 genedeficient
mice, purified them as described above to avoid melphalan induced cellular
toxicity and incubated them with different concentrations of melphalan. As shown in
figure 31A for concentrations ranging from 50 to 200 µg/ml melphalan the drug slightly
modifies cellular glutathione, but as compared to controls no significant depletion of GSX
could be observed after 18 hours. Interestingly, when cellular glutathione was depleted
by adding diethylmaleate (DEM) in a final concentration of 1 mM to primary liver cell
culture 30 min before melphalan incubation, melphalan-induced cellular toxicity was
significantly enhanced at concentrations higher than 25 µg/ml as compared to
incubations without DEM (figure 31B). To verify these data in the isolated perfused
mouse liver, we prepared livers of mice depleted of glutathione by an intraperitoneal
injection of 500 mg/kg phorone dissolved in 300 µl of corn oil and perfused them with
150 mg/kg melphalan. Intrahepatic depletion of glutathione also enhanced melphalan
toxicity, whereas depletion of glutathione alone by phorone application had no influence
on perfusion parameters and liver injury as indicated by lacking ALT activity in the
perfusate (figure 31C).
- 100 -

Results
In conclusion these results are in line with data obtained in the cell culture system
and negate ideas of apoptosis protection by glutathione depletion in this model.
4.14.2 Melphalan and cellular energy status
Initial experiments demonstrated melphalan induced toxicity being strongly
dependent on TNF and TNF receptors leading to hepatocyte apoptosis. Since our group
- 101 -

Results
was previously able to demonstrate that TNF-mediated hepatic apoptosis is blocked
when primary hepatocytes or mice were pre-treated with ATP-depleting carbohydrates
and cellular ATP content subsequently was depleted to about 20%. Therefore, we
investigated, if melphalan-induced hepatocyte apoptosis is also influenced by reduced
cellular energy state due to incubation with ATP-depleting carbohydrates.
To address this idea we isolated primary liver cells from C57Bl/6 mice, preincubated
them with 25 mM of ATP-depleting carbohydrates such as fructose, tagatose
or sorbitol for 30 min and added melphalan in usual concentrations to the cells.
Interestingly, under ATP-depleted conditions melphalan-induced cell death was
significantly decreased (figure 32A-C). To investigate if this effect was due to cellular
ATP depletion or possible other side-effects, e.g. osmolarity, we incubated liver cells
derived from wild type mice with different concentrations of melphalan in presence of the
non-ATP-depleting carbohydrates such as glucose, mannose or sorbitose (25 mM, t=-30
min) (figure 32D-F).
In that specific subsets of experiments we could show that cells pre-incubated
- 102 -

Results
With ATP-depleting carbohydrates were significantly protected from melphalan-induced
apoptosis, while cells pre-treated with non-ATP-depleting sugars were still sensitive for
melphalan. With these experiments we could demonstrate that ATP depletion protects
primary livers cells also when both TNF receptors are involved in apoptosis induction.
Since we could show that hepatocyte apoptosis can also be induced by melphalan
and exogenous TNF in absence of Kupffer cells, we were further interested, if
melphalan/solTNF-induced apoptosis can also be blocked by ATP depletion.
Therefore, we purified murine hepatocytes incubated them with 25 mM of fructose
for 30 min and stimulated them with melphalan and soluble TNF. As shown in figure 33,
cells were also protected significantly for apoptosis induced by melphalan and TNF.
- 103 -

Results
At this point it should be noted, that with respect to the high numbers of animal
experiments these experiments have not been verified in cells lacking TNFR1 or TNFR2
or both since it has been shown sufficiently that under Kupffer cell depleted conditions
hepatocyte death following melphalan/TNF treatment is due to activation of TNFR1 and
hepatocytes derived of animals lacking TNFR1 are not dying in response to
melphalan/TNF treatment.
Of greater importance for this project was the role of ATP depletion in the isolated
perfused mouse liver. Previous experiments in animals showed that the intrahepatic
ATP content can be lowered to about 20% 30 min after an intraperitoneal injection of 5
g/kg of tagatose, to about 40% 30 min after an intraperitoneal injection of 5 g/kg of
fructose and to about 30% 30 min after an intraperitoneal injection of 10 g/kg of fructose.
Each of these intracellular ATP concentrations has been shown to be sufficient to block
hepatocyte apoptosis in vivo due to TNF-mediated liver injury (Latta et al., personal
communication). To ensure sufficient ATP-depletion, mice were injected 30 min before
liver preparation with doses of 10 g/kg fructose or glucose as a control sugar. 30 min
after performing a closed recirculating fashion, freshly dissolved melphalan was added
in a dose of 150 mg/kg to the perfusate and perfusate samples were taken as indicated.
As expected, livers of mice injected with fructose were completely protected from
melphalan-induced organ-disruption up to nine hours (figure 34 ○), whereas livers of
mice pre-treated with glucose were still sensitive showing a mean ALT-release of 3,000
(± 1,874) U/L in the perfusate after 360 min (figure 34 ◊). Untreated livers from mice
depleted of ATP displayed only a slight toxicity after nine hours of perfusion, but this low
release of ALT might be also due to this long time of perfusion, since also untreated
control livers displayed slightly increased ALT-concentrations in the perfusate after nine
to ten hours (data not shown).
- 104 -

Results
To sum these data up, as expected from pervious data melphalan-induced
hepatocyte death could be inhibited by depletion of the intrahepatic ATP-pool. This
finding is of further interest, because protection of TNFR-mediated apoptosis could only
be demonstrated for primary cells, whole organs or living animals and not for cells or cell
lines derived from carcinomas. Therefore, it would be interesting to investigate if livers of
tumor-bearing mice show selective apoptosis of tumor cells when perfused with
melphalan under ATP-depleted conditions. We addressed this question and the results
obtained in that experiments are shown in the next section.
- 105 -

Results
4.15 Effect of melphalan on DEN-induced liver tumors in isolated liver
perfusion of tumor-bearing mice
ATP depletion by carbohydrates was only possible in primary liver cells and not in
cell lines derived from hepatic carcinoma. In addition, Adelman et al. published that
fructokinase activity was lacking in rat liver tumors (310). If this is the case, this
difference between normal healthy hepatic tissue and transformed hepatic carcinoma
cells might be useful for cancer therapy.
4.15.1 DEN induced massive tumor growth in murine livers following
intraperitoneal injection
To address the question if normal healthy liver tissue in contrary to tumorigenic
tissue might be protected from melphalan-induced apoptosis under ATP-depleting
conditions we performed tumor studies in C57Bl/6 mice induced by i.p. injection of DEN
as described in materials & methods. Tumors were grown for at least 20 weeks and
finally mice were sacrificed and tumor-bearing livers were perfused with melphalan
and/or TNF in the presence and absence of ATP-depleting carbohydrate fructose.
Mice injected with DEN displayed enhanced neoplasm in the liver (figure 35). The
tumors varied in size and numbers from small singular pre-neoplastic lesions (figure
35A+B) to solid well vascularised tumors spread over the whole organ (figure 35D-F).
Interestingly, perfusion parameters did not differ between the individual livers being
more or less affected by tumor growth. Additionally, untreated livers from tumor-bearing
mice perfused only with buffer did not show altered perfusion parameters compared to
untreated control livers of normal C57Bl/6 mice.
- 106 -

Results
- 107 -

Results
4.15.2 Isolated liver perfusion of tumor-bearing livers with melphalan/TNF
When perfused with melphalan alone livers displayed enhanced ALT release in the
perfusate, decreasing perfusate flow though the organ and finally breakdown of capillary
flow about 360 min following melphalan application (figure 36A ▲). On the contrary,
when perfused with melphalan under ATP-depleted conditions only a minor ALT release
could be detected till perfusion was stopped after 480 min not differing significantly from
control livers (figure 36A ♦, ○). Additionally, perfusate flow was stable till the end of
experiment. When perfused with melphalan and TNF, livers displayed enhanced ALT
release in the perfusate as compared to untreated control livers, resulting in breakdown
of perfusate flow about 390 min after melphalan application (figure 36B ▲). As
compared to livers perfused with melphalan alone perfusate flow was slightly enhanced
and the breakdown of capillary perfusate flow was also delayed. In Addition, there was
less ALT release detectable in the perfusate (figure 36A+B ▲). When perfused
simultaneously with melphalan and TNF under ATP-depleted conditions livers displayed
enhanced ALT release in the perfusate compared to untreated control livers, but
decreased and delayed ALT-release compared to livers perfused with melphalan and
TNF in the presence of cellular ATP (figure 36B).
- 108 -

Results
In conclusion, ATP depletion was not sufficient to protect organs completely from
melphalan/TNF-induced toxicity in this setting as this could be demonstrated in IPML of
livers derived from healthy animals. If this enhanced ALT release of ATP-depleted livers
was related to tumor cell death could not be estimated based on ALT activity in the
perfusate. This problem was further investigated by measurement of caspase-activation
in liver homogenates and histological examination of liver slices collected after the
experiment.
4.15.3 Activation of caspases in normal and neoplastic tissue
In order to investigate if healthy and tumor tissue was affected by melphalan or
melphalan combined with TNF, we analysed liver samples of normal and neoplastic
tissue collected after the experiment. Liver samples isolated from healthy tissue
displayed no significant onset of caspase activity (ctrl: 51±27; melphalan: 110±87;
melphalan/fructose: 88±42; melphalan/TNF 217±284; melphalan/TNF/fructose 90±93
pg/mlxmin). Additionally, also caspase activity detected in homogenates of isolated
neoplastic tissue differed not significantly between the individual treatment groups
(melphalan: 51±29; melphalan/fructose: 171±193; melphalan/TNF 335±413;
melphalan/TNF/fructose 100±109 pg/mlxmin). According to this results, it was not
possible conclude anything concerning the treatment-related onset of apoptosis in
normal and neoplatic tissue coout of the results obtained.
Therefore, we tried to answer the question if ATP depletion might contribute to a
selective death of cancer cells by histological examination of liver slices prepared from
each perfused mouse liver.
4.15.4 Histological examination of perfused livers derived from tumor-bearing
mice
Liver slices from perfused mouse livers were stained with H.E. and TUNEL to have
an selective marker for apoptosis. In order to grade tissue damage in these liver slices
we scored specimens on a scale from zero to five with zero being not damaged at all
and five being heavily damaged with total disruption of liver architecture and
apoptosis/necrosis been observed in most areas. Untreated control livers displayed an
intact liver architecture indicating that no or at least very minor toxicity occurred due to
- 109 -

Results
the perfusion conditions (score 0; figure 37E+F). Livers treated with melphalan alone
displayed apoptosis and necrosis in tumors as well as in normal tissue. Interestingly,
observed morphological changes varied between the individuals in this treatment group,
ranging from very minor to heavy tissue destruction (score 1-4, mostly score 2, figure
37A). But in all specimens observed toxicity was comparable between tumor and normal
tissue.
When melphalan was perfused in combination with TNF a similar situation could be
observed, but the level of toxicity was much higher (score 5, figure 37C). Toxicity
observed in liver samples of this treatment group matched very well with the early flow
stop seen during perfusion and the massive ALT release observed. But differences
between normal and tumor tissue could hardly be seen.
One of the most interesting questions was how ATP depletion would influence cell
death in normal and tumor tissue. According to the perfusion parameters, livers should
be protected in case of melphalan and partially protected in case of melphalan/TNF
treatment. Interestingly, this was not the case. Livers perfused with melphalan alone
under ATP-depleted conditions displayed also some light to moderate toxicity (score 2-3,
figure 37B), but compared to the non-ATP-depleted condition induced toxicity by
melphalan seemed to be decreased. Differences between neoplastic and normal liver
tissue could also not be observed.
In organs perfused with melphalan and TNF under ATP-depleted conditions, toxicity was
also observed in normal and tumor tissue, but tumor tissue seemed to be more affected
(tumor tissue score 3-4; normal tissue score 1-2, figure 37D). As compared to perfusions
under non-ATP-depleted conditions, the onset of cell death in tumor tissue was not
altered, but the normal surrounding tissue seemed to be more intact.
Concerning the mode of cell death it has to be mentioned that apoptosis as well
as necrosis could be observed. Especially in the presence of melphalan and TNF
necrosis was be found in tumor tissue, also affecting the vascular system. If this related
to TNF is not clear.
In conclusion, a clear statement if APT-depletion protects normal tissue from
melphalan-induced liver damage whilst keeping tumor tissue sensitive for induction of
cell death can not be made. Interestingly, when having a closer look on individual
experiments, results obtained in histological assessment of liver slices can be matched
to singular results of caspase activity and are also in line with perfusion parameters. But
- 110 -

Results
compared to other individual experiments of the same treatment schedule, differences in
the severity of tissue destruction became obvious.
- 111 -

5 Discussion
Discussion
Non-resectable cancers confined to the liver still represent a significant clinical
problem. In most of these cases, the liver is not the primary site of malignancies, but
represents the most susceptible site for metastasis. More than 140,000 new patients are
diagnosed with colorectal cancer each year in the United States and approximately more
than 25% of them develop liver metastases, of which most of them are non-resectable
(42, 408). Median survival of these patients is poor, ranging from 12 to 18 months, even
with local or systemic chemotherapy (158, 409-411). In most of these cases,
chemotherapy or a combination of surgery and chemotherapy is the only way of
therapeutic intervention. Since systemic application of therapeutics has been failed for
cancer therapy due to severe toxicity mainly affecting the gastrointestinal tract and the
immune system, several techniques have been developed for regional treatment of
cancers confined to various organs (46, 62, 91, 412, 413). For the liver, mainly two
techniques have been developed, which are commonly used in therapy today: isolated
hepatic perfusion (IHP) and hepatic artery infusion (HAI) (51, 174). These elegant
procedures have been shown to prevent systemic toxicity, due to leakage of applied
alkylating agents, or cytokines such as TNF in the circulation, and as such did not allow
systemic toxicity. However, the hepatic perfusion of melphalan, especially in
combination with TNF, has been shown to lead to a reversible hepatotoxicity in the
majority of patients, by means of a mechanism that remains elusive.
In this study, the effect of melphalan in primary murine hepatocytes, as well as in
isolated mouse liver perfusion was investigated to identify possible mechanisms leading
to cell death induced by this alkylating agent. The purpose of this study was to elucidate
mechanisms responsible for adverse reactions to melphalan, based on observations
gathered from numerous clinical studies and discussed in several reviews for being
responsible for enhanced adverse reactions. Last but not least, the knowledge
accumulated on death-receptors in the last decades opened up some new possibilities
of how to intervene with receptor-mediated apoptosis and might thus be interesting for
reducing side-effects related to cancer therapy.
- 131 -

Discussion
5.1 The role of TNF and TNF receptors in melphalan-mediated
hepatotoxicity
Hepatotoxicity, as well as hypotension, have been shown before to represent
major side-effects associated with the systemic use of TNF in cancer therapy (414).
Although TNF and CD95 ligand represent the key cytokines implicated in hepatotoxicity
and hepatocyte apoptosis. The former cytokine, the expression of which is upregulated
in many acute and chronic liver diseases, does not cause apoptosis in hepatocytes in
vitro or in vivo, unless during conditions of ischemia (415) or transcriptional/translational
inhibition (416). In contrast, the interaction between TNF and its TNFR1 has been shown
to be crucial in the priming of hepatocytes during liver regeneration (198, 417, 418) and
in the proliferation of oval cells during the neoplastic phase of liver carcinogenesis (419).
Interestingly, recent results have indicated a crucial role of TNF in classical toxicological
processes (420). Indeed, TNF was shown to be implicated in the regulation of products
inducing inflammation and fibrosis, but not in direct hepatocyte damage in carbon
tetrachloride hepatotoxicity (421). Also in alcohol-mediated liver toxicity, an important
role for TNF was recently suggested (422). Moreover, the interaction between TNF and
TNFR2 was suggested to be implicated in fumonisin hepatotoxicity in mice (423).
In this study, evidence for a novel mechanism is presented by which an
antineoplastic drug induces a significant hepatotoxicity in mice, mediated by membranebound
TNF on the surface of Kupffer cells. The importance of TNF in the observed
melphalan cytotoxicity is stressed by means of several observations: 1) Hepatocytes
lacking either TNFR1 or TNFR2 both TNF receptors were protected from melphalan
cytotoxicity. 2) Co-incubation of melphalan and TNF-neutralising anti-TNF antibodies
totally abolished melphalan-mediated hepatocellular apoptosis. 3) In absence of Kupffer
cells, identified as the source of TNF production, no hepatocyte apoptosis could be
detected. Last but not least, 4) lacking toxicity of macrophage-depleted hepatocytes
could either be restored by adding isolated Kupffer cells or by adding soluble TNF to
cellular incubations containing melphalan. With that set of experiments all four
necessary conditions have been investigated to identify TNF as the toxic mediator of
hepatocyte death following melphalan incubation: 1) loss of the TNF receptors abolished
melphalan toxicity (receptor-knock-out experiments); 2) neutralisation of the ligand
prevented melphalan toxicity (experiments with neutralising antibodies); 3) the source of
TNF was identified (experiments in absence of Kupffer cells); 4) Toxicity could be
restored when either the mediator or the natural source of the mediator was added to
the system only including hepatocytes and melphalan (TNF or Kupffer cells added to
purified hepatocytes in the presence of melphalan).
- 113 -

Discussion
Although TNFR1, that contains a death domain (218, 223), has been shown to
mediate apoptosis induced by soluble TNF in hepatocytes (212, 416), not only this
receptor type, but also TNFR2 was causally implicated in the melphalan-induced
hepatotoxicity, since livers from both tnfr1 and tnfr2 gene-deficient mice were resistant.
This means that although TNFR2 does not contain a death domain, it is nevertheless
implicated in melphalan-induced apoptosis. It was previously shown that this TNF
receptor type, as well as membrane-bound TNF were also implicated in experimental
liver injury caused by the plant lectin Con A, in inflammation and degeneration
processes in the central nervous system of transgenic mice (424-426) as well as in
experimental cerebral malaria (373). Moreover, transmembrane TNF was found to be
sufficient to mediate localised tissue toxicity and chronic inflammatory arthritis in
transgenic mice (426), as well as in concanavalin A hepatotoxicity (248). Since
membrane-bound TNF, induced by melphalan, has been suggested to preferentially
trigger TNFR2 and soluble TNF preferentially activates TNFR1, this could explain why
melphalan and soluble TNF had an additive effect in hepatotoxicity, since the activation
of TNFR2 has been reported to enhance the TNFR1-mediated apoptosis (reviewed in
(427)). Indeed, the stimulation of TNFR2 was recently shown to lead to the degradation
of TRAF2 subsequently blocking TNFR1-mediated apoptosis (260). Co-stimulation or
pre-stimulation of TNFR2 was shown to enhance caspase-8 processing, while TNFR2
was suggested to compete with TNFR1 for the recruitment of newly synthesised TRAF2bound
anti-apoptotic factors, thereby promoting the formation of a caspase-8-activating
TNFR1 complex (261). A similar co-operative signaling of both TNF receptors triggered
by transmembrane TNF was also reported in a murine model of arthritis (428).
Furthermore, studies with mutated TNF demonstrated that TNF mutants with a
selective binding capacity for TNFR1 displayed similar potency to wild type TNF in
causing cytotoxicity of a human laryngeal carcinoma-derived cell line (HEp-2) and
cytostasis in a human leukaemia cell line (U937), while exhibiting lower pro-inflammatory
activity than wild type TNF (252). Nevertheless, the role of TNFR2 in the cytocidal effect
of TNF still is a matter of debate. A study in TNF sensitive cells using broad panel of
antibodies against extracellular domains of TNFR2 suggests that TNFR2 contributes to
the cytocidal effect of TNF both by its own signaling and by regulating the access of TNF
to TNFR1 (242). This is also strengthened by studies demonstrating a severe
inflammatory syndrome involving mainly the pancreas, liver, kidney, and lung in
transgenic mice, resulting from production of the human p75TNF-R in relevant levels, as
compared to human diseases. In addition, this process was shown to evolve
independently of the presence of TNF, lymphotoxin alpha, or the TNFR1 (429).
Additionally it was found that overexpression of TNFR2 lead to TNF-dependent
apoptosis in transfected HeLa cells (257). Last but not least Grell et al. could
demonstrate in a TNF cytotoxicity model of KYM-1-derived cell line that both receptors
- 114 -

Discussion
are able to induce apoptosis, as revealed from a similar onset of DNA fragmentation and
typical morphologic criteria. This provides strong evidence for a distinct signaling
pathway of TNFR1 and TNFR2, indicating protein kinase(s)-mediated control of TNFR1
signaling and a tight linkage of TNFR2 to arachidonate metabolism (430).
But cell death is not the only cellular process involving TNFR1 and TNFR2 cooporation
(431-433). It could also be shown that ligand-independent TNFR55-mediated
co-operation takes place in TNFR75-induced granulocyte/macrophage-CSF secretion. In
contrast to ligand-dependent apoptosis in PC60 cells this co-operation is unidirectional.
According to the results of Declercq et al. this ligand-independent co-operation is
crucially dependent on an intact TNFR-associated factors 1 and 2 (TRAF1/TRAF2)binding
domain, while dominant negative TRAF2 has no inhibitory effect (258).
5.2 Poseidon’s trident - mechanism of melphalan-induced cytotoxicity
TNF was clearly identified as being the toxic principle of melphalan-induced
hepatocyte apoptosis by triggering death receptor TNFR1. But according to previous
findings on TNF-mediated apoptosis a number of questions remained open. As
documented in several publications, TNF alone failed in apoptosis induction in vitro or in
vivo unless co-administration of translational or transcriptional inhibitors such as alphaamanitin,
actinomycin D or the aminosugar galactosamine (416). Since in our
experimental setting none of these inhibitors have been added, and soluble TNF has
been induced apoptosis in purified hepatocytes in the presence but not in the absence of
melphalan, this indicated that melphalan simultaneously blocked transcription in liver
cells as has been demonstrated in HeLa and melanoma cells before (434, 435). On the
other hand, melphalan was shown to cross-link DNA in vitro (436) and is therefore likely
to be sufficient to block DNA transcription of anti-apoptotic proteins. So, melphalan fulfils
at least two jobs: stimulation of Kupffer cells and blocking transcription of cellular DNA.
In addition to these findings, another interesting observation was made. While
being unable to detect released soluble TNF upon incubations with melphalan, we
proposed a role for membrane-bound TNF that has also been shown to be biologically
active. Membrane-bound TNF mediates its biological action mainly through TNFR2
since this receptor has a much higher affinity for membrane-bound than for soluble TNF
(437). The striking role for membrane-bound TNF could be demonstrated when both
cell-types, hepatocytes and Kupffer cells, were co-incubated but spatially separated by
cell culture inserts allowing soluble TNF to pass through. In this case, no apoptosis was
induced in hepatocytes.
- 115 -

Discussion
The second result stressing this hypothesis was the observation that the study
drug was able to completely inhibit LPS-induced TNF production by Kupffer cells.
Moreover, when administered to LPS-challenged mice, also no TNF release could be
detected 90 min following LPS injection. This cytokine release was also inhibited when
melphalan was administered up to one hour following LPS challenge, demonstrating that
lack of soluble TNF was not due to transcriptional inhibition but due to inhibition of
cytokine release. Since soluble TNF has been shown to be cleaved from membranebound
TNF by means of the activity of the TNF converting enzyme (71, 72, 438) this
lead to the hypothesis that melphalan inhibits TACE activity. Indeed, we clearly
demonstrated that melphalan inhibited recombinant TACE activity in a cell-free system.
Interestingly, IC50 calculated from the obtained results fitted very well to results obtained
in cell culture experiments. The onset of hepatocyte apoptosis was observed at a
concentration of 25 to 50 µg/ml of melphalan, being exactly the concentration at which
the activity of the TACE enzyme was blocked in vitro.
However, there exist also observations arguing against this hypothesis: First,
melphalan inhibits both LPS-induced secreted and membrane-bound TNF production in
Kupffer cells, whereas TACE-inhibitors should only inhibit the former and second,
melphalan-induced membrane-bound TNF disappears already after 60 min, which would
not be the case upon inhibition of TACE. However, these findings might also be
explained by the fact that melphalan blocks efficiently transcription and translation
preventing cellular reactions due to external or internal stimuli already on DNA or mRNA
level while enhanced membrane bound TNF expression on the cellular surface in the
first two hours might result from presentation of already existing intracellular pools of
pre-formed TNF as suggested by Gordon et al. for mouse peritoneal mast cells releasing
their content upon IgE-dependent activation (79).
Melphalan on the one hand induced hepatotoxicity mediated via TNF and TNF
receptors in cell culture, but on the other hand never produced measurable amounts of
soluble TNF upon stimulation of isolated Kupffer cells or in the perfusate of isolated
livers treated with different doses of melphalan. Moreover, melphalan seemed to
abrogate TNF secretion of isolated Kupffer cells, peritoneal macrophages or isolated
livers upon stimulation with various concentrations of endotoxin. An inhibitory effect of
melphalan on the ELISA protocol has been excluded, since TNF has been detected by
the ELISA in samples containing melphalan and spiked with TNF. In our working
hypothesis we proposed that melphalan abolished TNF release by blocking the
matrixmetalloproteinase (MMP) TACE, cutting TNF off the membrane. The inhibition of
TACE by melphalan could finally be proven in a cell-free assay since the recombinant
protein and various fluorogenic substrates and inhibitors have been made commercially
available. Further more, we were also able to demonstrate lacking TNF release in vivo
- 116 -

Discussion
since mice challenged with endotoxin, SEB, or the plant lectin Con A were protected
from induced liver failure when pre-treated with 50 to 100 mg/kg of melphalan, Indicating
that this ability of TACE inhibition might also takes place in the living animal. In this
series of animal experiments we were also able to block endotoxin-induced TNF release
when melphalan was given up to one hour following LPS injection, arguing against the
fact that lack of TNF is only due to transcriptional or translational inhibition (439). Similar
findings were published by Nakama et al. for GalN/LPS induced liver injury. In this study,
the topoisomerase II inhibitor etoposide was shown to protect mice from GalN/LPS
induced liver failure when administered 50, 26 and 4 hours before GalN/LPS challenge
(440). Interestingly, this etoposide effect was also demonstrated in GalN/TNF-mediated
liver injury, which could be explained by upregulation of Bcl-xL, an anti-apoptotic protein
in the liver, in response to etoposide-treatment. In our model, we could clearly
demonstrate that low or undetectable cytokine levels in sera of mice occurred when
melphalan was co-administered, which is in contrast to the study mentioned above.
But TNF was not the only cytokine being reduced in response to melphalan.
Additionally, also IFNγ was reduced in GalN/SEB, Con A, and αCD95 challenged mice
and IL-1ß was found to be reduced in mice challenged with Con A. For the Con A model
it was described before that injection of this lectin caused increased plasma
concentrations of TNF, IL-6, IL-1, IL-2, and IFNγ, while TNF was identified as the central
mediator of Con A-induced liver injury (393, 441). Another study on Con A-induced
hepatotoxicity that investigated the role of TNF suggested Kupffer cells as the major
source of TNF produced in the liver (382). In general, the cross talk between
macrophages and T cells mostly mediated by IFNγ was found to be important for Con A
toxicity to occur (382), while TNF was identified as the cytokine being responsible for
triggering apoptosis in the hepatocyte. Therefore, our data are in line with these previous
findings. Interestingly, in the in vitro system the 26 kD membrane-bound form of TNF
was found to be crucial for the Con A induced toxicity, less the soluble 17 kD form (442),
while in vivo soluble TNF contributed to toxicity (394, 395, 441, 443). If this dependence
of both TNF receptors and the involvement of membrane-bound TNF in hepatocyte
apoptosis in vitro also contributes to the distinct morphological alterations observed in
both models remains elusive.
- 117 -

Discussion
5.3 Melphalan-induced upregulation of membrane TNF
In view of these results another question had to be answered. Melphalan was
shown to block TNF release probably by inhibition of matrix-metalloproteinase TACE in
vitro and in vivo. On the other hand, melphalan was shown to induce transcriptional
arrest in cells. The question that remained to be answered was whether melphalan also
induced an upregulation of TNF by gene activation or if only already existing pools of
pre-formed TNF were presented similar to findings mentioned by Gordon et al. (79).
Evidence for melphalan up-regulating TNF comes from observations published for
MOPC-315 tumor-bearing mice. Jovasevic and Mokyr could demonstrate that one hour
after low-dose melphalan application to mice increased mRNA levels for IFN-ß could be
detected in the spleen (352). IFN-ß in turn were shown to be necessary for upregulation
of TNF expression in MOPC-315 tumor cells. According to results obtained in cell-based
ELISA, a slight upregulation of TNF was also observed in the first two hours following
melphalan incubation, which was also confirmed in western blot analysis of RAW
stimulated with melphalan.
But also a series of previous results and recent findings argued against an active
upregulation of TNF. First of all, melphalan has been shown to react very rapidly with
DNA (436), disabling cellular response mechanisms involving transcription and
translation. This might explain why the melphalan-induced expression of membrane TNF
in Kupffer cells takes place only at early time points and subsequently fades out. This
implies that presented TNF originates from pre-formed mRNA pools present in Kupffer
cells, as reported by others (444). Second, pre-treatment of primary liver cells with
various phosphodiesterase-inhibitors one hour before adding melphalan did not show
any alteration of melphalan-induced TNF-mediated hepatocyte apoptosis (data not
shown). PDE-inhibitors have been shown to effectively block TNF upregulation of
isolated Kupffer cells in response to LPS-treatment. Third, atrial natriuretic peptide
(ANP), which has also been shown to effectively block TNF production of Kupffer cells in
response to various pathologic situations such as ischemia reperfusion injury in rat (445)
and LPS-triggered primary macrophages and RAW cells in vitro (445, 446), had also no
influence on melphalan-induced hepatotoxicity, stressing the fact that no active
upregulation of TNF takes place. On the other hand, western blot analysis of memTNF
revealed that two hours following melphalan incubation an enhanced band of
approximately 26 kD could be detected, confirming results previously obtained by cellbased
ELISA. Nevertheless, this might also be due to processing of already pre-formed
RNA pools as suggested by Ulich et al. (444).
In summary, it seems unlikely that an active upregulation of TNF in response to
melphalan takes place in our model system but this might also be a matter of melphalan
- 118 -

Discussion
concentrations used and pharmacological parameters like uptake and metabolism and
local cellular concentration.
5.4 Clinical relevance
Our findings might contribute to the fact why tumor therapy or local tumor therapy
is more efficient when melphalan is co-administered together with TNF, as compared to
treatment strategies using melphalan or TNF alone as revealed by several clinical
studies (65, 174, 447-449). On the other hand, results of one pre-clinical study
describing a lack of anti-tumor activity of human recombinant TNF, alone or in
combination with melphalan in a nude mouse human melanoma xenograft system
cannot be explained (450). On the contrary, in our model system apoptosis of primary
murine hepatocytes was also induced by melphalan and recombinant human TNF with
concentrations and kinetics comparable to murine TNF. These results are also
supported by findings of Leist et al. demonstrating TNFR1-mediated apoptosis of
primary murine hepatocytes also in response to incubation with recombinant human
TNF (212).
Apart from cellular results, the high extent of hepatotoxicity we observed in our
isolated perfused mouse liver model does not correspond to the relatively mild
hepatotoxicity reported in advanced studies in patients treated with melphalan and TNF
via hepatic perfusion. One of the reasons for this could be that this treatment procedure
frequently uses hyperthermia, i.e. a condition which could give rise to increased levels of
heat shock proteins (HSP). Recent results suggested that induction of heat shock
protein 70 in mice kept at 42°C, can protect them from the systemic toxicity of TNF,
without interfering with the tumoristatic effect in a murine melanoma tumor model (451).
Since we show in our pre-clinical setting that melphalan hepatotoxicity is mediated by
TNF, hyperthermia could thus potentially confer protection under clinical conditions.
Moreover, our observation that the melphalan-inducible membrane-bound TNF fades
out already 2 hours after melphalan treatment, could correlate with the reversible nature
of the melphalan-linked hepatotoxicity observed in patients.
To summarize this part, we could clearly demonstrate that TNF is the toxic
principle of melphalan-induced hepatocyte toxicity involving both TNF receptors.
Additionally, we could proof that there is a striking role for Kupffer cells presenting
membrane-bound TNF on their surface and that physical interaction of both cell types
(hepatocytes and Kupffer cells) is necessary to transmit TNF action to the hepatocyte,
while an active upregulation of TNF, as compared to LPS-treated macrophages of livers
seems unlikely or at least needs further investigation. Last but not least, we could
- 119 -

Discussion
demonstrate that melphalan is able to block TACE activity in vitro resulting in lack of
soluble TNF upon Kupffer cell stimulation in all models investigated and abundant liver
toxicity in various models of experimental liver failure in vivo.
5.5 Differential role of TNF receptors in the presence and absence of Kupffer
cells
According to numerous articles and reviews published on the role of TNF
receptors in apoptosis, it seems to be proven, that TNFR1 alone is sufficient to initiate a
cascade of biochemical reactions in response to binding of the specific ligand (192, 212,
230, 430). This interaction finally leads to activation of caspases and subsequently
cellular breakdown, ending up in apoptosis. Therefore it is interesting that also
numerous articles have been published demonstrating a role for both TNF receptors in
various experimental settings or pathological situations as mentioned above.
Interestingly, in most of these models a crucial role for TNFR2 was only demonstrated in
cells or animals being deficient of this receptor. The exact mechanism how TNFR2 is
involved in the mediated toxicity is currently not available.
Since membrane-bound TNF, induced by melphalan, has been suggested to
preferentially trigger TNFR2 and soluble TNF preferentially activates TNFR1, this could
explain why melphalan-mediated toxicity is dependent of TNFR2. According to this, the
activation of TNFR2 was reported to enhance the TNFR1 mediated apoptosis (reviewed
in (427)). Another study undertaken in 3T3 cells further indicated an alternative role for
TNFR2. This study revealed that although TNFR2 can significantly reduce the TNF
concentration required for cell killing, the mechanism by which this is accomplished is
not through the generation of an intracellular signal by TNFR2. Instead, TNFR2
regulates the rate of TNF association with TNFR1, possibly by increasing the local
concentration of TNF at the cell surface through rapid ligand association and
dissociation (255). This phenomenon is also known as ligand passing in literature.
As compared to findings published by Li et al, concerning TRAF2 degradation in
response to TNFR2 triggering (260) and Alexopoulou et al. concerning the positive cooperation
between both TNF receptors in murine arthritis (428) mentioned above, a
similar co-operation might occur in our model system.
When cells or organs depleted of Kupffer cells were pre-treated with melphalan
and stimulated with exogenous soluble TNF, the onset of apoptosis occured more
rapidly but with comparable extent. In this setting TNFR1 is the only necessary receptor,
clearly demonstrating that TNFR2 is not necessary for intracellular caspase activation. In
- 120 -

Discussion
contrast to the studies of Alexopoulou et al. mentioned above we did not observe a
delayed onset of injury when TNFR2 is missing, in our model system the toxicity was
totally abolished. This might be the case when TACE is inhibited by melphalan, so that
the cells were unable to release soluble TNF and transmembrane TNF was not
presented in sufficient amounts to trigger TNFR1 as demonstrated by western blot and
cell-based ELISA analysis.
This means that at least two distinct mechanisms exist. In the presence of Kupffer
cells and absence of exogenous TNF melphalan toxicity is mediated by Kupffer cellpresented
transmembrane TNF and both TNF receptors, while in the absence of Kupffer
cells and the presence of exogenous TNF only the presence of TNFR1 is a necessary
condition for inducing hepatocyte death (for illustration see figure 38). Interestingly, our
study also supports findings of Gorelik et al. who could demonstrate a striking role for
TNF production for the curative effectiveness of low-dose melphalan for mice bearing a
large MOPC-315 tumor (448).
Taken together these results might explain why melphalan in combination with
exogenous soluble TNF is more efficient in tumor therapy leading to higher regression
and response rates, as compared to the treatment strategies using melphalan alone. On
the other hand the presented data also fit to findings of higher liver toxicity observed in
patients post-surgery and combinatorial treatment strategies published by several
authors (172, 447, 452, 453) and therefore contribute to a better understanding of the
underlying mechanism.
- 121 -

Discussion
5.6 Intracellular signaling of TNF receptors following melphalan
incubation
Binding of TNF to its specific receptors is followed by a whole variety of different
cellular responses in various cell types. TNF was shown to be a key mediator in
inflammation and sepsis, immunological reactions, apoptosis, rheumatoid arthritis, liver
regeneration and so on (69, 70, 86-90). To fulfil these tasks by interaction with only two
receptors, there have to be other intracellular control centres regulating cellular
decisions upon TNF receptor activation. In the last decades numerous molecules,
pathways and transcription factors have been identified being involved in the regulation
of cellular responses upon TNF stimulation. In case of liver regeneration reactive oxygen
species, glutathione content, NFκB, STAT3 and AP1 have all been identified as major
regulators (reviewed in (198)), It was shown that autoregulation of tnf-alpha gene
transcription by selective signaling through TNFR1 requires p38 mitogen-activated
protein (MAP) kinase activity and the binding of the transcription factors ATF-2 and Jun
to the TNF-alpha cAMP-response element (CRE) promoter element. Consistent with
- 122 -

Discussion
these findings, TNFR1 engagement results in increased p38 MAP kinase activity and
p38-dependent phosphorylation of ATF-2 (454). Studies in mice deficient in NFκB
subunits have shown that this transcription factor is important for lymphocyte responses
to antigens and cytokine-inducible gene expression. In particular, the RelA (p65) subunit
is required for induction of tumor necrosis factor-alpha (TNF-alpha)-dependent genes
(455). HeLa cells transfected with TNFR2 could be shown to display enhanced
activation of NFκB and c-Jun kinase (257). Therefore, we investigated at least three of
these molecules that have been demonstrated to be involved in TNF-mediated
pathways: NFκB, MAPK and c-Jun kinase.
5.6.1 Nuclear factor κ B
Most normal and neoplastic cell types are resistant to TNF cytotoxicity unless
when co-treated with protein- or RNA-synthesis inhibitors, such as cycloheximide and
actinomycin D. Cellular resistance to TNF requires TNF receptor-associated factor 2
(TRAF2), which has been hypothesised to act mainly by mediating activation of the
transcription factors NFkB and activator protein 1 (AP1) (456). Especially NFκB p65 has
been demonstrated to be at least partly responsible for this effect by inducing a set of
genes upon TNF receptor activation that downregulate the apoptosis signal (457). In
case of TNFR1 triggering the intracellular adapter protein TRADD, which in turn interacts
with TRAF2, has been identified, leading to two distinct cellular responses, activation of
cell death and activation of NFκB (221, 222). Another group could demonstrate that
recruitment of the signal transducer FADD to the TNFR1 complex mediates apoptosis
but not NFκB activation. Two other signal transducers, RIP and TRAF2, instead mediate
NFκB activation (279). These two responses, however, seem to diverge downstream to
TRAF2 which might be mediated by MAP3K-related kinase NIK (266).
For TNFR2, also an ability of NFκB activation was reported. Of special interest
are results reporting the identification of a novel TNFR2 receptor isoform termed
icp75TNFR, which is generated by the use of an alternative transcriptional start site
within the TNFR2 gene and characterised by regulated intracellular expression. The
icp75TNFR binds to TNF and mediates intracellular signaling. Overexpression of the
icp75TNFR cDNA results in TNF-induced activation of NFκB in a TRAF2-dependent
manner (458).
Interestingly, these observations published on TNF receptor-mediated NFκB
activation are in line with findings in our model system. Since neither inhibitors of NFκB
nor cells/organs derived from inducible NFκB-k.o. mice have been displayed
morphological or biochemical changes in apoptosis induction in response to melphalan,
nor NFκB p65 translocation to the nucleus has not been altered by melphalan, we have
- 123 -

Discussion
to conclude that NFκB signaling does not play a decisive if any role in melphalan
mediated toxicity. This holds also true for incubations with melphalan and soluble TNF in
the absence of Kupffer cells. Since we have been demonstrate that TNFR2 has not
been involved in that specific setting a striking role of NFκB in the cross talk between
TNFR1 and TNFR2 under Kupffer cell-containing conditions can be excluded.
5.6.2 p38 mitogen activated protein kinase and c-Jun N-terminal kinase
The response of cells to extracellular stimuli is mediated in part by a number of
intracellular kinase and phosphatase enzymes. Within this area of research the
activation of mitogen-activated protein (MAP) kinases have been extensively described
and characterised as central components of the signal transduction pathways stimulated
by both growth factors and G protein-coupled receptor agonists. Signaling events
mediated by these kinases are fundamental to cellular functions, such as proliferation,
differentiation and cell death. More recently, homologues of the p42 and p44 isoforms of
MAP kinase have been described namely the stress-activated protein kinases (SAPKs)
or alternatively the c-jun N-terminal kinases (JNKs) and p38 MAP kinase. These MAP
kinase homologues are integral components of parallel MAP kinase cascades activated
in response to a number of cellular stresses including inflammatory cytokines (e.g.
Interleukin-1 (IL-1) and TNF), heat and chemical shock, bacterial endotoxin and
ischemia/cellular ATP depletion. These MAP kinase homologues mediate the
transduction of extracellular signals to the nucleus and are pivotal in the regulation of the
transcription events that determine functional outcome in response to such stresses
(reviewed in (459)). For renal ischemia reperfusion injury it has been shown that TNF
was released from the kidney in response to oxidants originating from renal ischemia
reperfusion injury. These oxidants in turn activated p38 MAP kinase and the TNF
transcription factor NFκB leading subsequently to TNF synthesis. In a positive feedback,
pro-inflammatory fashion, binding of TNF to specific TNF membrane receptors can
reactivate NFkB. This provides a mechanism by which TNF can upregulate its own
expression as well as facilitate the expression of other genes pivotal to the inflammatory
response (reviewed in (460, 461)). MAPKs have also been shown to take part in the
signal transduction of LPS-stimulated cell activation to produce bioactive substances
including TNF, while p38 MAPK is an important regulator of TNF expression. MAPKs
also take part in the complex network of signals induced by ionising radiation and other
cellular stresses, such as TNF and endoplasmatic reticulum (ER) stress (reviewed in
(462, 463)).
In our model system, inhibition of p38 MAPK had no significant influence on
melphalan-induced TNF-mediated hepatocyte apoptosis with caspase activity also not
- 124 -

Discussion
being altered. This fits very well into the concept that melphalan toxicity is not involving
an active upregulation of TNF, as suggested by data obtained from western blot and
cell-based ELISA experiments. Also positive feedback mechanisms as discussed for
pro-inflammatory events can be excluded. This is also in line with observations that
melphalan-induced Kupffer cell-mediated toxicity is not due to LPS contaminations of the
study drug, as excluded by Limulus amoebocyte lysate (LAL) bioassay measurements.
In view of these data it was surprising that inhibition of JNK displayed a protective
effect on melphalan-induced hepatocyte death in the presence as well as in the absence
of Kupffer cells. JNK is a stress-activated protein kinase that can be induced by
inflammatory cytokines, bacterial endotoxin, osmotic shock, UV radiation, and hypoxia.
Inhibition of JNK phosphorylation has been shown to prevent expression of the
inflammatory genes encoding for COX-2, IL-2, IFN-gamma, TNF-alpha, and prevented
the activation and differentiation of primary human CD4 cell cultures. In animal studies,
JNK inhibition blocked (bacterial) lipopolysaccharide-induced expression of tumor
necrosis factor-alpha and inhibited anti-CD3-induced apoptosis of CD4(+) CD8(+)
thymocytes (464). Studies on bile acid biosynthesis in cultured rat hepatocytes indicated
that the JNK pathway plays a pivotal role in regulating CYP7A1 levels, the rate-limiting
enzyme in the neutral pathway of bile acid biosynthesis (465). Further evidence for an
implication of JNK was shown in, fibroblast-like synoviocytes and synovium during
inflammation and tissue destruction in rheumatoid arthritis (466), in ischemia reperfusion
injury in mouse liver (467 , 468) and in apoptosis induced by DNA damage due to
stimulation with etoposide, teniposide, and UV irradiation (345). In the latter model, it
turned out that the stress-activated kinase pathway (SAPK/JNK) was required for the
maximal induction of apoptosis. If this would also be the case in melphalan-induced
hepatocyte death, this might explain why inhibition of JNK at least leads to a reduction in
hepatocyte death following melphalan incubation. This finding is also in line with the
observation that inhibition of JNK affects both mechanisms of melphalan-induced cell
death, i.e. those occurring either in the presence or in the absence of Kupffer cells.
Interestingly, TRAF2 was been demonstrated to be required for TNF-mediated activation
of c-Jun N-terminal kinase, but binding of TNF to TNFR2 induced ubiquitinylation and
proteasomal degradation of TRAF2 (260). As mentioned above, cellular resistance to
TNF required TRAF2. Inhibition of TRAF2 function(s) by signaling-deficient
oligomerisation partners or by molecules affecting TRAF2 recruitment to the TNFR1
complex completely abrogated the cytoprotective response. (456).
In conclusion this means that if TRAF2 has been degraded in response to TNFR2
stimulation, the cellular resistance to TNF has been lost and apoptosis has been
promoted. This could partially explain why TNFR2 is crucially involved in melphalanmediated
cell death since membrane TNF has been shown to play a crucial role in this
scenario and the latter has been demonstrated to bind preferentially to TNFR2. But this
- 125 -

Discussion
scenario would not explain why partial protection of hepatocytes from cell death occurs
by means of JNK-inhibition. Instead this partial protection could be explained, assuming
that TRAF2 is not degraded completely and remaining amounts being sufficient to
activate JNK. However, this was not investigated and remains speculative.
5.7 Modification of melphalan-toxicity by altered cellular redox or energy
state
5.7.1 Glutathione
The major low molecular weight thiol inside cells, the tripeptide glutathione (GSH),
is of importance for protection of the cell against oxidative challenge, for thiol
homeostasis required to guarantee basic functions, and for defence mechanisms
against xenobiotics (reviewed in (296)). In the last decades numerous examples of toxic
liver injury have been published involving intracellular depletion of GSX caused by
carbon tetrachloride, acetaminophen overdoses or allylalcohol demonstrating the
importance of an intact glutathione status (294). Intracellular depletion of GSH as such
has been shown to have little metabolic consequences unless an additional stress is
superimposed. The kinetic properties of GSH-dependent enzymes imply that loss of up
to 90% of intracellular GSH may still be compatible with cellular integrity. Mitochondrial
GSH accounting for about 10% of total cellular GSH may define the threshold beyond
which toxicity commences. Such a severe depletion of GSH has been described for
some diseases such as liver dysfunction, AIDS (295) or pulmonary fibrosis. In case of
receptor-mediated apoptosis, protection from cell death could be recently demonstrated
in case of CD95-mediated apoptosis and TNF receptor-mediated apoptosis (297).
Interestingly, this protective effect could only be shown for apoptosis while necrosis was
shown to be enhanced (298). Recently the underlying mechanism of this cytoprotective
effect could be identified, since the processing of pro-caspase-8 activation at the DISC
has been demonstrated to be dependent of an intact GSX-status (469). In another
cellular model, glutathione depletion has been shown to be associated with decreased
Bcl-2 expression and increased apoptosis (303).
Also in cancer therapy the effect of glutathione during drug treatment has been
intensively studied. For melphalan it could be shown that this molecule interacts very
rapidly with glutathione, also without the need of GSX-S-transferase (470). A study
performed on the rate of glutathione conjugation of melphalan demonstrated significant
- 126 -

Discussion
amounts of conjugation also in the absence of cellular enzymes which could not be
enhanced by addition of melanoma cell homogenates (181).
Based on the knowledge accumulated on the role of glutathione on receptormediated
apoptosis we also investigated the role of glutathione depletion in melphalaninduced
hepatocellular apoptosis. Cells or organs depleted of glutathione by DEM or
phorone displayed enhanced cytotoxicity, when incubated with even low concentrations
of melphalan, as compared to non-depleted cells or organs. These findings are in line
with enhanced gastrointestinal toxicity in mice observed upon treatment with regional
hyperthermia and melphalan following BSO-mediated glutathione depletion (176-178)
and studies obtained in U-937 human promonocytic cells (471). Conjugation of cytostatic
drugs to glutathione was also discussed as a potential mechanism of tumor cell
resistance (472), but this was negated since data from rats and humans have been
suggested that hepatic GSH conjugation plays a very minor (if any) role in the
elimination of melphalan. Therefore, it seems unlikely that modulation of GSH levels
affects the rate of elimination of this drug (182). This is also in line with our results
demonstrating that addition of melphalan in concentrations of up to 200 µg/ml to isolated
purified hepatocytes is not sufficient to deplete intracellular GSX.
In conclusion, our data obtained with melphalan under glutathione-depleted
conditions fit in very well with observations in cell-free and cellular studies as well as
with pre-clinical and clinical studies undertaken with melphalan. All experimental
systems demonstrated that melphalan is able to conjugate spontaneously to glutathione
and depletion of glutathione is leading to enhanced cytotoxicity. Therefore, depletion of
GSX is most likely not suitable for cancer therapy.
5.7.2 ATP
The cellular ATP content is another apoptosis-related parameter (473, 474) that
may be modified selectively in the liver. ATP depletion has been found to affect at least
three different pathways. First, there is evidence that ATP is needed in the formation of
the so-called apoptosome in which pro-caspase-9 is activated (238, 239, 475).
Accordingly, cellular ATP depletion frequently prevented activation of execution
caspases downstream of cytochrome c release (476, 477). Such cells with damaged
mitochondria are always destined to die, but when ATP is lacking, a switch from
apoptosis to necrosis has been observed (476-478). Second, in some models the ATPdependent
step might be upstream of cytochrome c release (476, 477), and third in
CD95-mediated apoptosis being independent of the mitochondrial pathway, ATP
depletion has no effect on caspase activation (473) or cell death (479). In studies on
- 127 -

Discussion
receptor-mediated apoptosis we could previously demonstrate, that TNF-induced liver
cell injury was completely blocked upon ATP-depletion, while CD95-mediated
hepatocyte apoptosis was enhanced (305). Interestingly, this could also been
demonstrated in vivo since ATP depletion induced by pre-treatment of mice by singular
i.p. injection of ATP-depleting carbohydrates has been prevented TNF-mediated liver
injury (305).
Since we have been demonstrated, that melphalan-induced hepatocyte death is
also dependent of TNF and TNF receptors, we investigated the effect of ATP-depleting
and non-ATP-depleting sugars in our model system. Analogous to previous data on Act
D/TNF-induced hepatocyte apoptosis we could demonstrate a protective effect of the
ATP-depleting sugars fructose and tagatose in the melphalan-model, while non-ATPdepleting
sugars like glucose, mannose or sorbose displayed no protective effect.
Interestingly, this cytoprotective effect could only be demonstrated for primary liver cells
and not for HepG2 cells and Hepa1,6 cells (data not shown), which is in line with
findings previously published for transformed cells or cells derived from hepatoma (305),
as well as primary neurones (480), and lymphoid cells (474). These results further
indicate that melphalan-induced apoptosis is only dependent of TNF and TNF receptor
mediated apoptosis, since CD95-induced apoptosis was shown to be enhanced upon
lack of intracellular ATP. Also of interest is the fact that ATP depletion protects cells or
organs also in a model which is dependent of both TNF receptors. This could previously
be demonstrated only for Con A-induced murine liver failure in vivo (Latta et al., personal
communication).
In this thesis evidence is also provided that isolated murine livers can be
protected from TNF-mediated or melphalan-induced liver injury by ATP depletion. This is
of greater interest, since this model has been closely related to the situation of patients
undergoing IHP during cancer therapy. The cytoprotective effect of ATP-depleting
carbohydrates could be demonstrated by a lacking efflux of intra-hepatic ALT in the
perfusate, by lacking caspase-activity in liver samples and in histological examination
demonstrating an intact liver morphology. Moreover, perfusion parameters like perfusate
flow through the liver were observed to be improved, at least after several hours of
perfusion.
- 128 -

Discussion
5.8 Influence of ATP depletion on melphalan/TNF-mediated organ toxicity
in tumor-bearing mice
In this study, we provide evidence that normal hepatocytes in contrast to
transformed neoplastic cells might be protected by ATP depletion from melphalaninduced
apoptosis. This idea was investigated in studies performed in isolated perfusion
of livers derived from tumor-bearing mice. Mice injected intraperitoneally with DEN
displayed massive neoplastic lesions spread over the whole liver. The treatment of mice
with DEN has been shown to produce solid, well-vascularised tumors about 20 weeks
following i.p. injection. Tumor induction in our study varied from small pre-neoplastic
lesions to huge tumors spread over the whole organ. As expected, in the absence of
TNF melphalan induced liver toxicity in tumor-bearing mice was comparable to livers
isolated from healthy mice concerning the amount and the time-course of ALT release in
the perfusate. Additionally, pre-treatment with fructose also protected hepatocytes from
melphalan-induced toxicity. These data might reflect observations in patients undergoing
IHP only being treated with melphalan alone. In these patients also low tumor response
rates and low hepatocyte toxicity have been observed (54, 57). When livers were
perfused in the presence of TNF, the observed liver toxicity was much higher and ATP
depletion by fructose pre-treatment was not able to prevent organ toxicity completely.
But this was also expected, because tumor cells should undergo apoptosis or necrosis
in this setting. When compared to the situation in the patient, high liver toxicity or
fulminant organ failure leading to death was only observed when cytostatic drugs were
combined with exogenous soluble TNF for therapy (65, 172, 174, 447). Since it could not
be determined whether ALT release in the perfusate resulted from tumor cells or healthy
hepatocytes we tried to measure caspase activity in isolated healthy liver tissue and
tumor tissue after the experiment. Unfortunately, this approach failed because resulting
values of caspase activities were not consistent in three independent assays.
Histological examination of liver slices collected after the experiment also revealed no
unified results. Interestingly, untreated control livers displayed less to no toxicity with an
overall intact liver architecture, indicating that perfusion conditions alone did not
influence tumor cell or hepatocyte death. This is also in line with previous findings of
liver slices collected from healthy mice perfused as controls for up to nine hours.
Interestingly, when tumor-bearing livers were perfused with melphalan, no obvious
differences could be observed between tumor tissue and healthy cells in most cases. In
some material, necrosis of the endothelium could be observed, but if this was due to
melphalan treatment or the mode of perfusion is not clear since this necrosis has been
also observed in some livers perfused with melphalan under ATP-depleted conditions.
When livers were depleted of ATP before melphalan perfusion, we also could not
observe an unified situation in all individuals of the same treatment group. Some of the
material displayed no apoptosis at all, some livers displayed apoptosis in healthy tissue
- 129 -

Discussion
as well as at the tumor site and in some material we could find enhanced cytotoxicity in
tumors, as compared to healthy tissue. Interestingly, this was either observed in mice
bearing huge tumors spread over the whole lobe or in one case of very low tumor
induction, bearing only very small lesions with diameters smaller than 1 mm. The latter
might be explained by the observation that very small lesions in the liver are only
connected to the hepatic artery (45) and therefore might not be very well perfused when
livers were connected via the hepatic veins. When livers were perfused in combination
with melphalan and TNF, much higher toxicity could be observed. This is also in line with
observed reduced perfusate flow though the liver following 4 hours of perfusion. But
whether this contributed to the extended toxicity, for example through changes in the
micro-circulation, is not clear. Again, mice pre-treated with fructose for ATP depletion in
this treatment group displayed slightly enhanced toxicity at the tumor site, as compared
to the healthy tissue.
In summary, according to the results obtained from the isolated liver perfusion
with melphalan/TNF under ATP-depleted and non-depleted conditions, pre-treatment of
livers/mice with fructose seemed to have a positive effect on the reduction of unwanted
side-effects by melphalan/TNF on unaffected liver tissue. But when having a closer look
by histological examination this selective protective effect between normal healthy tissue
and transformed tumor tissue is less clear. To obtain more significant results larger
studies have to be performed, also with various concentrations of used drugs and
cytokines.
In general, the understanding of the mechanism and time-course of melphalan
hepatotoxicity in the presence and the absence of exogenous TNF can thus provide a
basis for the design of clinical treatment regimens that minimise or even abrogate the
hitherto inevitable adverse effects of this therapy.
- 130 -

6 Summary
Summary
Isolated hepatic perfusion of non-resectable liver cancer using the combination of
hyperthermia, tumor necrosis factor (TNF) and melphalan can be associated with a
treatment-related hepatotoxicity. The present thesis investigated whether, apart from
TNF, also melphalan is cytotoxic in primary murine liver cells in vitro. Additionally, it was
investigated whether further mediators are involved, which cell types contribute to the
cytotoxicity and which mode of cell death takes place. The elucidation of the mechanism
of melphalan-induced liver cell destruction as proposed here may allow a rational
therapy of patients suffering from inoperable hepatic tumors or metastases.
In detail the following results were obtained:
1. Melphalan induced a concentration-dependent apoptosis in isolated murine liver
cells in vitro and isolated perfused mouse livers in situ.
2. Cells or organs derived from TNF receptor 1, TNF receptor 2 and TNF receptor 1
and 2 gene-deficient mice, isolated primary cells or organs of mice depleted of
Kupffer cells were resistant against melphalan-induced toxicity. Incubation of cells
with melphalan in presence of neutralising antiTNF antibodies abrogated
melphalan-induced apoptosis.
3. Direct contact of hepatocytes and Kupffer cells was mandatory for the observed
toxicity.
4. In absence of Kupffer cells, melphalan-induced toxicity was restored in the
presence of exogenous recombinant TNF dependent only on TNF receptor 1.
5. The activity of recombinant TNF alpha converting enzyme (TACE) was shown to
be inhibited by melphalan. In situ and in vivo, the production of soluble TNF upon
LPS-stimulation and hepatotoxicity upon treatment of mice by various proinflammatory
stimuli was blocked by melphalan.
6. After pre-treatment of cells or mice with ATP-depleting carbohydrates such as
fructose or tagatose, primary hepatocytes and isolated livers were protected from
melphalan-induced apoptosis in contrast to non-ATP-depleting sugars such as
glucose, mannose or sorbose. On the contrary, depletion of intracellular
- 131 -

Summary
glutathione by diethylmalate in vitro or phorone in the isolated liver perfusion in
situ enhanced melphalan-induced apoptosis.
7. Transformed cells or cell lines derived from hepatoma cells such as Hepa1-6 or
HepG2 cells were not protected by ATP-depletion. This fact offers the possibility
to selectively and transiently protect healthy cells against melphalan in the ATPdepleted
state.
8. Such a strategy was tested using an ex-vivo approach of tumor-bearing mice in
isolated liver perfusion in the ATP-depleted state and preliminary data were
obtained which support the concept.
- 132 -

7 Zusammenfassung
Zusammenfassung
Isolierte hepatische Perfusion mit Melphalan, Tumor Nekrose Faktor (TNF) und
Hyperthermie zur Behandlung von Patienten mit nicht-operablen Lebertumoren ist oft
mit akuter Hepatotoxizität assoziiert. In der vorliegenden Arbeit wurde die Frage
untersucht, ob Melphalan unabhängig von TNF toxisch auf primäre murine Leberzellen
wirkt. Zusätzlich wurde die Art der Schädigung, sowie beteilige Mediatoren und
Zelltypen untersucht. Die Aufklärung des Mechanismus, auf dem die Leberschädigung
induziert durch Melphalan beruht, ermöglicht eine bessere Therapie von Patienten mit
inoperablen Tumoren oder Metastasen in der Leber.
Im Einzelnen wurden folgende Ergebnisse erhalten:
1. Melphalan induzierte konzentrationsabhängig Apoptose in isolierten primären
Leberzellen in vitro und in isolierten Organperfusion in situ
2. Zellen oder Organe, die aus Mäusen mit einem genetischen Defekt im TNF
Rezeptor 1, TNF Rezeptor 2 oder beiden TNF Rezeptoren isoliert wurden,
waren vor Melphalan induzierter Schädigung geschützt. Organe, in denen die
Kupfferzellen experimentell depletiert wurden, zeigten ebenfalls keine
Melphalan-induzierte Schädigung. Coinkubation von Melphalan mit neutralisierenden
anti-TNF Antikörpern verhinderte die Melphalan-induzierte
Schädigung ebenso.
3. Ein direkter Kontakt zwischen Hepatozyt und Kupfferzelle war notwendig, um
die Melphalan-induzierte Schädigung zu vermitteln.
4. In Abwesenheit von Kupfferzellen konnte die Melphalan-induzierte
Schädigung durch Zugabe von exogenem rekombinantem TNF wieder
hergestellt werden. In diesem Falle war die induzierte Schädigung nur von
TNF Rezeptor 1 abhängig.
5. Die Aktivität von rekombinanter TNF-Convertase (TACE) wurde durch
Melphalan inhibiert. In situ und in vivo wurde die Inhibition von TACE durch
die ausbleibende TNF-Freisetzung nach Stimulation mit Endotoxin und durch
den Schutz vor TNF-vermittelter Leberschädigung nach Behandlung von
Mäusen mit pro-inflammatorischen Substanzen gezeigt.
- 133-

Zusammenfassung
6. Vorbehandlung von Zellen oder Mäusen mit ATP-depletierenden Zuckern wie
Fruktose oder Tagatose schützte primäre Hepatozyten oder isoliert
perfundierte Lebern vor Melphalan-induzierter Schädigung, während die
Vorbehandlung mit nicht-depletierenden Zuckern wie Glukose, Mannose oder
Sorbose keinen Einfluß hatte. Im Gegensatz dazu hatte eine Vorbehandlung
zur Depletion des intrazellulären Glutathionspiegels mit Diethylmaleat in vitro
oder Phoron in situ keinen protektiven Effekt, sondern erhöhte die Schädigung
durch Melphalan.
7. Transformierte Zellen oder Zelllinien, die aus Lebertumoren kultiviert wurden
wie Hepa 1-6 oder HepG2 Zellen, waren nicht durch ATP depletion geschützt.
Diese Tatsache bietet die Möglichkeit, gesunde Hepatozyten unter ATP
depletion selektiv und transient gegen Melphalan zu schützen.
8. Eine derartige Strategie wurde in einem ex-vivo Ansatz untersucht, in dem
Tumor-induzierte Mäuse in der isolierten Leberperfusion mit Melphalan und
TNF unter ATP-depletierten Bedingungen behandelt wurden. Vorläufige Daten
aus dieser Studie bestätigten grundsätzlich das Konzept.
- 134 -

8 References
References
1. Mattioli A. Kolonial- oder Vernichtungskrieg? Entgrenzte Kriegsgewalt - Der
italienische Giftgaseinsatz in Abessinien 1935-1936. Vierteljahreshefte für
Zeitgeschichte 2003;3:311-338.
2. Adair CPJ, Bagg H. Experimental and clinical studies on the treatment of cancer
by dichlorethylsulfide (mustard gas). Ann Surg 1931;93:190-199.
3. Gilman A, Philips FS. The biological actions and therapeutic applications of the
B-chloroethyl amines and sulfides. Science 1946;103:409-415.
4. Chen JJ, Sun Y, Nabel GJ. Regulation of the proinflammatory effects of Fas
ligand (CD95L). Science 1998;282:1714-1717.
5. Price CC, Gaucher GM, Koneru P, Shibakawa R, Sowa JR, Yamaguchi M.
Relative reactivities for monofunctional nitrogen mustard alkylation of nucleic acid
components. Biochim Biophys Acta 1968;166:327-359.
6. Lawley PD, Brookes P. Acidic dissociation of 7:9-dialkylguanines and its possible
relation to mutagenic properties of alkylating agents. Nature 1961;192:1081-
1082.
7. Bauer GB, Povirk LF. Specificity and kinetics of interstrand and intrastrand
bifunctional alkylation by nitrogen mustards at a G-G-C sequence. Nucleic Acids
Res 1997;25:1211-1218.
8. Henne T, Schmahl D. Occurrence of second primary malignancies in man--a
second look. Cancer Treat Rev 1985;12:77-94.
9. Wang P, Bauer GB, Bennett RA, Povirk LF. Thermolabile adenine adducts and
A.T base pair substitutions induced by nitrogen mustard analogues in an SV40based
shuttle plasmid. Biochemistry 1991;30:11515-11521.
10. Wang P, Bauer GB, Kellogg GE, Abraham DJ, Povirk LF. Effect of distamycin on
chlorambucil-induced mutagenesis in pZ189: evidence of a role for minor groove
alkylation at adenine N-3. Mutagenesis 1994;9:133-139.
11. Meier HL, Gross CL, Papirmeister B. 2,2'-Dichlorodiethyl sulfide (sulfur mustard)
decreases NAD+ levels in human leukocytes. Toxicol Lett 1987;39:109-122.
12. Szabo C, Dawson VL. Role of poly(ADP-ribose) synthetase in inflammation and
ischaemia- reperfusion. Trends Pharmacol Sci 1998;19:287-298.
13. de Murcia G, Menissier de Murcia J. Poly(ADP-ribose) polymerase: a molecular
nick-sensor. Trends Biochem Sci 1994;19:172-176.
14. Lautier D, Lagueux J, Thibodeau J, Menard L, Poirier GG. Molecular and
biochemical features of poly (ADP-ribose) metabolism. Mol Cell Biochem
1993;122:171-193.
- 135-

References
15. Sims JL, Berger SJ, Berger NA. Poly(ADP-ribose) Polymerase inhibitors preserve
nicotinamide adenine dinucleotide and adenosine 5'-triphosphate pools in DNAdamaged
cells: mechanism of stimulation of unscheduled DNA synthesis.
Biochemistry 1983;22:5188-5194.
16. Yamamoto H, Uchigata Y, Okamoto H. Streptozotocin and alloxan induce DNA
strand breaks and poly(ADP-ribose) synthetase in pancreatic islets. Nature
1981;294:284-286.
17. Burkart V, Wang ZQ, Radons J, Heller B, Herceg Z, Stingl L, Wagner EF, et al.
Mice lacking the poly(ADP-ribose) polymerase gene are resistant to pancreatic
beta-cell destruction and diabetes development induced by streptozocin. Nat Med
1999;5:314-319.
18. Charron MJ, Bonner-Weir S. Implicating PARP and NAD+ depletion in type I
diabetes. Nat Med 1999;5:269-270.
19. Bergel F, Stock JA. Cyto-active Amino-acid and Peptide Derivatives Part I.
Substituted Phenylalanins. J Chem Soc 1954:2409-2417.
20. Bergel F. Some chemical aspects of abnormal growth. Lect Sci Basis Med
1954;4:54-73.
21. Stock CC. Experimental cancer chemotherapy. Adv Cancer Res 1954;2:425-492.
22. Larionov LF, Shkodinskaja EN, Troosheikina VI, Khokhlov AS, Vasina OS,
Novikova MA. Studies on the anti-tumour activity of p-di-(2-chloroethyl)
aminophenylalanine (sarcolysine). Lancet 1955;269:169-171.
23. Schmidt LH, Fradkin R, Sullivan R, Flowers A. Comparative Pharmacology of
Alkylating Agents. II. Therapeutic Activities of Alkylating Agents and Reference
Compounds against Various Tumor Systems. Cancer Chemother Rep
1965;18:SUPPL 2:403+.
24. Schmidt LH, Fradkin R, Sullivan R, Flowers A. Comparative Pharmacology of
Alkylating Agents. I. Cancer Chemother Rep 1965;18:SUPPL 2:1-401.
25. Schmidt LH, Fradkin R, Sullivan R, Flowers A. Comparative Pharmacology of
Alkylating Agents. 3. Toxicity Data on Monkeys and Dogs. Cancer Chemother
Rep 1965;18:SUPPL 2:1017+.
26. Bosanquet AG. Stability of melphalan solutions during preparation and storage. J
Pharm Sci 1985;74:348-351.
27. Flora KP, Smith SL, Cradock JC. Application of a simple high-performance liquid
chromatographic method for the determination of melphalan in the presence of its
hydrolysis products. J Chromatogr 1979;177:91-97.
28. Goldenberg GJ, Vanstone CL, Israels LG, Iise D, Bihler I. Evidence for a
transport carrier of nitrogen mustard in nitrogen mustard-sensitive and -resistant
L5178Y lymphoblasts. Cancer Res 1970;30:2285-2291.
- 136 -

References
29. Wolpert MK, Ruddon RW. A study on the mechanism of resistance to nitrogen
mustard (HN2) in Ehrlich ascites tumor cells: comparison of uptake of HN2-14-C
into sensitive and resistant cells. Cancer Res 1969;29:873-879.
30. Furner RL, Brown RK. L-phenylalanine mustard (L-PAM): the first 25 years.
Cancer Treat Rep 1980;64:559-574.
31. Hajek R, Adam Z, Vasova I, Kral Z. Evolution of multiple myeloma treatment from
melphalan monotherapy to bone marrow transplantation. Acta Med Austriaca
1996;23:85-91.
32. Alexanian R, Haut A, Khan AU, Lane M, McKelvey EM, Migliore PJ, Stuckey WJ,
Jr., et al. Treatment for multiple myeloma. Combination chemotherapy with
different melphalan dose regimens. Jama 1969;208:1680-1685.
33. Boccadoro M, Marmont F, Tribalto M, Avvisati G, Andriani A, Barbui T, Cantonetti
M, et al. Multiple myeloma: VMCP/VBAP alternating combination chemotherapy
is not superior to melphalan and prednisone even in high-risk patients. J Clin
Oncol 1991;9:444-448.
34. Alexanian R, Barlogie B, Dixon D. High-dose glucocorticoid treatment of resistant
myeloma. Ann Intern Med 1986;105:8-11.
35. Alexanian R, Barlogie B, Tucker S. VAD-based regimens as primary treatment for
multiple myeloma. Am J Hematol 1990;33:86-89.
36. McElwain TJ, Powles RL. High-dose intravenous melphalan for plasma-cell
leukaemia and myeloma. Lancet 1983;2:822-824.
37. Barlogie B, Hall R, Zander A, Dicke K, Alexanian R. High-dose melphalan with
autologous bone marrow transplantation for multiple myeloma. Blood
1986;67:1298-1301.
38. Cunningham D, Paz-Ares L, Gore ME, Malpas J, Hickish T, Nicolson M, Meldrum
M, et al. High-dose melphalan for multiple myeloma: long-term follow-up data. J
Clin Oncol 1994;12:764-768.
39. Harousseau JL, Milpied N, Laporte JP, Collombat P, Facon T, Tigaud JD,
Casassus P, et al. Double-intensive therapy in high-risk multiple myeloma. Blood
1992;79:2827-2833.
40. Lienard D, Eggermont AM, Kroon BB, Schraffordt Koops H, Lejeune FJ. Isolated
limb perfusion in primary and recurrent melanoma: indications and results. Semin
Surg Oncol 1998;14:202-209.
41. Schraffordt Koops H, Eggermont AM, Lienard D, Kroon BB, Hoekstra HJ, Van
Geel AN, Nieweg OE, et al. Hyperthermic isolated limb perfusion for the
treatment of soft tissue sarcomas. Semin Surg Oncol 1998;14:210-214.
42. Weinreich DM, Alexander HR. Transarterial perfusion of liver metastases. Semin
Oncol 2002;29:136-144.
- 137 -

References
43. Ausman RK. Development of a technic for isolated perfusion of the liver. N Y
State J Med 1961;61:3993-3997.
44. Healey JE, Jr., Smith JL, Jr., Clark L, Jr., Stehlin JS, Jr., White EC. Hepatic tissue
tolerance to thio-TEPA administered by the isolation-perfusion technique. J Surg
Res 1961;1:111-116.
45. Breedis C, Young G. The blood supply of neoplasms in the liver. Am J Pathol
1954;30:969-977.
46. Christoforidis D, Martinet O, Lejeune FJ, Mosimann F. Isolated liver perfusion for
non-resectable liver tumours: a review. Eur J Surg Oncol 2002;28:875-890.
47. Kowal CD, Bertino JR. Possible benefits of hyperthermia to chemotherapy.
Cancer Res 1979;39:2285-2289.
48. Giovanella BC, Stehlin JS, Jr., Morgan AC. Selective lethal effect of supranormal
temperatures on human neoplastic cells. Cancer Res 1976;36:3944-3950.
49. van der Veen AH, de Wilt JH, Eggermont AM, van Tiel ST, Seynhaeve AL, ten
Hagen TL. TNF-alpha augments intratumoural concentrations of doxorubicin in
TNF-alpha-based isolated limb perfusion in rat sarcoma models and enhances
anti-tumour effects. Br J Cancer 2000;82:973-980.
50. Marinelli A, van Dierendonck JH, van Brakel GM, Irth H, Kuppen PJ, Tjaden UR,
van de Velde CJ. Increasing the effective concentration of melphalan in
experimental rat liver tumours: comparison of isolated liver perfusion and hepatic
artery infusion. Br J Cancer 1991;64:1069-1075.
51. Marinelli A, van de Velde CJ, Kuppen PJ, Franken HC, Souverijn JH, Eggermont
AM. A comparative study of isolated liver perfusion versus hepatic artery infusion
with mitomycin C in rats. Br J Cancer 1990;62:891-896.
52. Skibba JL, Condon RE. Hyperthermic isolation-perfusion in vivo of the canine
liver. Cancer 1983;51:1303-1309.
53. Skibba JL, Quebbeman EJ. Tumoricidal effects and patient survival after
hyperthermic liver perfusion. Arch Surg 1986;121:1266-1271.
54. Hafstrom LR, Holmberg SB, Naredi PL, Lindner PG, Bengtsson A, Tidebrant G,
Schersten TS. Isolated hyperthermic liver perfusion with chemotherapy for liver
malignancy. Surg Oncol 1994;3:103-108.
55. Hafstrom L, Engaras B, Holmberg SB, Gustavsson B, Jonsson PE, Lindner P,
Naredi P, et al. Treatment of liver metastases from colorectal cancer with hepatic
artery occlusion, intraportal 5-fluorouracil infusion, and oral allopurinol. A
randomized clinical trial. Cancer 1994;74:2749-2756.
56. Marinelli A, de Brauw LM, Beerman H, Keizer HJ, van Bockel JH, Tjaden UR, van
de Velde CJ. Isolated liver perfusion with mitomycin C in the treatment of
colorectal cancer metastases confined to the liver. Jpn J Clin Oncol 1996;26:341-
350.
- 138 -

References
57. Leff RS, Thompson JM, Johnson DB, Mosley KR, Daly MB, Knight WAd, Ruxer
RL, Jr., et al. Phase II trial of high-dose melphalan and autologous bone marrow
transplantation for metastatic colon carcinoma. J Clin Oncol 1986;4:1586-1591.
58. de Vries MR, Rinkes IH, van de Velde CJ, Wiggers T, Tollenaar RA, Kuppen PJ,
Vahrmeijer AL, et al. Isolated hepatic perfusion with tumor necrosis factor alpha
and melphalan: experimental studies in pigs and phase I data from humans.
Recent Results Cancer Res 1998;147:107-119.
59. Alexander HR, Brown CK, Bartlett DL, Libutti SK, Figg WD, Raje S, Turner E.
Augmented capillary leak during isolated hepatic perfusion (IHP) occurs via tumor
necrosis factor-independent mechanisms. Clin Cancer Res 1998;4:2357-2362.
60. Lienard D, Ewalenko P, Delmotte JJ, Renard N, Lejeune FJ. High-dose
recombinant tumor necrosis factor alpha in combination with interferon gamma
and melphalan in isolation perfusion of the limbs for melanoma and sarcoma. J
Clin Oncol 1992;10:52-60.
61. Lienard D, Lejeune FJ, Ewalenko P. In transit metastases of malignant
melanoma treated by high dose rTNF alpha in combination with interferongamma
and melphalan in isolation perfusion. World J Surg 1992;16:234-240.
62. Alexander HR, Jr., Fraker DL, Bartlett DL. Isolated limb perfusion for malignant
melanoma. Semin Surg Oncol 1996;12:416-428.
63. Eggermont AM, Schraffordt Koops H, Klausner JM, Kroon BB, Schlag PM,
Lienard D, van Geel AN, et al. Isolated limb perfusion with tumor necrosis factor
and melphalan for limb salvage in 186 patients with locally advanced soft tissue
extremity sarcomas. The cumulative multicenter European experience. Ann Surg
1996;224:756-764;.
64. Eggermont AM, Schraffordt Koops H, Lienard D, Kroon BB, van Geel AN,
Hoekstra HJ, Lejeune FJ. Isolated limb perfusion with high-dose tumor necrosis
factor-alpha in combination with interferon-gamma and melphalan for
nonresectable extremity soft tissue sarcomas: a multicenter trial. J Clin Oncol
1996;14:2653-2665.
65. Lindner P, Fjalling M, Hafstrom L, Kierulff-Nielsen H, Mattsson J, Schersten T,
Rizell M, et al. Isolated hepatic perfusion with extracorporeal oxygenation using
hyperthermia, tumour necrosis factor alpha and melphalan. Eur J Surg Oncol
1999;25:179-185.
66. Hoption Cann SA, van Netten JP, van Netten C. Dr William Coley and tumour
regression: a place in history or in the future. Postgrad Med J 2003;79:672-680.
67. Nauts HC, Fowler GA, Bogatko FH. A review of the influence of bacterial infection
and of bacterial products (Coley's toxins) on malignant tumors in man; a critical
analysis of 30 inoperable cases treated by Coley's mixed toxins, in which
diagnosis was confirmed by microscopic examination selected for special study.
Acta Med Scand Suppl 1953;144:1-103.
- 139 -

References
68. Carswell EA, Old LJ, Kassel RL, Green S, Fiore N, Williamson B. An endotoxininduced
serum factor that causes necrosis of tumors. Proc Natl Acad Sci U S A
1975;72:3666-3670.
69. Old LJ. Tumor necrosis factor (TNF). Science 1985;230:630-632.
70. Beutler B, Cerami A. The biology of cachectin/TNF--a primary mediator of the
host response. Annu Rev Immunol 1989;7:625-655.
71. Black RA, Rauch CT, Kozlosky CJ, Peschon JJ, Slack JL, Wolfson MF, Castner
BJ, et al. A metalloproteinase disintegrin that releases tumour-necrosis factoralpha
from cells. Nature 1997;385:729-733.
72. Black RA. Tumor necrosis factor-alpha converting enzyme. Int J Biochem Cell
Biol 2002;34:1-5.
73. Decker K. Biologically active products of stimulated liver macrophages (Kupffer
cells). Eur J Biochem 1990;192:245-261.
74. Djeu JY, Serbousek D, Blanchard DK. Release of tumor necrosis factor by
human polymorphonuclear leukocytes. Blood 1990;76:1405-1409.
75. Djeu JY. Tumor necrosis factor and Candida albicans. Behring Inst Mitt
1991:222-227.
76. Djeu JY, Matsushima K, Oppenheim JJ, Shiotsuki K, Blanchard DK. Functional
activation of human neutrophils by recombinant monocyte-derived neutrophil
chemotactic factor/IL-8. J Immunol 1990;144:2205-2210.
77. Dubravec DB, Spriggs DR, Mannick JA, Rodrick ML. Circulating human
peripheral blood granulocytes synthesize and secrete tumor necrosis factor
alpha. Proc Natl Acad Sci U S A 1990;87:6758-6761.
78. Ohno I, Tanno Y, Yamauchi K, Takishima T. Production of tumour necrosis factor
by mastocytoma P815 cells. Immunology 1990;69:312-315.
79. Gordon JR, Galli SJ. Mast cells as a source of both preformed and
immunologically inducible TNF-alpha/cachectin. Nature 1990;346:274-276.
80. Steffen M, Ottmann OG, Moore MA. Simultaneous production of tumor necrosis
factor-alpha and lymphotoxin by normal T cells after induction with IL-2 and anti-
T3. J Immunol 1988;140:2621-2624.
81. Kobayashi Y, Asada M, Osawa T. Production of lymphotoxin and tumour necrosis
factor by a T-cell hybridoma. Immunology 1987;60:213-217.
82. Brennan FM, Maini RN, Feldmann M. TNF alpha--a pivotal role in rheumatoid
arthritis? Br J Rheumatol 1992;31:293-298.
83. Shirai T, Yamaguchi H, Ito H, Todd CW, Wallace RB. Cloning and expression in
Escherichia coli of the gene for human tumour necrosis factor. Nature
1985;131:803-806.
- 140 -

References
84. Pennica D, Nedwin GE, Hayflick JS, Seeburg PH, Derynck R, Palladino MA, Kohr
WJ, et al. Human tumour necrosis factor: precursor structure, expression and
homology to lymphotoxin. Nature 1984;312:724-729.
85. Marmenout A, Fransen L, Tavernier J, Van der Heyden J, Tizard R, Kawashima
E, Shaw A, et al. Molecular cloning and expression of human tumor necrosis
factor and comparison with mouse tumor necrosis factor. Eur J Biochem
1985;152:515-522.
86. Fiers W. Tumor necrosis factor. Characterization at the molecular, cellular and in
vivo level. FEBS Lett 1991;285:199-212.
87. Camussi G, Albano E, Tetta C, Bussolino F. The molecular action of tumor
necrosis factor-alpha. Eur J Biochem 1991;202:3-14.
88. Beutler B, Cerami A. Tumor necrosis, cachexia, shock, and inflammation: a
common mediator. Annu Rev Biochem 1988;57:505-518.
89. Old LJ. Tumour necrosis factor. Polypeptide mediator network. Nature
1987;326:330-331.
90. Cerami A, Beutler B. The role of cachectin/TNF in endotoxic shock and cachexia.
Immunol Today 1988;9:28-31.
91. Lejeune FJ, Ruegg C, Lienard D. Clinical applications of TNF-alpha in cancer.
Curr Opin Immunol 1998;10:573-580.
92. Williams TW, Granger GA. Lymphocyte in vitro cytotoxicity: lymphotoxins of
several mammalian species. Nature 1968;219:1076-1077.
93. Itoh N, Yonehara S, Ishii A, Yonehara M, Mizushima S, Sameshima M, Hase A,
et al. The polypeptide encoded by the cDNA for human cell surface antigen Fas
can mediate apoptosis. Cell 1991;66:233-243.
94. Wiley SR, Schooley K, Smolak PJ, Din WS, Huang CP, Nicholl JK, Sutherland
GR, et al. Identification and characterization of a new member of the TNF family
that induces apoptosis. Immunity 1995;3:673-682.
95. Aggarwal BB. Signalling pathways of the TNF superfamily: a double-edged
sword. Nat Rev Immunol 2003;3:745-756.
96. Gray PW, Aggarwal BB, Benton CV, Bringman TS, Henzel WJ, Jarrett JA, Leung
DW, et al. Cloning and expression of cDNA for human lymphotoxin, a lymphokine
with tumour necrosis activity. Nature 1984;312:721-724.
97. Aggarwal BB, Henzel WJ, Moffat B, Kohr WJ, Harkins RN. Primary structure of
human lymphotoxin derived from 1788 lymphoblastoid cell line. J Biol Chem
1985;260:2334-2344.
98. Aggarwal BB, Moffat B, Harkins RN. Human lymphotoxin. Production by a
lymphoblastoid cell line, purification, and initial characterization. J Biol Chem
1984;259:686-691.
- 141 -

References
99. Locksley RM, Killeen N, Lenardo MJ. The TNF and TNF receptor superfamilies:
integrating mammalian biology. Cell 2001;104:487-501.
100. Aggarwal BB. Comparative analysis of the structure and function of TNF-alpha
and TNF-beta. Immunol Ser 1992;56:61-78.
101. Yonehara S, Ishii A, Yonehara M. A cell-killing monoclonal antibody (anti-Fas) to
a cell surface antigen co-downregulated with the receptor of tumor necrosis
factor. J Exp Med 1989;169:1747-1756.
102. Pitti RM, Marsters SA, Ruppert S, Donahue CJ, Moore A, Ashkenazi A. Induction
of apoptosis by Apo-2 ligand, a new member of the tumor necrosis factor
cytokine family. J Biol Chem 1996;271:12687-12690.
103. Trauth BC, Klas C, Peters AM, Matzku S, Moller P, Falk W, Debatin KM, et al.
Monoclonal antibody-mediated tumor regression by induction of apoptosis.
Science 1989;245:301-305.
104. Oehm A, Behrmann I, Falk W, Pawlita M, Maier G, Klas C, Li-Weber M, et al.
Purification and molecular cloning of the APO-1 cell surface antigen, a member of
the tumor necrosis factor/nerve growth factor receptor superfamily. Sequence
identity with the Fas antigen. J Biol Chem 1992;267:10709-10715.
105. Suda T, Takahashi T, Golstein P, Nagata S. Molecular cloning and expression of
the Fas ligand, a novel member of the tumor necrosis factor family. Cell
1993;75:1169-1178.
106. Takahashi T, Tanaka M, Inazawa J, Abe T, Suda T, Nagata S. Human Fas
ligand: gene structure, chromosomal location and species specificity. Int Immunol
1994;6:1567-1574.
107. Deushi T, Yamaguchi T, Kamiya K, Iwasaki A, Mizoguchi T, Nakayama M,
Watanabe I, et al. A new aminoglycoside antibiotic, sannamycin C and its 4-Nglycyl
derivative. J Antibiot (Tokyo) 1980;33:1274-1280.
108. Galle PR, Hofmann WJ, Walczak H, Schaller H, Otto G, Stremmel W, Krammer
PH, et al. Involvement of the CD95 (APO-1/Fas) receptor and ligand in liver
damage. J Exp Med 1995;182:1223-1230.
109. Jaeschke H. Inflammation in response to hepatocellular apoptosis. Hepatology
2002;35:964-966.
110. Pinkoski MJ, Brunner T, Green DR, Lin T. Fas and Fas ligand in gut and liver. Am
J Physiol Gastrointest Liver Physiol 2000;278:G354-G366.
111. Leist M, Gantner F, Kunstle G, Bohlinger I, Tiegs G, Bluethmann H, Wendel A.
The 55-kD tumor necrosis factor receptor and CD95 independently signal murine
hepatocyte apoptosis and subsequent liver failure. Mol Med 1996;2:109-124.
112. Takehara T, Hayashi N. Fas and fas ligand in human hepatocellular carcinoma. J
Gastroenterol 2001;36:727-728.
- 142 -

References
113. Alderson MR, Armitage RJ, Maraskovsky E, Tough TW, Roux E, Schooley K,
Ramsdell F, et al. Fas transduces activation signals in normal human T
lymphocytes. J Exp Med 1993;178:2231-2235.
114. Desbarats J, Wade T, Wade WF, Newell MK. Dichotomy between naive and
memory CD4(+) T cell responses to Fas engagement. Proc Natl Acad Sci U S A
1999;96:8104-8109.
115. Freiberg RA, Spencer DM, Choate KA, Duh HJ, Schreiber SL, Crabtree GR,
Khavari PA. Fas signal transduction triggers either proliferation or apoptosis in
human fibroblasts. J Invest Dermatol 1997;108:215-219.
116. Desbarats J, Newell MK. Fas engagement accelerates liver regeneration after
partial hepatectomy. Nat Med 2000;6:920-923.
117. Ibuki N, Yamamoto K, Yabushita K, Okano N, Okamoto R, Shimada N, Hakoda
T, et al. In situ expression of Granzyme B and Fas-ligand in the liver of viral
hepatitis. Liver 2002;22:198-204.
118. de Vries EG, Timmer T, Mulder NH, van Geelen CM, van der Graaf WT,
Spierings DC, de Hooge MN, et al. Modulation of death receptor pathways in
oncology. Drugs Today (Barc) 2003;39 Suppl C:95-109.
119. Wang XZ, Chen XC, Chen YX, Zhang LJ, Li D, Chen FL, Chen ZX, et al.
Overexpression of HBxAg in hepatocellular carcinoma and its relationship with
Fas/FasL system. World J Gastroenterol 2003;9:2671-2675.
120. Strand S, Hofmann WJ, Hug H, Muller M, Otto G, Strand D, Mariani SM, et al.
Lymphocyte apoptosis induced by CD95 (APO-1/Fas) ligand-expressing tumor
cells--a mechanism of immune evasion? Nat Med 1996;2:1361-1366.
121. Muller M, Strand S, Hug H, Heinemann EM, Walczak H, Hofmann WJ, Stremmel
W, et al. Drug-induced apoptosis in hepatoma cells is mediated by the CD95
(APO-1/Fas) receptor/ligand system and involves activation of wild-type p53. J
Clin Invest 1997;99:403-413.
122. Kischkel FC, Lawrence DA, Chuntharapai A, Schow P, Kim KJ, Ashkenazi A.
Apo2L/TRAIL-dependent recruitment of endogenous FADD and caspase-8 to
death receptors 4 and 5. Immunity 2000;12:611-620.
123. Mildner M, Eckhart L, Lengauer B, Tschachler E. Hepatocyte growth
factor/scatter factor inhibits UVB-induced apoptosis of human keratinocytes but
not of keratinocyte-derived cell lines via the phosphatidylinositol 3-kinase/AKT
pathway. J Biol Chem 2002;277:14146-14152.
124. Arbuckle E, Langlois NE, Eremin O, Heys SD. Evidence for Fas counter attack in
vivo from a study of colorectal cancer. Oncol Rep 2000;7:45-47.
125. O'Connell J, O'Sullivan GC, Collins JK, Shanahan F. The Fas counterattack: Fasmediated
T cell killing by colon cancer cells expressing Fas ligand. J Exp Med
1996;184:1075-1082.
- 143 -

References
126. Cardi G, Heaney JA, Schned AR, Ernstoff MS. Expression of Fas(APO-1/CD95)
in tumor-infiltrating and peripheral blood lymphocytes in patients with renal cell
carcinoma. Cancer Res 1998;58:2078-2080.
127. Hahne M, Rimoldi D, Schroter M, Romero P, Schreier M, French LE, Schneider
P, et al. Melanoma cell expression of Fas(Apo-1/CD95) ligand: implications for
tumor immune escape. Science 1996;274:1363-1366.
128. Pan G, O'Rourke K, Chinnaiyan AM, Gentz R, Ebner R, Ni J, Dixit VM. The
receptor for the cytotoxic ligand TRAIL. Science 1997;276:111-113.
129. Sheridan JP, Marsters SA, Pitti RM, Gurney A, Skubatch M, Baldwin D,
Ramakrishnan L, et al. Control of TRAIL-induced apoptosis by a family of
signaling and decoy receptors. Science 1997;277:818-821.
130. Walczak H, Degli-Esposti MA, Johnson RS, Smolak PJ, Waugh JY, Boiani N,
Timour MS, et al. TRAIL-R2: a novel apoptosis-mediating receptor for TRAIL.
Embo J 1997;16:5386-5397.
131. Degli-Esposti MA, Smolak PJ, Walczak H, Waugh J, Huang CP, DuBose RF,
Goodwin RG, et al. Cloning and characterization of TRAIL-R3, a novel member of
the emerging TRAIL receptor family. J Exp Med 1997;186:1165-1170.
132. Degli-Esposti MA, Dougall WC, Smolak PJ, Waugh JY, Smith CA, Goodwin RG.
The novel receptor TRAIL-R4 induces NF-kappaB and protects against TRAILmediated
apoptosis, yet retains an incomplete death domain. Immunity
1997;7:813-820.
133. Marsters SA, Sheridan JP, Pitti RM, Huang A, Skubatch M, Baldwin D, Yuan J, et
al. A novel receptor for Apo2L/TRAIL contains a truncated death domain. Curr
Biol 1997;7:1003-1006.
134. Golstein P. Cell death: TRAIL and its receptors. Curr Biol 1997;7:R750-R753.
135. Griffith TS, Lynch DH. TRAIL: a molecule with multiple receptors and control
mechanisms. Curr Opin Immunol 1998;10:559-563.
136. Sprick MR, Weigand MA, Rieser E, Rauch CT, Juo P, Blenis J, Krammer PH, et
al. FADD/MORT1 and caspase-8 are recruited to TRAIL receptors 1 and 2 and
are essential for apoptosis mediated by TRAIL receptor 2. Immunity 2000;12:599-
609.
137. Kuang AA, Diehl GE, Zhang J, Winoto A. FADD is required for DR4- and DR5mediated
apoptosis: lack of trail-induced apoptosis in FADD-deficient mouse
embryonic fibroblasts. J Biol Chem 2000;275:25065-25068.
138. Peter ME. The TRAIL DISCussion: It is FADD and caspase-8! Cell Death Differ
2000;7:759-760.
139. Holler N, Zaru R, Micheau O, Thome M, Attinger A, Valitutti S, Bodmer JL, et al.
Fas triggers an alternative, caspase-8-independent cell death pathway using the
kinase RIP as effector molecule. Nat Immunol 2000;1:489-495.
- 144 -

References
140. Walczak H, Miller RE, Ariail K, Gliniak B, Griffith TS, Kubin M, Chin W, et al.
Tumoricidal activity of tumor necrosis factor-related apoptosis-inducing ligand in
vivo. Nat Med 1999;5:157-163.
141. Tanaka S, Sugimachi K, Shirabe K, Shimada M, Wands JR. Expression and
antitumor effects of TRAIL in human cholangiocarcinoma. Hepatology
2000;32:523-527.
142. Yamanaka T, Shiraki K, Sugimoto K, Ito T, Fujikawa K, Ito M, Takase K, et al.
Chemotherapeutic agents augment TRAIL-induced apoptosis in human
hepatocellular carcinoma cell lines. Hepatology 2000;32:482-490.
143. Gores GJ, Kaufmann SH. Is TRAIL hepatotoxic? Hepatology 2001;34:3-6.
144. Mori E, Thomas M, Motoki K, Nakazawa K, Tahara T, Tomizuka K, Ishida I, et al.
Human normal hepatocytes are susceptible to apoptosis signal mediated by both
TRAIL-R1 and TRAIL-R2. Cell Death Differ 2004;11:203-207.
145. Zheng SJ, Wang P, Tsabary G, Chen YH. Critical roles of TRAIL in hepatic cell
death and hepatic inflammation. J Clin Invest 2004;113:58-64.
146. Gonzalez VM, Fuertes MA, Alonso C, Perez JM. Is cisplatin-induced cell death
always produced by apoptosis? Mol Pharmacol 2001;59:657-663.
147. Oh KW, Qian T, Brenner DA, Lemasters JJ. Salicylate enhances necrosis and
apoptosis mediated by the mitochondrial permeability transition. Toxicol Sci
2003;73:44-52.
148. Mundt B, Kuhnel F, Zender L, Paul Y, Tillmann H, Trautwein C, Manns MP, et al.
Involvement of TRAIL and its receptors in viral hepatitis. Faseb J 2003;17:94-96.
149. Higuchi H, Bronk SF, Taniai M, Canbay A, Gores GJ. Cholestasis Increases
Tumor Necrosis Factor-Related Apoptotis-Inducing Ligand (TRAIL)-R2/DR5
Expression and Sensitizes the Liver to TRAIL- Mediated Cytotoxicity. J
Pharmacol Exp Ther 2002;303:461-467.
150. Lin T, Gu J, Zhang L, Huang X, Stephens LC, Curley SA, Fang B. Targeted
expression of green fluorescent protein/tumor necrosis factor-related apoptosisinducing
ligand fusion protein from human telomerase reverse transcriptase
promoter elicits antitumor activity without toxic effects on primary human
hepatocytes. Cancer Res 2002;62:3620-3625.
151. Srivastava RK. TRAIL/Apo-2L: mechanisms and clinical applications in cancer.
Neoplasia 2001;3:535-546.
152. Nagane M, Huang HJ, Cavenee WK. The potential of TRAIL for cancer
chemotherapy. Apoptosis 2001;6:191-197.
153. Kagawa S, He C, Gu J, Koch P, Rha SJ, Roth JA, Curley SA, et al. Antitumor
activity and bystander effects of the tumor necrosis factor-related apoptosisinducing
ligand (TRAIL) gene. Cancer Res 2001;61:3330-3338.
- 145 -

References
154. Lawrence D, Shahrokh Z, Marsters S, Achilles K, Shih D, Mounho B, Hillan K, et
al. Differential hepatocyte toxicity of recombinant Apo2L/TRAIL versions. Nat
Med 2001;7:383-385.
155. Ozoren N, Kim K, Burns TF, Dicker DT, Moscioni AD, El-Deiry WS. The caspase
9 inhibitor Z-LEHD-FMK protects human liver cells while permitting death of
cancer cells exposed to tumor necrosis factor-related apoptosis-inducing ligand.
Cancer Res 2000;60:6259-6265.
156. Hugh TJ, Kinsella AR, Poston GJ. Management strategies for colorectal liver
metastases--Part I. Surg Oncol 1997;6:19-30.
157. Weiss L, Grundmann E, Torhorst J, Hartveit F, Moberg I, Eder M, Fenoglio-
Preiser CM, et al. Haematogenous metastatic patterns in colonic carcinoma: an
analysis of 1541 necropsies. J Pathol 1986;150:195-203.
158. Yoon SS, Tanabe KK. Surgical treatment and other regional treatments for
colorectal cancer liver metastases. Oncologist 1999;4:197-208.
159. Habib NA, Hodgson HJ, Lemoine N, Pignatelli M. A phase I/II study of hepatic
artery infusion with wtp53-CMV-Ad in metastatic malignant liver tumours. Hum
Gene Ther 1999;10:2019-2034.
160. Greenway B. Hepatic metastases from colorectal cancer: resection or not. Br J
Surg 1988;75:513-519.
161. Cady B, Stone MD. The role of surgical resection of liver metastases in colorectal
carcinoma. Semin Oncol 1991;18:399-406.
162. Wagner JS, Adson MA, Van Heerden JA, Adson MH, Ilstrup DM. The natural
history of hepatic metastases from colorectal cancer. A comparison with resective
treatment. Ann Surg 1984;199:502-508.
163. Oxley EM, Ellis H. Prognosis of carcinoma of the large bowel in the presence of
liver metastases. Br J Surg 1969;56:149-152.
164. Flanagan L, Jr., Foster JH. Hepatic resection for metastatic cancer. Am J Surg
1967;113:551-557.
165. Ragnhammar P, Hafstrom L, Nygren P, Glimelius B. A systematic overview of
chemotherapy effects in colorectal cancer. Acta Oncol 2001;40:282-308.
166. Hafstrom L, Holmberg SB, Lindner P, Naredi P, Asztely M, Bengtsson A, Nielsen
HK, et al. Isolated regional liver perfusion with tumour necrosis factor-alpha
followed by hyperthermic melphalan perfusion. Reg Cancer Treat. 1994;7:172-
176.
167. Ahmed N, Berridge MV. Ceramides that mediate apoptosis reduce glucose
uptake and transporter affinity for glucose in human leukaemic cell lines but not in
neutrophils. Pharmacol Toxicol 2000;86:114-121.
- 146 -

References
168. Borel Rinkes IH, de Vries MR, Jonker AM, Swaak TJ, Hack CE, Nooyen PT,
Wiggers T, et al. Isolated hepatic perfusion in the pig with TNF-alpha with and
without melphalan. Br J Cancer 1997;75:1447-1453.
169. Goto A, Shomori K, Ohkumo T, Tanaka F, Sato K, Ito H. Hyperthermia-induced
apoptosis occurs both in a p53 gene-dependent and -independent manner in
three human gastric carcinoma cell lines. Oncol Rep 1999;6:335-339.
170. Kobayashi D, Watanabe N, Yamauchi N, Okamoto T, Tsuji N, Sasaki H, Sato T,
et al. Heat-induced apoptosis via caspase-3 activation in tumour cells carrying
mutant p53. Int J Hyperthermia 2000;16:471-480.
171. van Bree C, van der Maat B, Ceha HM, Franken NA, Haveman J, Bakker PJ.
Inactivation of p53 and of pRb protects human colorectal carcinoma cells against
hyperthermia-induced cytotoxicity and apoptosis. J Cancer Res Clin Oncol
1999;125:549-555.
172. Hafstrom L, Naredi P. Isolated hepatic perfusion with extracorporeal oxygenation
using hyperthermia TNF alpha and melphalan: Swedish experience. Recent
Results Cancer Res 1998;147:120-126.
173. Oldhafer KJ, Lang H, Frerker M, Moreno L, Chavan A, Flemming P, Nadalin S, et
al. First experience and technical aspects of isolated liver perfusion for extensive
liver metastasis. Surgery 1998;123:622-631.
174. Alexander HR, Jr., Bartlett DL, Libutti SK, Fraker DL, Moser T, Rosenberg SA.
Isolated hepatic perfusion with tumor necrosis factor and melphalan for
unresectable cancers confined to the liver. J Clin Oncol 1998;16:1479-1489.
175. Plaat BE, Molenaar WM, Mastik MF, Koudstaal J, van den Berg E, Koops HS,
Hoekstra HJ. Hyperthermic isolated limb perfusion with tumor necrosis factoralpha
and melphalan in patients with locally advanced soft tissue sarcomas:
treatment response and clinical outcome related to changes in proliferation and
apoptosis. Clin Cancer Res 1999;5:1650-1657.
176. Skapek SX, VanDellen AF, McMahon DP, Postels DG, Griffith OW, Bigner DD,
Friedman HS. Melphalan-induced toxicity in nude mice following pretreatment
with buthionine sulfoximine. Cancer Chemother Pharmacol 1991;28:15-21.
177. Laskowitz DT, Elion GB, Dewhirst MW, Griffith OW, Cattley RC, Bigner DD,
Friedman HS. Enhancement of melphalan-induced gastrointestinal toxicity in
mice treated with regional hyperthermia and BSO-mediated glutathione depletion.
Int J Hyperthermia 1992;8:111-120.
178. Smith AC, Liao JT, Page JG, Wientjes MG, Grieshaber CK. Pharmacokinetics of
buthionine sulfoximine (NSC 326231) and its effect on melphalan-induced toxicity
in mice. Cancer Res 1989;49:5385-5391.
179. Kramer RA, Schuller HM, Smith AC, Boyd MR. Effects of buthionine sulfoximine
on the nephrotoxicity of 1-(2-chloroethyl)-3-(trans-4-methylcyclohexyl)-1nitrosourea
(MeCCNU). J Pharmacol Exp Ther 1985;234:498-506.
- 147 -

References
180. Vahrmeijer AL, van Dierendonck JH, Schutrups J, van de Velde CJ, Mulder GJ.
Effect of glutathione depletion on inhibition of cell cycle progression and induction
of apoptosis by melphalan (L-phenylalanine mustard) in human colorectal cancer
cells. Biochem Pharmacol 1999;58:655-664.
181. Guenthner TM, Whalen R, Jevtovic-Todorovic V. Direct measurement of
melphalan conjugation with glutathione: studies with human melanoma cells and
mammalian liver. J Pharmacol Exp Ther 1992;260:1331-1336.
182. Vahrmeijer AL, Snel CA, Steenvoorden DP, Beijnen JH, Pang KS, Schutrups J,
Tirona R, et al. Lack of glutathione conjugation of melphalan in the isolated in situ
liver perfusion in humans. Cancer Res 1996;56:4709-4714.
183. Virchow R. Cellular Pathology as based upon Physiological and Pathological
Histology. Dover Publications, New York 1991:356-382.
184. Glücksmann A. Cell death in normal vertebrate ontogeny. Biol Rev 1951;26:59-
86.
185. Ellis RE, Yuan JY, Horvitz HR. Mechanisms and functions of cell death. Annu
Rev Cell Biol 1991;7:663-698.
186. Saunders JW, Jr. Death in embryonic systems. Science 1966;154:604-612.
187. Kerr JF, Wyllie AH, Currie AR. Apoptosis: a basic biological phenomenon with
wide-ranging implications in tissue kinetics. Br J Cancer 1972;26:239-257.
188. Patel T, Gores GJ. Apoptosis and hepatobiliary disease. Hepatology
1995;21:1725-1741.
189. Majno G, Joris I. Apoptosis, oncosis, and necrosis. An overview of cell death. Am
J Pathol 1995;146:3-15.
190. Schwartzman RA, Cidlowski JA. Apoptosis: the biochemistry and molecular
biology of programmed cell death. Endocr Rev 1993;14:133-151.
191. Thompson CB. Apoptosis in the pathogenesis and treatment of disease. Science
1995;267:1456-1462.
192. Chen G, Goeddel DV. TNF-R1 signaling: a beautiful pathway. Science
2002;296:1634-1635.
193. Krammer PH. CD95(APO-1/Fas)-mediated apoptosis: live and let die. Adv
Immunol 1999;71:163-210.
194. Wajant H. The Fas signaling pathway: more than a paradigm. Science
2002;296:16358-11636.
195. Danial NN, Korsmeyer SJ. Cell death: critical control points. Cell 2004;116:205-
219.
- 148 -

References
196. Earnshaw WC, Martins LM, Kaufmann SH. Mammalian caspases: structure,
activation, substrates, and functions during apoptosis. Annu Rev Biochem
1999;68:383-424.
197. Siegel RM, Muppidi J, Roberts M, Porter M, Wu Z. Death receptor signaling and
autoimmunity. Immunol Res 2003;27:499-512.
198. Fausto N. Liver regeneration. J Hepatol 2000;32:19-31.
199. Hengartner MO. The biochemistry of apoptosis. Nature 2000;407:770-776.
200. Cohen GM. Caspases: the executioners of apoptosis. Biochem J 1997;326:1-16.
201. Galle PR. Apoptosis in liver disease. J Hepatol 1997;27:405-412.
202. Fleisher TA. Apoptosis. Ann Allergy Asthma Immunol 1997;78:245-249.
203. Hall AG. Review: The role of glutathione in the regulation of apoptosis. Eur J Clin
Invest 1999;29:238-245.
204. Wang TH, Wang HS. Apoptosis: (1). Overview and clinical significance. J Formos
Med Assoc 1999;98:381-393.
205. Wang TH, Wang HS. Apoptosis: (2) characteristics of apoptosis. J Formos Med
Assoc 1999;98:531-542.
206. Williams G. SC, Grimes E., McCarthy N. Apoptosis: final control point in cell
biology. Trends in cell biology 1992;2:263-267.
207. Chao DT, Korsmeyer SJ. BCL-2 family: regulators of cell death. Annu Rev
Immunol 1998;16:395-419.
208. Pettmann B, Henderson CE. Neuronal cell death. Neuron 1998;20:633-647.
209. Green DR. Apoptotic pathways: the roads to ruin. Cell 1998;94:695-698.
210. Cohen JJ. Apoptosis. Immunol Today 1993;14:126-130.
211. Leong KG, Karsan A. Signaling pathways mediated by tumor necrosis factor
alpha. Histol Histopathol 2000;15:1303-1325.
212. Leist M, Gantner F, Jilg S, Wendel A. Activation of the 55 kDa TNF receptor is
necessary and sufficient for TNF-induced liver failure, hepatocyte apoptosis, and
nitrite release. J Immunol 1995;154:1307-1316.
213. Laster SM, Wood JG, Gooding LR. Tumor necrosis factor can induce both
apoptic and necrotic forms of cell lysis. J Immunol 1988;141:2629-2634.
214. Itoh N, Nagata S. A novel protein domain required for apoptosis. Mutational
analysis of human Fas antigen. J Biol Chem 1993;268:10932-10937.
215. Natoli G, Costanzo A, Guido F, Moretti F, Levrero M. Apoptotic, non-apoptotic,
and anti-apoptotic pathways of tumor necrosis factor signalling. Biochem
Pharmacol 1998;56:915-920.
- 149 -

References
216. Sidoti-de Fraisse C, Rincheval V, Risler Y, Mignotte B, Vayssiere JL. TNF-alpha
activates at least two apoptotic signaling cascades. Oncogene 1998;17:1639-
1651.
217. Tartaglia LA, Rothe M, Hu YF, Goeddel DV. Tumor necrosis factor's cytotoxic
activity is signaled by the p55 TNF receptor. Cell 1993;73:213-216.
218. Tartaglia LA, Ayres TM, Wong GH, Goeddel DV. A novel domain within the 55 kd
TNF receptor signals cell death. Cell 1993;74:845-853.
219. Jiang Y, Woronicz JD, Liu W, Goeddel DV. Prevention of constitutive TNF
receptor 1 signaling by silencer of death domains. Science 1999;283:543-546.
220. Tschopp J, Martinon F, Hofmann K. Apoptosis: Silencing the death receptors.
Curr Biol 1999;9:R381-384.
221. Hsu H, Xiong J, Goeddel DV. The TNF receptor 1-associated protein TRADD
signals cell death and NF-kappa B activation. Cell 1995;81:495-504.
222. Hsu H, Shu HB, Pan MG, Goeddel DV. TRADD-TRAF2 and TRADD-FADD
interactions define two distinct TNF receptor 1 signal transduction pathways. Cell
1996;84:299-308.
223. Boldin MP, Varfolomeev EE, Pancer Z, Mett IL, Camonis JH, Wallach D. A novel
protein that interacts with the death domain of Fas/APO1 contains a sequence
motif related to the death domain. J Biol Chem 1995;270:7795-7798.
224. Chinnaiyan AM, O'Rourke K, Tewari M, Dixit VM. FADD, a novel death domaincontaining
protein, interacts with the death domain of Fas and initiates apoptosis.
Cell 1995;81:505-512.
225. Muzio M, Chinnaiyan AM, Kischkel FC, O'Rourke K, Shevchenko A, Ni J, Scaffidi
C, et al. FLICE, a novel FADD-homologous ICE/CED-3-like protease, is recruited
to the CD95 (Fas/APO-1) death--inducing signaling complex. Cell 1996;85:817-
827.
226. Hofmann K, Bucher P, Tschopp J. The CARD domain: a new apoptotic signalling
motif. Trends Biochem Sci 1997;22:155-156.
227. Medema JP, Scaffidi C, Kischkel FC, Shevchenko A, Mann M, Krammer PH,
Peter ME. FLICE is activated by association with the CD95 death-inducing
signaling complex (DISC). Embo J 1997;16:2794-2804.
228. Wajant H, Johannes FJ, Haas E, Siemienski K, Schwenzer R, Schubert G, Weiss
T, et al. Dominant-negative FADD inhibits TNFR60-, Fas/Apo1- and TRAIL-
R/Apo2-mediated cell death but not gene induction. Curr Biol 1998;8:113-116.
229. Muzio M, Stockwell BR, Stennicke HR, Salvesen GS, Dixit VM. An induced
proximity model for caspase-8 activation. J Biol Chem 1998;273:2926-2930.
230. Rath PC, Aggarwal BB. TNF-induced signaling in apoptosis. J Clin Immunol
1999;19:350-364.
- 150 -

References
231. Pan G, O'Rourke K, Dixit VM. Caspase-9, Bcl-XL, and Apaf-1 form a ternary
complex. J Biol Chem 1998;273:5841-5845.
232. Srinivasula SM, Hegde R, Saleh A, Datta P, Shiozaki E, Chai J, Lee RA, et al. A
conserved XIAP-interaction motif in caspase-9 and Smac/DIABLO regulates
caspase activity and apoptosis. Nature 2001;410:112-116.
233. Shi Y. A structural view of mitochondria-mediated apoptosis. Nat Struct Biol
2001;8:394-401.
234. Joza N, Susin SA, Daugas E, Stanford WL, Cho SK, Li CY, Sasaki T, et al.
Essential role of the mitochondrial apoptosis-inducing factor in programmed cell
death. Nature 2001;410:549-554.
235. Harris MH, Thompson CB. The role of the Bcl-2 family in the regulation of outer
mitochondrial membrane permeability. Cell Death Differ 2000;7:1182-1191.
236. Li H, Zhu H, Xu CJ, Yuan J. Cleavage of BID by caspase 8 mediates the
mitochondrial damage in the Fas pathway of apoptosis. Cell 1998;94:491-501.
237. Zou H, Henzel WJ, Liu X, Lutschg A, Wang X. Apaf-1, a human protein
homologous to C. elegans CED-4, participates in cytochrome c-dependent
activation of caspase-3. Cell 1997;90:405-413.
238. Srinivasula SM, Ahmad M, Fernandes-Alnemri T, Alnemri ES. Autoactivation of
procaspase-9 by Apaf-1-mediated oligomerization. Mol Cell 1998;1:949-957.
239. Li P, Nijhawan D, Budihardjo I, Srinivasula SM, Ahmad M, Alnemri ES, Wang X.
Cytochrome c and dATP-dependent formation of Apaf-1/caspase-9 complex
initiates an apoptotic protease cascade. Cell 1997;91:479-489.
240. Budihardjo I, Oliver H, Lutter M, Luo X, Wang X. Biochemical pathways of
caspase activation during apoptosis. Annu Rev Cell Dev Biol 1999;15:269-290.
241. Lorenzo HK, Susin SA, Penninger J, Kroemer G. Apoptosis inducing factor (AIF):
a phylogenetically old, caspase-independent effector of cell death. Cell Death
Differ 1999;6:516-524.
242. Bigda J, Beletsky I, Brakebusch C, Varfolomeev Y, Engelmann H, Holtmann H,
Wallach D. Dual role of the p75 tumor necrosis factor (TNF) receptor in TNF
cytotoxicity. J Exp Med 1994;180:445-460.
243. Medvedev AE, Sundan A, Espevik T. Involvement of the tumor necrosis factor
receptor p75 in mediating cytotoxicity and gene regulating activities. Eur J
Immunol 1994;24:2842-2849.
244. Heller RA, Song K, Fan N, Chang DJ. The p70 tumor necrosis factor receptor
mediates cytotoxicity. Cell 1992;70:47-56.
245. Lucas R, Garcia I, Donati YR, Hribar M, Mandriota SJ, Giroud C, Buurman WA,
et al. Both TNF receptors are required for direct TNF-mediated cytotoxicity in
microvascular endothelial cells. Eur J Immunol 1998;28:3577-3586.
- 151 -

References
246. Zheng L, Fisher G, Miller RE, Peschon J, Lynch DH, Lenardo MJ. Induction of
apoptosis in mature T cells by tumour necrosis factor. Nature 1995;377:348-351.
247. Shalaby MR, Sundan A, Loetscher H, Brockhaus M, Lesslauer W, Espevik T.
Binding and regulation of cellular functions by monoclonal antibodies against
human tumor necrosis factor receptors. J Exp Med 1990;172:1517-1520.
248. Küsters S, Tiegs G, Alexopoulou L, Pasparakis M, Douni E, Kunstle G,
Bluethmann H, et al. In vivo evidence for a functional role of both tumor necrosis
factor (TNF) receptors and transmembrane TNF in experimental hepatitis. Eur J
Immunol 1997;27:2870-2875.
249. Künstle G, Hentze H, Germann PG, Tiegs G, Meergans T, Wendel A.
Concanavalin A hepatotoxicity in mice: tumor necrosis factor-mediated organ
failure independent of caspase-3-like protease activation. Hepatology
1999;30:1241-1251.
250. Collawn JF, Stangel M, Kuhn LA, Esekogwu V, Jing SQ, Trowbridge IS, Tainer
JA. Transferrin receptor internalization sequence YXRF implicates a tight turn as
the structural recognition motif for endocytosis. Cell 1990;63:1061-1072.
251. Espevik T, Brockhaus M, Loetscher H, Nonstad U, Shalaby R. Characterization
of binding and biological effects of monoclonal antibodies against a human tumor
necrosis factor receptor. J Exp Med 1990;171:415-426.
252. Barbara JA, Smith WB, Gamble JR, Van Ostade X, Vandenabeele P, Tavernier
J, Fiers W, et al. Dissociation of TNF-alpha cytotoxic and proinflammatory
activities by p55 receptor- and p75 receptor-selective TNF-alpha mutants. Embo
J 1994;13:843-850.
253. Van Ostade X, Vandenabeele P, Tavernier J, Fiers W. Human tumor necrosis
factor mutants with preferential binding to and activity on either the R55 or R75
receptor. Eur J Biochem 1994;220:771-779.
254. Loetscher H, Stueber D, Banner D, Mackay F, Lesslauer W. Human tumor
necrosis factor alpha (TNF alpha) mutants with exclusive specificity for the 55kDa
or 75-kDa TNF receptors. J Biol Chem 1993;268:26350-26357.
255. Tartaglia LA, Pennica D, Goeddel DV. Ligand passing: the 75-kDa tumor
necrosis factor (TNF) receptor recruits TNF for signaling by the 55-kDa TNF
receptor. J Biol Chem 1993;268:18542-18548.
256. Vandenabeele P, Declercq W, Beyaert R, Fiers W. Two tumour necrosis factor
receptors: structure and function. Trends Cell Biol 1995;5:392-399.
257. Haridas V, Darnay BG, Natarajan K, Heller R, Aggarwal BB. Overexpression of
the p80 TNF receptor leads to TNF-dependent apoptosis, nuclear factor-kappa B
activation, and c-Jun kinase activation. J Immunol 1998;160:3152-3162.
258. Declercq W, Denecker G, Fiers W, Vandenabeele P. Cooperation of both TNF
receptors in inducing apoptosis: involvement of the TNF receptor-associated
factor binding domain of the TNF receptor 75. J Immunol 1998;161:390-399.
- 152 -

References
259. Weiss T, Grell M, Siemienski K, Muhlenbeck F, Durkop H, Pfizenmaier K,
Scheurich P, et al. TNFR80-dependent enhancement of TNFR60-induced cell
death is mediated by TNFR-associated factor 2 and is specific for TNFR60. J
Immunol 1998;161:3136-3142.
260. Li X, Yang Y, Ashwell JD. TNF-RII and c-IAP1 mediate ubiquitination and
degradation of TRAF2. Nature 2002;416:345-347.
261. Fotin-Mleczek M, Henkler F, Samel D, Reichwein M, Hausser A, Parmryd I,
Scheurich P, et al. Apoptotic crosstalk of TNF receptors: TNF-R2-induces
depletion of TRAF2 and IAP proteins and accelerates TNF-R1-dependent
activation of caspase-8. J Cell Sci 2002;115:2757-2770.
262. Rothe M, Wong SC, Henzel WJ, Goeddel DV. A novel family of putative signal
transducers associated with the cytoplasmic domain of the 75 kDa tumor
necrosis factor receptor. Cell 1994;78:681-692.
263. Laegreid A, Medvedev A, Nonstad U, Bombara MP, Ranges G, Sundan A,
Espevik T. Tumor necrosis factor receptor p75 mediates cell-specific activation of
nuclear factor kappa B and induction of human cytomegalovirus enhancer. J Biol
Chem 1994;269:7785-7791.
264. Lin Y, Devin A, Rodriguez Y, Liu ZG. Cleavage of the death domain kinase RIP
by caspase-8 prompts TNF-induced apoptosis. Genes Dev 1999;13:2514-2526.
265. Song HY, Regnier CH, Kirschning CJ, Goeddel DV, Rothe M. Tumor necrosis
factor (TNF)-mediated kinase cascades: bifurcation of nuclear factor-kappaB and
c-jun N-terminal kinase (JNK/SAPK) pathways at TNF receptor-associated factor
2. Proc Natl Acad Sci U S A 1997;94:9792-9796.
266. Malinin NL, Boldin MP, Kovalenko AV, Wallach D. MAP3K-related kinase
involved in NF-kappaB induction by TNF, CD95 and IL-1. Nature 1997;385:540-
544.
267. Woronicz JD, Gao X, Cao Z, Rothe M, Goeddel DV. IkappaB kinase-beta: NFkappaB
activation and complex formation with IkappaB kinase-alpha and NIK.
Science 1997;278:866-869.
268. DiDonato JA, Hayakawa M, Rothwarf DM, Zandi E, Karin M. A cytokineresponsive
IkappaB kinase that activates the transcription factor NF-kappaB.
Nature 1997;388:548-554.
269. Mercurio F, Zhu H, Murray BW, Shevchenko A, Bennett BL, Li J, Young DB, et al.
IKK-1 and IKK-2: cytokine-activated IkappaB kinases essential for NF-kappaB
activation. Science 1997;278:860-866.
270. Regnier CH, Song HY, Gao X, Goeddel DV, Cao Z, Rothe M. Identification and
characterization of an IkappaB kinase. Cell 1997;90:373-383.
271. Baldwin AS, Jr. The NF-kappa B and I kappa B proteins: new discoveries and
insights. Annu Rev Immunol 1996;14:649-683.
- 153 -

References
272. Karin M, Ben-Neriah Y. Phosphorylation meets ubiquitination: the control of NF-
[kappa]B activity. Annu Rev Immunol 2000;18:621-663.
273. Ghosh S, Karin M. Missing pieces in the NF-kappaB puzzle. Cell 2002;109
Suppl:S81-96.
274. Vig E, Green M, Liu Y, Donner DB, Mukaida N, Goebl MG, Harrington MA.
Modulation of tumor necrosis factor and interleukin-1-dependent NF-kappaB
activity by mPLK/IRAK. J Biol Chem 1999;274:13077-13084.
275. Feinstein E, Kimchi A, Wallach D, Boldin M, Varfolomeev E. The death domain: a
module shared by proteins with diverse cellular functions. Trends Biochem Sci
1995;20:342-344.
276. Trofimova M, Sprenkle AB, Green M, Sturgill TW, Goebl MG, Harrington MA.
Developmental and tissue-specific expression of mouse pelle-like protein kinase.
J Biol Chem 1996;271:17609-17612.
277. Natoli G, Costanzo A, Ianni A, Templeton DJ, Woodgett JR, Balsano C, Levrero
M. Activation of SAPK/JNK by TNF receptor 1 through a noncytotoxic TRAF2dependent
pathway. Science 1997;275:200-203.
278. Minden A, Lin A, McMahon M, Lange-Carter C, Derijard B, Davis RJ, Johnson
GL, et al. Differential activation of ERK and JNK mitogen-activated protein
kinases by Raf-1 and MEKK. Science 1994;266:1719-1723.
279. Liu ZG, Hsu H, Goeddel DV, Karin M. Dissection of TNF receptor 1 effector
functions: JNK activation is not linked to apoptosis while NF-kappaB activation
prevents cell death. Cell 1996;87:565-576.
280. Reinhard C, Shamoon B, Shyamala V, Williams LT. Tumor necrosis factor alphainduced
activation of c-jun N-terminal kinase is mediated by TRAF2. Embo J
1997;16:1080-1092.
281. Wajant H, Scheurich P. Tumor necrosis factor receptor-associated factor (TRAF)
2 and its role in TNF signaling. Int J Biochem Cell Biol 2001;33:19-32.
282. Deveraux QL, Reed JC. IAP family proteins--suppressors of apoptosis. Genes
Dev 1999;13:239-252.
283. Deveraux QL, Roy N, Stennicke HR, Van Arsdale T, Zhou Q, Srinivasula SM,
Alnemri ES, et al. IAPs block apoptotic events induced by caspase-8 and
cytochrome c by direct inhibition of distinct caspases. Embo J 1998;17:2215-
2223.
284. Deveraux QL, Stennicke HR, Salvesen GS, Reed JC. Endogenous inhibitors of
caspases. J Clin Immunol 1999;19:388-398.
285. Nishitoh H, Saitoh M, Mochida Y, Takeda K, Nakano H, Rothe M, Miyazono K, et
al. ASK1 is essential for JNK/SAPK activation by TRAF2. Mol Cell 1998;2:389-
395.
- 154 -

References
286. Derijard B, Raingeaud J, Barrett T, Wu IH, Han J, Ulevitch RJ, Davis RJ.
Independent human MAP-kinase signal transduction pathways defined by MEK
and MKK isoforms. Science 1995;267:682-685.
287. Kyriakis JM, Avruch J. Protein kinase cascades activated by stress and
inflammatory cytokines. Bioessays 1996;18:567-577.
288. Winston BW, Chan ED, Johnson GL, Riches DW. Activation of p38mapk, MKK3,
and MKK4 by TNF-alpha in mouse bone marrow-derived macrophages. J
Immunol 1997;159:4491-4497.
289. Stokoe D, Stephens LR, Copeland T, Gaffney PR, Reese CB, Painter GF,
Holmes AB, et al. Dual role of phosphatidylinositol-3,4,5-trisphosphate in the
activation of protein kinase B. Science 1997;277:567-570.
290. Ludwig S, Engel K, Hoffmeyer A, Sithanandam G, Neufeld B, Palm D, Gaestel M,
et al. 3pK, a novel mitogen-activated protein (MAP) kinase-activated protein
kinase, is targeted by three MAP kinase pathways. Mol Cell Biol 1996;16:6687-
6697.
291. Heinrich M, Neumeyer J, Jakob M, Hallas C, Tchikov V, Winoto-Morbach S,
Wickel M, et al. Cathepsin D links TNF-induced acid sphingomyelinase to Bidmediated
caspase-9 and -3 activation. Cell Death Differ 2004;11:550-563.
292. Anderson ME. Glutathione and glutathione delivery compounds. Adv Pharmacol
1997;38:65-78.
293. Meister A. On the antioxidant effects of ascorbic acid and glutathione. Biochem
Pharmacol 1992;44:1905-1915.
294. Jones TW, Thor H, Orrenius S. Cellular defense mechanisms against toxic
substances. Arch Toxicol Suppl 1986;9:259-271.
295. Droge W, Schulze-Osthoff K, Mihm S, Galter D, Schenk H, Eck HP, Roth S, et al.
Functions of glutathione and glutathione disulfide in immunology and
immunopathology. Faseb J 1994;8:1131-1138.
296. Uhlig S, Wendel A. The physiological consequences of glutathione variations.
Life Sci 1992;51:1083-1094.
297. Hentze H, Kunstle G, Volbracht C, Ertel W, Wendel A. CD95-Mediated murine
hepatic apoptosis requires an intact glutathione status. Hepatology 1999;30:177-
185.
298. Hentze H, Gantner F, Kolb SA, Wendel A. Depletion of hepatic glutathione
prevents death receptor-dependent apoptotic and necrotic liver injury in mice. Am
J Pathol 2000;156:2045-2056.
299. McCloskey P, Edwards RJ, Tootle R, Selden C, Roberts E, Hodgson HJ.
Resistance of three immortalized human hepatocyte cell lines to acetaminophen
and N-acetyl-p-benzoquinoneimine toxicity. J Hepatol 1999;31:841-851.
- 155 -

References
300. Abdel Rahim AG, Arthur JR, Mills CF. Effects of dietary copper, cadmium, iron,
molybdenum and manganese on selenium utilization by the rat. J Nutr
1986;116:403-411.
301. Chuang JI, Chang TY, Liu HS. Glutathione depletion-induced apoptosis of Haras-transformed
NIH3T3 cells can be prevented by melatonin. Oncogene
2003;22:1349-1357.
302. He YY, Huang JL, Ramirez DC, Chignell CF. Role of reduced glutathione efflux in
apoptosis of immortalized human keratinocytes induced by UVA. J Biol Chem
2003;278:8058-8064.
303. Celli A, Que FG, Gores GJ, LaRusso NF. Glutathione depletion is associated with
decreased Bcl-2 expression and increased apoptosis in cholangiocytes. Am J
Physiol 1998;275:G749-G757.
304. Chen Q, Galleano M, Cederbaum AI. Cytotoxicity and apoptosis produced by
arachidonic acid in HepG2 cells overexpressing human cytochrome P-4502E1.
Alcohol Clin Exp Res 1998;22:782-784.
305. Latta M, Kunstle G, Leist M, Wendel A. Metabolic depletion of ATP by fructose
inversely controls CD95- and tumor necrosis factor receptor 1-mediated hepatic
apoptosis. J Exp Med 2000;191:1975-1986.
306. Latta M, Kresse M, Künstle G, Lucas R, Wendel A. Models of cytokine-induced
hepatic failure. Disease Progression and Carcinogenesis in the Gastrointestinal
Tract (P.R. Galle, W.E. Schmidt, G. Gerken, B. Wiedenmann, eds.) Proceedings
of the FALK Symposium 132 held in Freiburg, Germany, Oct. 9-10, 2002
2003:86-99.
307. Smith LH, Jr., Ettinger RH, Seligson D. A comparison of the metabolism of
fructose and glucose in hepatic disease and diabetes mellitus. J Clin Invest
1953;32:273-282.
308. Miller M, Murphy JR, Craig JW, Woodward H. Studies on experimental diabetic
acidosis. Comparison of effect of fructose and glucose in initial hours of
treatment. J Clin Endocr 1953;13:866.
309. Woods HF, Eggleston LV, Krebs HA. The cause of hepatic accumulation of
fructose 1-phosphate on fructose loading. Biochem J 1970;119:501-510.
310. Adelman RC, Morris HP, Weinhouse S. Fructokinase, triokinase, and aldolases in
liver tumors of the rat. Cancer Res 1967;27:2408-2413.
311. Bode JC, Zelder O, Rumpelt HJ, Wittkamp U. Depletion of liver adenosine
phosphates and metabolic effects of intravenous infusion of fructose or sorbitol in
man and in the rat. Eur J Clin Invest 1973;3:436-441.
312. Pan G, Ni J, Wei YF, Yu G, Gentz R, Dixit VM. An antagonist decoy receptor and
a death domain-containing receptor for TRAIL. Science 1997;277:815-818.
313. Leverkus M, Neumann M, Mengling T, Rauch CT, Brocker EB, Krammer PH,
Walczak H. Regulation of tumor necrosis factor-related apoptosis-inducing ligand
- 156 -

References
sensitivity in primary and transformed human keratinocytes. Cancer Res
2000;60:553-559.
314. Scaffidi C, Schmitz I, Krammer PH, Peter ME. The role of c-FLIP in modulation of
CD95-induced apoptosis. J Biol Chem 1999;274:1541-1548.
315. Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli KJ, Debatin KM, et
al. Two CD95 (APO-1/Fas) signaling pathways. Embo J 1998;17:1675-1687.
316. Irmler M, Thome M, Hahne M, Schneider P, Hofmann K, Steiner V, Bodmer JL, et
al. Inhibition of death receptor signals by cellular FLIP. Nature 1997;388:190-195.
317. Goyal L. Cell death inhibition: keeping caspases in check. Cell 2001;104:805-
808.
318. Harris NL, Stein H, Coupland SE, Hummel M, Favera RD, Pasqualucci L, Chan
WC. New approaches to lymphoma diagnosis. Hematology (Am Soc Hematol
Educ Program) 2001:194-220.
319. Cavalli F, Isaacson PG, Gascoyne RD, Zucca E. MALT Lymphomas. Hematology
(Am Soc Hematol Educ Program) 2001:241-258.
320. Dierlamm J, Baens M, Wlodarska I, Stefanova-Ouzounova M, Hernandez JM,
Hossfeld DK, De Wolf-Peeters C, et al. The apoptosis inhibitor gene API2 and a
novel 18q gene, MLT, are recurrently rearranged in the
t(11;18)(q21;q21)p6ssociated with mucosa-associated lymphoid tissue
lymphomas. Blood 1999;93:3601-3609.
321. Gutierrez MI, Cherney B, Hussain A, Mostowski H, Tosato G, Magrath I, Bhatia
K. Bax is frequently compromised in Burkitt's lymphomas with irreversible
resistance to Fas-induced apoptosis. Cancer Res 1999;59:696-703.
322. MacNamara B, Wang W, Chen Z, Hou M, Mazur J, Gruber A, Porwit-MacDonald
A. Telomerase activity in relation to pro- and anti-apoptotic protein expression in
high grade non-Hodgkin's lymphomas. Haematologica 2001;86:386-393.
323. Meijerink JP, Mensink EJ, Wang K, Sedlak TW, Sloetjes AW, de Witte T,
Waksman G, et al. Hematopoietic malignancies demonstrate loss-of-function
mutations of BAX. Blood 1998;91:2991-2997.
324. Snow AL, Chen LJ, Nepomuceno RR, Krams SM, Esquivel CO, Martinez OM.
Resistance to Fas-mediated apoptosis in EBV-infected B cell lymphomas is due
to defects in the proximal Fas signaling pathway. J Immunol 2001;167:5404-
5411.
325. Takakuwa T, Dong Z, Takayama H, Matsuzuka F, Nagata S, Aozasa K. Frequent
mutations of Fas gene in thyroid lymphoma. Cancer Res 2001;61:1382-1385.
326. Bertoni F, Conconi A, Luminari S, Realini C, Roggero E, Baldini L, Carobbio S, et
al. Lack of CD95/FAS gene somatic mutations in extranodal, nodal and splenic
marginal zone B cell lymphomas. Leukemia 2000;14:446-448.
- 157 -

References
327. Gronbaek K, Straten PT, Ralfkiaer E, Ahrenkiel V, Andersen MK, Hansen NE,
Zeuthen J, et al. Somatic Fas mutations in non-Hodgkin's lymphoma: association
with extranodal disease and autoimmunity. Blood 1998;92:3018-3024.
328. Xerri L, Carbuccia N, Parc P, Birg F. Search for rearrangements and/or allelic
loss of the fas/APO-1 gene in 101 human lymphomas. Am J Clin Pathol
1995;104:424-430.
329. Shin MS, Kim HS, Kang CS, Park WS, Kim SY, Lee SN, Lee JH, et al.
Inactivating mutations of CASP10 gene in non-Hodgkin lymphomas. Blood
2002;99:4094-4099.
330. Dales JP, Palmerini F, Devilard E, Hassoun J, Birg F, Xerri L. Caspases:
conductors of the cell death machinery in lymphoma cells. Leuk Lymphoma
2001;41:247-253.
331. Kuchino Y, Kitanaka C. Apoptosis in cancer. Hum Cell 1996;9:223-228.
332. Arends MJ, Wyllie AH. Apoptosis: mechanisms and roles in pathology. Int Rev
Exp Pathol 1991;32:223-254.
333. Mesner PW, Jr., Budihardjo, II, Kaufmann SH. Chemotherapy-induced apoptosis.
Adv Pharmacol 1997;41:461-499.
334. Liebermann DA, Hoffman B, Steinman RA. Molecular controls of growth arrest
and apoptosis: p53-dependent and independent pathways. Oncogene
1995;11:199-210.
335. Kato M, Sato H. Regulation of apoptosis by p53. Exp Med 1995;13:1866-1872.
336. Scrable HJ, Sapienza C, Cavenee WK. Genetic and epigenetic losses of
heterozygosity in cancer predisposition and progression. Adv Cancer Res
1990;54:25-62.
337. Sugiyama A, Kume A, Nemoto K, Lee SY, Asami Y, Nemoto F, Nishimura S, et
al. Isolation and characterization of s-myc, a member of the rat myc gene family.
Proc Natl Acad Sci U S A 1989;86:9144-9148.
338. Fourel G, Trepo C, Bougueleret L, Henglein B, Ponzetto A, Tiollais P, Buendia
MA. Frequent activation of N-myc genes by hepadnavirus insertion in woodchuck
liver tumours. Nature 1990;347:294-298.
339. Asai A, Miyagi Y, Sugiyama A, Nagashima Y, Kanemitsu H, Obinata M, Mishima
K, et al. The s-Myc protein having the ability to induce apoptosis is selectively
expressed in rat embryo chondrocytes. Oncogene 1994;9:2345-2352.
340. Kitanaka C, Sugiyama A, Kanazu S, Miyagi Y, Mishima K, Asai A, Kuchino Y. s-
Myc acts as a transcriptional activator and its sequence-specific DNA binding is
required for induction of programmed cell death in glioma cells. Cell Death Differ
1995;2:123-131.
341. Kuchino Y, Kitanaka C, Asai A. [Application of myc family genes to therapy of
malignant brain tumors]. Tanpakushitsu Kakusan Koso 1995;40:2713-2719.
- 158 -

References
342. Kaufmann SH, Earnshaw WC. Induction of apoptosis by cancer chemotherapy.
Exp Cell Res 2000;256:42-49.
343. Friesen C, Fulda S, Debatin KM. Deficient activation of the CD95 (APO-1/Fas)
system in drug-resistant cells. Leukemia 1997;11:1833-1841.
344. Friesen C, Herr I, Krammer PH, Debatin KM. Involvement of the CD95 (APO-
1/FAS) receptor/ligand system in drug-induced apoptosis in leukemia cells. Nat
Med 1996;2:574-577.
345. Kasibhatla S, Brunner T, Genestier L, Echeverri F, Mahboubi A, Green DR. DNA
damaging agents induce expression of Fas ligand and subsequent apoptosis in T
lymphocytes via the activation of NF-kappa B and AP-1. Mol Cell 1998;1:543-
551.
346. Fulda S, Sieverts H, Friesen C, Herr I, Debatin KM. The CD95 (APO-1/Fas)
system mediates drug-induced apoptosis in neuroblastoma cells. Cancer Res
1997;57:3823-3829.
347. Fulda S, Scaffidi C, Pietsch T, Krammer PH, Peter ME, Debatin KM. Activation of
the CD95 (APO-1/Fas) pathway in drug- and gamma-irradiation-induced
apoptosis of brain tumor cells. Cell Death Differ 1998;5:884-893.
348. Fulda S, Los M, Friesen C, Debatin KM. Chemosensitivity of solid tumor cells in
vitro is related to activation of the CD95 system. Int J Cancer 1998;76:105-114.
349. Bennett MW, O'Connell J, O'Sullivan GC, Brady C, Roche D, Collins JK,
Shanahan F. The Fas counterattack in vivo: apoptotic depletion of tumorinfiltrating
lymphocytes associated with Fas ligand expression by human
esophageal carcinoma. J Immunol 1998;160:5669-5675.
350. Michael-Robinson JM, Pandeya N, Cummings MC, Walsh MD, Young JP,
Leggett BA, Purdie DM, et al. Fas ligand and tumour counter-attack in colorectal
cancer stratified according to microsatellite instability status. J Pathol
2003;201:46-54.
351. Hug H, Strand S, Grambihler A, Galle J, Hack V, Stremmel W, Krammer PH, et
al. Reactive oxygen intermediates are involved in the induction of CD95 ligand
mRNA expression by cytostatic drugs in hepatoma cells. J Biol Chem
1997;272:28191-28193.
352. Jovasevic VM, Mokyr MB. Melphalan-induced expression of IFN-beta in MOPC-
315 tumor-bearing mice and its importance for the up-regulation of TNF-alpha
expression. J Immunol 2001;167:4895-4901.
353. Houghton JA, Harwood FG, Tillman DM. Thymineless death in colon carcinoma
cells is mediated via fas signaling. Proc Natl Acad Sci U S A 1997;94:8144-8149.
354. Tillman DM, Petak I, Houghton JA. A Fas-dependent component in 5fluorouracil/leucovorin-induced
cytotoxicity in colon carcinoma cells. Clin Cancer
Res 1999;5:425-430.
- 159 -

References
355. McGahon AJ, Costa Pereira AP, Daly L, Cotter TG. Chemotherapeutic druginduced
apoptosis in human leukaemic cells is independent of the Fas (APO-
1/CD95) receptor/ligand system. Br J Haematol 1998;101:539-547.
356. Eischen CM, Kottke TJ, Martins LM, Basi GS, Tung JS, Earnshaw WC, Leibson
PJ, et al. Comparison of apoptosis in wild-type and Fas-resistant cells:
chemotherapy-induced apoptosis is not dependent on Fas/Fas ligand
interactions. Blood 1997;90:935-943.
357. Shain KH, Landowski TH, Buyuksal I, Cantor AB, Dalton WS. Clonal variability in
CD95 expression is the major determinant in Fas-medicated, but not
chemotherapy-medicated apoptosis in the RPMI 8226 multiple myeloma cell line.
Leukemia 2000;14:830-840.
358. Seglen PO. Preparation of rat liver cells. 3. Enzymatic requirements for tissue
dispersion. Exp Cell Res 1973;82:391-398.
359. Klaunig JE, Goldblatt PJ, Hinton DE, Lipsky MM, Chacko J, Trump BF. Mouse
liver cell culture. I. Hepatocyte isolation. In Vitro 1981;17:913-925.
360. Klaunig JE, Goldblatt PJ, Hinton DE, Lipsky MM, Trump BF. Mouse liver cell
culture. II. Primary culture. In Vitro 1981;17:926-934.
361. Leist M. Inflammatory stimuli and mediators in the mouse liver. Konstanz:
University of Konstanz; 1993.
362. Osypiw JC, Allen RL, Billington D. Subpopulations of rat hepatocytes separated
by Percoll density-gradient centrifugation show characteristics consistent with
different acinar locations. Biochem J 1994;304:617-624.
363. Van Rooijen N, Sanders A. Liposome mediated depletion of macrophages:
mechanism of action, preparation of liposomes and applications. J Immunol
Methods 1994;174:83-93.
364. Doolittle RL, Richter GW. Isolation and culture of Kupffer cells and hepatocytes
from single rat livers: with observations on iron-loaded kupffer cells. Lab Invest
1981;45:558-566.
365. Schümann J, Wolf D, Pahl A, Brune K, Papadopoulos T, van Rooijen N, Tiegs G.
Importance of kupffer cells for T-cell-dependent liver injury in mice. Am J Pathol
2000;157:1671-1683.
366. Tiegs G, Hentschel J, Wendel A. A T cell-dependent experimental liver injury in
mice inducible by concanavalin A. J Clin Invest 1992;90:196-203.
367. Bergmeyer HU. Evaluation of optimum conditions of two-substrate enzyme
reactions. J Clin Chem Clin Biochem 1977;15:405-410.
368. Bergmeyer HU, Scheibe P, Wahlefeld AW. Optimization of methods for aspartate
aminotransferase and alanine aminotransferase. Clin Chem 1978;24:58-73.
369. Bergmeyer HU. Methods of enzymatic analysis. Weinheim: Verlag Chemie
Weinheim, 1983: 605.
- 160 -

References
370. Thornberry NA. Caspases: a decade of death research. Cell Death Differ
1999;6:1023-1027.
371. Jin G, Huang X, Black R, Wolfson M, Rauch C, McGregor H, Ellestad G, et al. A
continuous fluorimetric assay for tumor necrosis factor-alpha converting enzyme.
Anal Biochem 2002;302:269-275.
372. Tietze F. Enzymic method for quantitative determination of nanogram amounts of
total and oxidized glutathione: applications to mammalian blood and other
tissues. Anal Biochem 1969;27:502-522.
373. Lucas R, Juillard P, Decoster E, Redard M, Burger D, Donati Y, Giroud C, et al.
Crucial role of tumor necrosis factor (TNF) receptor 2 and membrane- bound TNF
in experimental cerebral malaria. Eur J Immunol 1997;27:1719-1725.
374. Lucas R, Heirwegh K, Neirynck A, Remels L, Van Heuverswyn H, De Baetselier
P. Generation and characterization of a neutralizing rat anti-rmTNF-alpha
monoclonal antibody. Immunology 1990;71:218-223.
375. Watanabe-Fukunaga R, Brannan CI, Copeland NG, Jenkins NA, Nagata S.
Lymphoproliferation disorder in mice explained by defects in Fas antigen that
mediates apoptosis. Nature 1992;356:314-317.
376. Vanhaesebroeck B, Van Bladel S, Lenaerts A, Suffys P, Beyaert R, Lucas R, Van
Roy F, et al. Two discrete types of tumor necrosis factor-resistant cells derived
from the same cell line. Cancer Res 1991;51:2469-2477.
377. Jones AL, Selby P. Tumour necrosis factor: clinical relevance. Cancer Surv
1989;8:817-836.
378. Lehmann V, Freudenberg MA, Galanos C. Lethal toxicity of lipopolysaccharide
and tumor necrosis factor in normal and D-galactosamine-treated mice. J Exp
Med 1987;165:657-663.
379. Tiegs G, Wolter M, Wendel A. Tumor necrosis factor is a terminal mediator in
galactosamine/endotoxin-induced hepatitis in mice. Biochem Pharmacol
1989;38:627-631.
380. Decker K, Keppler D. Galactosamine hepatitis: key role of the nucleotide
deficiency period in the pathogenesis of cell injury and cell death. Rev Physiol
Biochem Pharmacol 1974;71:77-106.
381. Leist M, Auer-Barth S, Wendel A. Tumor necrosis factor production in the
perfused mouse liver and its pharmacological modulation by methylxanthines. J
Pharmacol Exp Ther 1996;176:968-976.
382. Gantner F, Leist M, Kusters S, Vogt K, Volk HD, Tiegs G. T cell stimulus-induced
crosstalk between lymphocytes and liver macrophages results in augmented
cytokine release. Exp Cell Res 1996;229:137-146.
383. Leist M, Gantner F, Bohlinger I, Tiegs G, Germann PG, Wendel A. Tumor
necrosis factor-induced hepatocyte apoptosis precedes liver failure in
experimental murine shock models. Am J Pathol 1995;146:1220-1234.
- 161 -

References
384. Tiegs G, Niehorster M, Wendel A. Leukocyte alterations do not account for
hepatitis induced by endotoxin or TNF alpha in galactosamine-sensitized mice.
Biochem Pharmacol 1990;40:1317-1322.
385. Schümann J, Bluethmann H, Tiegs G. Synergism of Pseudomonas aeruginosa
exotoxin A with endotoxin, superantigen, or TNF results in TNFR1- and TNFR2dependent
liver toxicity in mice. Immunol Lett 2000;74:165-172.
386. Gantner F, Leist M, Jilg S, Germann PG, Freudenberg MA, Tiegs G. Tumor
necrosis factor-induced hepatic DNA fragmentation as an early marker of T celldependent
liver injury in mice. Gastroenterology 1995;109:166-176.
387. Miethke T, Wahl C, Heeg K, Echtenacher B, Krammer PH, Wagner H. T cellmediated
lethal shock triggered in mice by the superantigen staphylococcal
enterotoxin B: critical role of tumor necrosis factor. J Exp Med 1992;175:91-98.
388. Pfeffer K, Matsuyama T, Kundig TM, Wakeham A, Kishihara K, Shahinian A,
Wiegmann K, et al. Mice deficient for the 55 kd tumor necrosis factor receptor are
resistant to endotoxic shock, yet succumb to L. monocytogenes infection. Cell
1993;73:457-467.
389. Nagaki M, Muto Y, Ohnishi H, Yasuda S, Sano K, Naito T, Maeda T, et al.
Hepatic injury and lethal shock in galactosamine-sensitized mice induced by the
superantigen staphylococcal enterotoxin B. Gastroenterology 1994;106:450-458.
390. Schümann J, Angermuller S, Bang R, Lohoff M, Tiegs G. Acute hepatotoxicity of
Pseudomonas aeruginosa exotoxin A in mice depends on T cells and TNF. J
Immunol 1998;161:5745-5754.
391. Koesling M, Rott O, Fleischer B. Macrophages are dispensable for superantigenmediated
stimulation and anergy induction of peripheral T cells in vivo. Cell
Immunol 1994;157:29-37.
392. Bette M, Schafer MK, van Rooijen N, Weihe E, Fleischer B. Distribution and
kinetics of superantigen-induced cytokine gene expression in mouse spleen. J
Exp Med 1993;178:1531-1539.
393. Gantner F, Leist M, Lohse AW, Germann PG, Tiegs G. Concanavalin A-induced
T-cell-mediated hepatic injury in mice: the role of tumor necrosis factor.
Hepatology 1995;21:190-198.
394. Toyabe S, Seki S, Iiai T, Takeda K, Shirai K, Watanabe H, Hiraide H, et al.
Requirement of IL-4 and liver NK1+ T cells for concanavalin A-induced hepatic
injury in mice. J Immunol 1997;159:1537-1542.
395. Knolle PA, Gerken G, Loser E, Dienes HP, Gantner F, Tiegs G, Meyer zum
Buschenfelde KH, et al. Role of sinusoidal endothelial cells of the liver in
concanavalin A-induced hepatic injury in mice. Hepatology 1996;24:824-829.
396. Santucci L, Fiorucci S, Cammilleri F, Servillo G, Federici B, Morelli A. Galectin-1
exerts immunomodulatory and protective effects on concanavalin A-induced
hepatitis in mice. Hepatology 2000;31:399-406.
- 162 -

References
397. Fiorucci S, Santucci L, Antonelli E, Distrutti E, Del Sero G, Morelli O, Romani L,
et al. NO-aspirin protects from T cell-mediated liver injury by inhibiting caspasedependent
processing of Th1-like cytokines. Gastroenterology 2000;118:404-
421.
398. Ogasawara J, Watanabe-Fukunaga R, Adachi M, Matsuzawa A, Kasugai T,
Kitamura Y, Itoh N, et al. Lethal effect of the anti-Fas antibody in mice. Nature
1993;364:806-809.
399. Enari M, Hug H, Nagata S. Involvement of an ICE-like protease in Fas-mediated
apoptosis. Nature 1995;375:78-81.
400. Rensing-Ehl A, Hess S, Ziegler-Heitbrock HW, Riethmuller G, Engelmann H.
Fas/Apo-1 activates nuclear factor kappa B and induces interleukin-6 production.
J Inflamm 1995;45:161-174.
401. Kondo T, Suda T, Fukuyama H, Adachi M, Nagata S. Essential roles of the Fas
ligand in the development of hepatitis. Nat Med 1997;3:409-413.
402. Nishimura Y, Hirabayashi Y, Matsuzaki Y, Musette P, Ishii A, Nakauchi H, Inoue
T, et al. In vivo analysis of Fas antigen-mediated apoptosis: effects of agonistic
anti-mouse Fas mAb on thymus, spleen and liver. Int Immunol 1997;9:307-316.
403. Ichikawa K, Yoshida-Kato H, Ohtsuki M, Ohsumi J, Yamaguchi J, Takahashi S,
Tani Y, et al. A novel murine anti-human Fas mAb which mitigates
lymphadenopathy without hepatotoxicity. Int Immunol 2000;12:555-562.
404. Kakinuma C, Takagaki K, Yatomi T, Nakamura N, Nagata S, Uemura A,
Shibutani Y. Acute toxicity of an anti-Fas antibody in mice. Toxicol Pathol
1999;27:412-420.
405. Cardier JE, Schulte T, Kammer H, Kwak J, Cardier M. Fas (CD95, APO-1)
antigen expression and function in murine liver endothelial cells: implications for
the regulation of apoptosis in liver endothelial cells. Faseb J 1999;13:1950-1960.
406. Wanner GA, Mica L, Wanner-Schmid E, Kolb SA, Hentze H, Trentz O, Ertel W.
Inhibition of caspase activity prevents CD95-mediated hepatic microvascular
perfusion failure and restores Kupffer cell clearance capacity. Faseb J
1999;13:1239-1248.
407. Wolf D, Schumann J, Koerber K, Kiemer AK, Vollmar AM, Sass G, Papadopoulos
T, et al. Low-molecular-weight hyaluronic acid induces nuclear factor-kappaB-
dependent resistance against tumor necrosis factor alpha-mediated liver injury in
mice. Hepatology 2001;34:535-547.
408. Alexander HR, Jr., Bartlett DL, Libutti SK. Current status of isolated hepatic
perfusion with or without tumor necrosis factor for the treatment of unresectable
cancers confined to liver. Oncologist 2000;5:416-424.
409. Choti MA, Bulkley GB. Management of hepatic metastases. Liver Transpl Surg
1999;5:65-80.
- 163 -

References
410. Venook AP. Update on hepatic intra-arterial chemotherapy. Oncology (Huntingt)
1997;11:947-957; discussion 961-942, 964, 970.
411. Cancer M-AGi. Reappraisal of hepatic arterial infusion in the treatment of
nonresectable liver metastases from colorectal cancer. Meta-Analysis Group in
Cancer. J Natl Cancer Inst 1996;88:252-258.
412. Lienard D, Eggermont AM, Koops HS, Kroon B, Towse G, Hiemstra S, Schmitz
P, et al. Isolated limb perfusion with tumour necrosis factor-alpha and melphalan
with or without interferon-gamma for the treatment of in-transit melanoma
metastases: a multicentre randomized phase II study. Melanoma Res
1999;9:491-502.
413. Wu PC, McCart A, Hewitt SM, Turner E, Libutti SK, Bartlett DL, Alexander HR.
Isolated organ perfusion does not result in systemic microembolization of tumor
cells. Ann Surg Oncol 1999;6:658-663.
414. Spriggs DR, Sherman ML, Michie H, Arthur KA, Imamura K, Wilmore D, Frei E,
3rd, et al. Recombinant human tumor necrosis factor administered as a 24-hour
intravenous infusion. A phase I and pharmacologic study. J Natl Cancer Inst
1988;80:1039-1044.
415. Rudiger HA, Clavien PA. Tumor necrosis factor alpha, but not Fas, mediates
hepatocellular apoptosis in the murine ischemic liver. Gastroenterology
2002;122:202-210.
416. Leist M, Gantner F, Bohlinger I, Germann PG, Tiegs G, Wendel A. Murine
hepatocyte apoptosis induced in vitro and in vivo by TNF-alpha requires
transcriptional arrest. J Immunol 1994;153:1778-1788.
417. Yamada Y, Fausto N. Deficient liver regeneration after carbon tetrachloride injury
in mice lacking type 1 but not type 2 tumor necrosis factor receptor. Am J Pathol
1998;152:1577-1589.
418. Yamada Y, Webber EM, Kirillova I, Peschon JJ, Fausto N. Analysis of liver
regeneration in mice lacking type 1 or type 2 tumor necrosis factor receptor:
requirement for type 1 but not type 2 receptor. Hepatology 1998;28:959-970.
419. Knight B, Yeoh GC, Husk KL, Ly T, Abraham LJ, Yu C, Rhim JA, et al. Impaired
preneoplastic changes and liver tumor formation in tumor necrosis factor receptor
type 1 knockout mice. J Exp Med 2000;192:1809-1818.
420. Luster MI, Simeonova PP, Gallucci RM, Bruccoleri A, Blazka ME, Yucesoy B,
Matheson JM. The role of tumor necrosis factor alpha in chemical-induced
hepatotoxicity. Ann N Y Acad Sci 2000;919:214-220.
421. Simeonova PP, Gallucci RM, Hulderman T, Wilson R, Kommineni C, Rao M,
Luster MI. The role of tumor necrosis factor-alpha in liver toxicity, inflammation,
and fibrosis induced by carbon tetrachloride. Toxicol Appl Pharmacol
2001;177:112-120.
- 164 -

References
422. Enomoto N, Takei Y, Hirose M, Ikejima K, Miwa H, Kitamura T, Sato N.
Thalidomide prevents alcoholic liver injury in rats through suppression of Kupffer
cell sensitization and TNF-alpha production. Gastroenterology 2002;123:291-300.
423. Sharma RP, Bhandari N, Riley RT, Voss KA, Meredith FI. Tolerance to fumonisin
toxicity in a mouse strain lacking the P75 tumor necrosis factor receptor.
Toxicology 2000;143:183-194.
424. Probert L, Akassoglou K, Kassiotis G, Pasparakis M, Alexopoulou L, Kollias G.
TNF-alpha transgenic and knockout models of CNS inflammation and
degeneration. J Neuroimmunol 1997;72:137-141.
425. Akassoglou K, Douni E, Bauer J, Lassmann H, Kollias G, Probert L. Exclusive
tumor necrosis factor (TNF) signaling by the p75TNF receptor triggers
inflammatory ischemia in the CNS of transgenic mice. Proc Natl Acad Sci U S A
2003;100:709-714.
426. Akassoglou K, Probert L, Kontogeorgos G, Kollias G. Astrocyte-specific but not
neuron-specific transmembrane TNF triggers inflammation and degeneration in
the central nervous system of transgenic mice. J Immunol 1997;158:438-445.
427. Wajant H, Pfizenmaier K, Scheurich P. Tumor necrosis factor signaling. Cell
Death Differ 2003;10:45-65.
428. Alexopoulou L, Pasparakis M, Kollias G. A murine transmembrane tumor
necrosis factor (TNF) transgene induces arthritis by cooperative p55/p75 TNF
receptor signaling. Eur J Immunol 1997;27:2588-2592.
429. Douni E, Kollias G. A critical role of the p75 tumor necrosis factor receptor
(p75TNF-R) in organ inflammation independent of TNF, lymphotoxin alpha, or the
p55TNF-R. J Exp Med 1998;188:1343-1352.
430. Grell M, Zimmermann G, Hulser D, Pfizenmaier K, Scheurich P. TNF receptors
TR60 and TR80 can mediate apoptosis via induction of distinct signal pathways.
J Immunol 1994;153:1963-1972.
431. Vilcek J, Feldmann M. Historical review: Cytokines as therapeutics and targets of
therapeutics. Trends Pharmacol Sci 2004;25:201-209.
432. MacEwan DJ. TNF receptor subtype signalling: differences and cellular
consequences. Cell Signal 2002;14:477-492.
433. MacEwan DJ. TNF ligands and receptors--a matter of life and death. Br J
Pharmacol 2002;135:855-875.
434. Garcia ST, McQuillan A, Panasci L. Correlation between the cytotoxicity of
melphalan and DNA crosslinks as detected by the ethidium bromide fluorescence
assay in the F1 variant of B16 melanoma cells. Biochem Pharmacol
1988;37:3189-3192.
435. Frankfurt OS. Detection of DNA damage in individual cells by flow cytometric
analysis using anti-DNA monoclonal antibody. Exp Cell Res 1987;170:369-380.
- 165 -

References
436. Masta A, Gray PJ, Phillips DR. Nitrogen mustard inhibits transcription and
translation in a cell free system. Nucleic Acids Res 1995;23:3508-3515.
437. Grell M, Wajant H, Zimmermann G, Scheurich P. The type 1 receptor (CD120a)
is the high-affinity receptor for soluble tumor necrosis factor. Proc Natl Acad Sci
U S A 1998;95:570-575.
438. Moss ML, Jin SL, Milla ME, Bickett DM, Burkhart W, Carter HL, Chen WJ, et al.
Cloning of a disintegrin metalloproteinase that processes precursor tumournecrosis
factor-alpha. Nature 1997;385:733-736.
439. Tiegs G. Immunotoxicology of host-response-mediated experimental liver injury.
J Hepatol 1994;21:890-903.
440. Nakama T, Hirono S, Moriuchi A, Hasuike S, Nagata K, Hori T, Ido A, et al.
Etoposide prevents apoptosis in mouse liver with D-
galactosamine/lipopolysaccharide-induced fulminant hepatic failure resulting in
reduction of lethality. Hepatology 2001;33:1441-1450.
441. Mizuhara H, O'Neill E, Seki N, Ogawa T, Kusunoki C, Otsuka K, Satoh S, et al. T
cell activation-associated hepatic injury: mediation by tumor necrosis factors and
protection by interleukin 6. J Exp Med 1994;179:1529-1537.
442. Aversa G, Punnonen J, de Vries JE. The 26-kD transmembrane form of tumor
necrosis factor alpha on activated CD4+ T cell clones provides a costimulatory
signal for human B cell activation. J Exp Med 1993;177:1575-1585.
443. Sass G, Koerber K, Bang R, Guehring H, Tiegs G. Inducible nitric oxide synthase
is critical for immune-mediated liver injury in mice. J Clin Invest 2001;107:439-
447.
444. Ulich TR, Guo KZ, Irwin B, Remick DG, Davatelis GN. Endotoxin-induced
cytokine gene expression in vivo. II. Regulation of tumor necrosis factor and
interleukin-1 alpha/beta expression and suppression. Am J Pathol
1990;137:1173-1185.
445. Kiemer AK, Vollmar AM, Bilzer M, Gerwig T, Gerbes AL. Atrial natriuretic peptide
reduces expression of TNF-alpha mRNA during reperfusion of the rat liver upon
decreased activation of NF-kappaB and AP-1. J Hepatol 2000;33:236-246.
446. Kiemer AK, Vollmar AM. Effects of different natriuretic peptides on nitric oxide
synthesis in macrophages. Endocrinology 1997;138:4282-4290.
447. de Vries MR, Borel Rinkes IH, Swaak AJ, Hack CE, Van De Velde CJ, Wiggers
T, Tollenaar RA, et al. Acute-phase response patterns in isolated hepatic
perfusion with tumour necrosis factor alpha (TNF-alpha) and melphalan in
patients with colorectal liver metastases. Eur J Clin Invest 1999;29:553-560.
448. Gorelik L, Rubin M, Prokhorova A, Mokyr MB. Importance of TNF production for
the curative effectiveness of low dose melphalan therapy for mice bearing a large
MOPC-315 tumor. J Immunol 1995;154:3941-3951.
- 166 -

References
449. Manusama ER, Nooijen PT, Stavast J, Durante NM, Marquet RL, Eggermont AM.
Synergistic antitumour effect of recombinant human tumour necrosis factor alpha
with melphalan in isolated limb perfusion in the rat. Br J Surg 1996;83:551-555.
450. Furrer M, Altermatt HJ, Ris HB, Althaus U, Ruegg C, Lienard D, Lejeune FJ. Lack
of antitumour activity of human recombinant tumour necrosis factor-alpha, alone
or in combination with melphalan in a nude mouse human melanoma xenograft
system. Melanoma Res 1997;7 Suppl 2:S43-S49.
451. Van Molle W, Wielockx B, Mahieu T, Takada M, Taniguchi T, Sekikawa K, Libert
C. HSP70 protects against TNF-induced lethal inflammatory shock. Immunity
2002;16:685-695.
452. Kuppen PJ, Jonges LE, van de Velde CJ, Vahrmeijer AL, Tollenaar RA, Borel
Rinkes IH, Eggermont AM. Liver and tumour tissue concentrations of TNF-alpha
in cancer patients treated with TNF-alpha and melphalan by isolated liver
perfusion. Br J Cancer 1997;75:1497-1500.
453. Lans TE, Bartlett DL, Libutti SK, Gnant MF, Liewehr DJ, Venzon DJ, Turner EM,
et al. Role of tumor necrosis factor on toxicity and cytokine production after
isolated hepatic perfusion. Clin Cancer Res 2001;7:784-790.
454. Brinkman BM, Telliez JB, Schievella AR, Lin LL, Goldfeld AE. Engagement of
tumor necrosis factor (TNF) receptor 1 leads to ATF-2- and p38 mitogenactivated
protein kinase-dependent TNF-alpha gene expression. J Biol Chem
1999;274:30882-30886.
455. Beg AA, Baltimore D. An essential role for NF-kappaB in preventing TNF-alphainduced
cell death. Science 1996;274:782-784.
456. Natoli G, Costanzo A, Guido F, Moretti F, Bernardo A, Burgio VL, Agresti C, et al.
Nuclear factor kB-independent cytoprotective pathways originating at tumor
necrosis factor receptor-associated factor 2. J Biol Chem 1998;273:31262-31272.
457. Van Antwerp DJ, Martin SJ, Verma IM, Green DR. Inhibition of TNF-induced
apoptosis by NF-kappa B. Trends Cell Biol 1998;8:107-111.
458. Seitz C, Muller P, Krieg RC, Mannel DN, Hehlgans T. A novel p75TNF receptor
isoform mediating NFkappa B activation. J Biol Chem 2001;276:19390-19395.
459. Paul A, Wilson S, Belham CM, Robinson CJ, Scott PH, Gould GW, Plevin R.
Stress-activated protein kinases: activation, regulation and function. Cell Signal
1997;9:403-410.
460. Donnahoo KK, Shames BD, Harken AH, Meldrum DR. Review article: the role of
tumor necrosis factor in renal ischemia-reperfusion injury. J Urol 1999;162:196-
203.
461. Meldrum DR, Donnahoo KK. Role of TNF in mediating renal insufficiency
following cardiac surgery: evidence of a postbypass cardiorenal syndrome. J
Surg Res 1999;85:185-199.
- 167 -

References
462. Dent P, Yacoub A, Contessa J, Caron R, Amorino G, Valerie K, Hagan MP, et al.
Stress and radiation-induced activation of multiple intracellular signaling
pathways. Radiat Res 2003;159:283-300.
463. Takeda K, Matsuzawa A, Nishitoh H, Ichijo H. Roles of MAPKKK ASK1 in stressinduced
cell death. Cell Struct Funct 2003;28:23-29.
464. Bennett BL, Sasaki DT, Murray BW, O'Leary EC, Sakata ST, Xu W, Leisten JC,
et al. SP600125, an anthrapyrazolone inhibitor of Jun N-terminal kinase. Proc
Natl Acad Sci U S A 2001;98:13681-13686.
465. Gupta S, Stravitz RT, Dent P, Hylemon PB. Down-regulation of cholesterol
7alpha-hydroxylase (CYP7A1) gene expression by bile acids in primary rat
hepatocytes is mediated by the c-Jun N-terminal kinase pathway. J Biol Chem
2001;276:15816-15822.
466. Han Z, Boyle DL, Chang L, Bennett B, Karin M, Yang L, Manning AM, et al. c-Jun
N-terminal kinase is required for metalloproteinase expression and joint
destruction in inflammatory arthritis. J Clin Invest 2001;108:73-81.
467. Onishi I, Tani T, Hashimoto T, Shimizu K, Yagi M, Yamamoto K, Yoshioka K.
Activation of c-Jun N-terminal kinase during ischemia and reperfusion in mouse
liver. FEBS Lett 1997;420:201-204.
468. Zwacka RM, Zhang Y, Zhou W, Halldorson J, Engelhardt JF.
Ischemia/reperfusion injury in the liver of BALB/c mice activates AP-1 and
nuclear factor kappaB independently of IkappaB degradation. Hepatology
1998;28:1022-1030.
469. Hentze H, Schmitz I, Latta M, Krueger A, Krammer PH, Wendel A. Glutathione
dependence of caspase-8 activation at the death-inducing signaling complex. J
Biol Chem 2002;222:5588-5595.
470. Awasthi S, Bajpai KK, Piper JT, Singhal SS, Ballatore A, Seifert WE, Jr., Awasthi
YC, et al. Interactions of melphalan with glutathione and the role of glutathione Stransferase.
Drug Metab Dispos 1996;24:371-374.
471. Troyano A, Fernandez C, Sancho P, de Blas E, Aller P. Effect of glutathione
depletion on antitumor drug toxicity (apoptosis and necrosis) in U-937 human
promonocytic cells. The role of intracellular oxidation. J Biol Chem
2001;276:47107-47115.
472. Calvert P, Yao KS, Hamilton TC, O'Dwyer PJ. Clinical studies of reversal of drug
resistance based on glutathione. Chem Biol Interact 1998;111-112:213-224.
473. Eguchi Y, Shimizu S, Tsujimoto Y. Intracellular ATP levels determine cell death
fate by apoptosis or necrosis. Cancer Res 1997;57:1835-1840.
474. Leist M, Single B, Castoldi AF, Kuhnle S, Nicotera P. Intracellular adenosine
triphosphate (ATP) concentration: a switch in the decision between apoptosis and
necrosis. J Exp Med 1997;185:1481-1486.
- 168 -

References
475. Liu X, Kim CN, Yang J, Jemmerson R, Wang X. Induction of apoptotic program in
cell-free extracts: requirement for dATP and cytochrome c. Cell 1996;86:147-157.
476. Stridh H, Fava E, Single B, Nicotera P, Orrenius S, Leist M. Tributyltin-induced
apoptosis requires glycolytic adenosine trisphosphate production. Chem Res
Toxicol 1999;12:874-882.
477. Leist M, Single B, Naumann H, Fava E, Simon B, Kuhnle S, Nicotera P. Nitric
oxide inhibits execution of apoptosis at two distinct ATP- dependent steps
upstream and downstream of mitochondrial cytochrome c release. Biochem
Biophys Res Commun 1999;258:215-221.
478. Leist M, Single B, Naumann H, Fava E, Simon B, Kuhnle S, Nicotera P. Inhibition
of mitochondrial ATP generation by nitric oxide switches apoptosis to necrosis.
Exp Cell Res 1999;249:396-403.
479. Ferrari D, Stepczynska A, Los M, Wesselborg S, Schulze-Osthoff K. Differential
regulation and ATP requirement for caspase-8 and caspase-3 activation during
CD95- and anticancer drug-induced apoptosis. J Exp Med 1998;188:979-984.
480. Volbracht C, Leist M, Nicotera P. ATP controls neuronal apoptosis triggered by
microtubule breakdown or potassium deprivation. Mol Med 1999;5:477-489.
- 169 -