Part 3. Signal transduction leading to cell death.
WP6: Cell death pathways: linking cell death to inflammation and therapy
Current knowledge indicates that most of the molecular signalling complexes leading to the activation
of NF-kB and inflammation are also involved in the initiation of apoptotic and necrotic cell death
processes. Moreover, in certain conditions cell death, may even be considered as an important
inflammation-modulating process. In models of ischemia-reperfusion, prevention of cell death
contributes to reduced inflammation suggesting that cell death may precede or amplify the
inflammation process (Daemen et al., 1999). Similarly,
ongoing inflammation during rheumatoid
arthritis has been attributed to necrotic cell death processes, for example by the release of HMGB1
from the dying cells (Andersson et al., 2004). Therefore, a better understanding of the molecular
mechanisms governing the impact of cell death on the inflammatory process could provide new
molecular basis for therapeutic intervention. The general aim of this workpackage is to integrate
different levels of cell death research: morphology of cell death, intracellular signal transduction,
intercellular communication (phagocytosis, inflammation, immunomodulation), involvement of cell
death in pathologies and use of cell death related molecules as therapeutic targets to influence
inflammation. Since the core signalling pathways paradigms have been worked out for apoptosis, the
challenge today is to determine if and how the process is manipulated by intracellular pathogens to
their advantage,
to discover
alternative cell death pathways underlining the various cell death
processes and the molecular links and crosstalks with inflammatory signalling.
WP6.1. Molecular signalling in cell death
This work package combines the efforts of different groups in the program (ULG1, ULG2, UG1, UG2,
KUL) to integrate the knowledge and share the tools and methods, to study the induction or modulation
of the cell death processes elicited by different stimuli such as bacterial and viral infection,
photodynamic therapy (PDT), endoplasmatic reticulum (ER) stress, TLR ligands and TNF ligation. We
believe sharing data in an early stage on cell death pathways elicited by different stimuli will be of
synergistic value to the whole consortium in identifying points of convergence or crosstalks.
The following items will be addressed:
RIP1–mediated programmed necrosis. RIP1 is a serine/threonine kinase containing a dead domain
(DD). It plays a central role in NF-B activation and necrotic cell death induced by TNF, dsRNA and
hypoxia. In TLR3 response to viral infection or dsRNA, RIP1 is required in addition for signalling to
apoptosis (Kalai et al. 2002, Kaiser and Offermann (2005). Therefore, UG2 will study the involvement
of RIP1 in necrotic cell death induced by photochemically generated oxidative stress with KUL.
Structure-function analysis of RIP1 carried out by UG2 suggests that the NF-B activation and necrotic
cell death promoting functions of the protein can be separated. The UG2 group is developing
conditional transgenic mice overexpressing the RIP1 mutants lacking necrotic signalling and
conditional knock-in mice. When the mice are available, they will be also analysed in the various
disease models and infection models available in the consortium including viral and bacterial infection
models.
ER stress mediated cell death. Cell death signalling pathways initiated by ER stress are still poorly
understood. The ER is especially developed in B cell lymphoma and myeloma cells. Therefore, with
the hope that a better understanding of these pathways may lead to development of new treatments
for these types of cancer, UG2 and KUL will study the involvement of PKR, PERK, IRE-1, ATF-6,
caspases, p53 and BCL2 family members, in cell death induced by ER-stress agents such as
thapsigargin, tunicamycin, brefeldin A and hypericin-PDT in B cell lymphoma and myeloma cells.
The KUL group has recently reported that ER-Ca2+ depletion caused by hypericin-PDT, a paradigm of
an anticancer treatment utilizing reactive oxygen species (ROS) to kill the cancer cells or endothelial
cells (Dolmans et al., 2003), engages an autophagic pathway which causes cell death in apoptoticdeficient cells (Buytaert et al., 2006). Given the still uncertain role of autophagy as a tumor suppression
mechanism and as a death pathway in response to anticancer therapy KUL will focus on the molecular
mechanisms and role of autophagy as principal backup lethal program activated by PDT. The crucial
question is how and when autophagy, which is in essence a survival mechanism following cellular
stress, turns into a veritable cell death mechanism and whether it uses caspase-dependent or
caspase-independent cell death pathways. Thus, the KUL group will undertake a thorough comparison
of the effects of PDT in MEFs or human cancer cell lines with RNAi silenced cells, or in apoptosisdeficient cells (with UG2). Analysis of the crosstalk between the apoptotic and autophagic machinery
will clarify whether these pathways actively suppress/influence each other or are promoted
independently. The role of calpains and of the death kinases RIP1 and DAPK as mediators of
autophagic cell death, and the interaction between Bcl-2 proteins with Beclin-1 will be investigated. The
fact that RIP1, a pronecrotic kinase, also plays a role in autophagic cell death (Yu et al., 2004), and
that caspase-inhibition apparently promotes autophagy, suggests an intricate relationship between
these three cell death pathways. The UG2 group will also evaluate the RIP1 mutants for this function
and the effects of caspases in the regulation of Beclin-1following growth factor depletion or cellular
stress (PDT with KUL).
Death pathways in T-lymphocytes and neutrophils. During the inflammatory reaction, activated tissueresident macrophages and mast cells release various mediators (histamine, leukotrienes, chemokines,
TNF,..) that are perceived by neutrophils, which in turn respond by a ROS production and
degranulation. ROS strongly influence lymphocytes and other cells that are present in the tissue or in
the close vicinity. Recently, the ULG1 laboratory has shown that T lymphocytes exposure to ROS
leads to apoptosis by a still undefined pathway. This ROS mediated pathway can be suppressed by
the expression of the lipid phosphatase SHIP-1, which also counteracts Fas ligand induced T cell
apoptosis (Gloire et al., 2006). Hence, ULG1 will be involved in the molecular characterization of (i)
the apoptosis pathways in T lymphocytes subjected to an oxidative stress with a particular attention to
the role of the SHIP-1 phosphatase (ii) the domain of SHIP-1 that are required for acting as protector of
apoptosis, and (iii) the signalling pathways influenced by SHIP-1. In addition, ULG1 will determine
whether or not SHIP-1 is also protecting T lymphocytes from apoptosis induced by a series of
compounds known to trigger apoptosis at various cellular sites (mitochondria, ER, nucleus...) (in
collaboration with KUL).
Preliminary in vitro results from ULG2 reveal that the engagement of P2X1 receptors contributes to the
activation and cell death of human blood neutrophils. Within the framework of the WP6.1, ULG2 will
characterize the cell death modalities, apoptotic, necrotic or autophagic, of wild-type and P2X1-deficient
neutrophils ex vivo (see also WP4.6). The intracellular signalling pathways activated upon stimulation
of murine and human neutrophils with ATP or stable P2X-subtype selective analogs, used alone and in
combination with relevant pro-inflammatory molecules or Toll-like receptor agonists, will be studied. In
particular, the role of caspase-1 activation by the NALP-containing inflammasomes (caspase-1dependent production of IL-1 and IL-18) and caspase-11 up-regulation, and signalling through
caspase-8 and caspase-3 will be analysed (with UG2).
Caspase-1 in inflammation and cell death (pyroptosis). Caspase-1 is a crucial mediator in inflammation
due to the proteolytic activation of pro-IL1. In humans several caspase-1 regulatory gene products
have been reported that originated recently by gene duplication of the caspase-1 gene and that
modulates caspase-1 activation. These gene products consist of the prodomain of caspase-1 and are
therefore called caspase-1 CARD-only proteins and include COP, INCA and ICEBERG (Lamkanfi et
al., 2004a). UG2 has also demonstrated that the prodomain of caspase-1 is able to recruit RIP2 and
TRAF6 leading to activation of NF-B, a feature shared by COP, but not by the more distantly related
INCA and ICEBERG (Lamkanfi et al. 2004a, 2006). Recently caspase-1 has also been implicated in
pyroptosis of Salmonella-infected macrophages (Franchi et al. 2006). In the UG2 group, the molecular
mechanism of this bacteria-induced caspase-1 activation is unraveled and the type of cell death is
analysed by time laps. Moreover, CARD-only mutants of caspase-1 are developed that interfere both
with caspase-1 activation and caspase-1-mediated NF-B activation, and which apparently block
induction of pyroptosis. The development of transgenic mice of this CARD-only mutant will unable UG2
to evaluate the modulatory effects on inflammation and pyroptosis, and to test various disease and
infection models available in the consortium.
Terminal differentiation of the keratinocytes. Terminal differentiation of the keratinocytes can be
considered as a special type of programmed cell death in which the dead body corpses (corneocytes)
are not removed by phagocytosis but remain and function as a barrier against water loss and infection
(Lippens et al. 2005). UG2 group has identified caspase-14 as being specifically expressed and
proteolytically activated during terminal differentiation of keratinocytes. UG2 has developed a caspase14 knockout mice, which has a severe barrier phenotype. This knockout mouse will be analysed in
inflammation
models
(psoriasis
induced
by
ubiquimod
a
TLRx
ligand),
infection
models
(Pseudomonas), damage models (UVB, PDT in collaboration with KUL) and skin cancer development
(melanomas, papillomas, squamous and basal cell carcinomas) whether or not in p53 deficient
background.
WP6.2 Identification of protein complexes and key executioner molecules in cell death and
inflammation
Cell death and inflammation processes are often initiated by molecular complexes that consist of a
platform/sensor molecule (TNF receptor, Apaf-1, NALPs), adaptor molecules combining two
recognition motifs (TRADD, FADD, RAIDD, PYCARD, CARDINAL) and effector proteins (caspase or
RIP kinases). Cellular stress, infection, death domain ligands, organelle perturbation lead to the
activation of a platform, often an ATP-dependent process, the recruitment of appropriate adaptors, and
the recruitment and activation of the effector proteins. This initiates cellular responses such as NF-B
activation, necrosis or apoptosis. In this WP6.2 the protein complexes initiating necrotic cell death
(necrosome) or cellular stress (stressosome) following DNA damage, heat, ischemia-reperfusion,
oxidative stress, hypoxia, will be studied.
The necrosome and stressosome. In preliminary experiments UG2 has already described a complex
containing PIDD, RAIDD, caspase-2, IKK, IKK, NEMO, Hsp90, RIP1. This complex, called
stressosome, has been demonstrated by gel filtration and coimmunoprecipitation experiments, and is
formed in conditions of heat treatment, hydrogen peroxide stress, hypoxia. This stressosome complex
has in theory many different outcomes in inflammation (NF-B), apoptosis (caspase-2) and necrosis
(RIP1). UG2 is now exploring this complex formation in cell deficient for certain of these proteins (by
RNAi) and evaluate the effect on the biological outcome (NF-B activation, apoptosis, necrosis). In
collaboration with KUL, UG2 will evaluate whether the stressosome is formed in UVB exposed normal
or transformed keratinocytes, its contribution to sunburn cell formation and its crosstalk with the
proapoptotic p38 MAPK (Van Laethem et al., 2004) and p53 pathways. In order to identify the crucial
signal transduction proteins and substrates in necrotic cell death, the UG2 group is now setting up a
RNAi library screening. In parallel UG2 is also following a biochemical approach to identify a possible
necrosome complex containing RIP1 using gel filtration, coimmunoprecipitation and two dimensional
blue native gel electrophroresis/PAGE and mass spectrometry (see also WP1.1)
WP6.3 Intercellular crosstalk between cell death and inflammation
Several factors that modulate the innate immune system have been identified in dying cells. These
include intracellular factors HMGB1, Hsp70 and modified phospholipids. Cell death coupled with
phagocytosis of the dying cell is an important alternative mechanism for clearance when an infected
cell fails to eliminate an intracellular pathogen. Understanding better how cell death and phagocytosis
of dying cells promote clearance of pathogens and development of immunity and how these pathways
are deregulated during infection, may help in designing new and better vaccination strategies and
antitumor approaches.
Signals from dying cells. Within UG1 and UG2 a microarray analysis has been performed on
macrophages (naïve macrophages, M1 “killer” macrophages, M2 “healer” macrophages) (Mantovani et
al., 2005) that have been exposed to living, apoptotic or necrotic dying cells. The preliminary results
show that apoptotic dying cells strongly activate genes involved in tissue repair (VEGF, MMPs, etc.).
UG1 and UG2 want to validate these results and expand them to other relevant cellular coincubation
systems (neutrophils, endothelial cells, dendritic cells).
KUL will explore the way by which the robust expression of HSP70 in response to PDT affects cell
death (intracellular) and modulate the immune response (intercellular). KUL will explore how surface
expression/release of HSP70 in PDT-stressed living cells or dying cells (and the dependence on the
cell death modalities) modulates their recognition by macrophages and release of pro-inflammatory
cytokines. The use of hsp70-/- MEFs (with UG2), stably transfected siRNA-hsp70 cancer cell lines and
apoptosis-deficient cells, will be very instrumental for this analysis. Furthermore, KUL will disclose
whether the cytoprotective function of HSP70 in PDT relays on its powerful chaperone activity, the
inhibition of key modulators of the apoptotic machinery and/or the protection from lysosomal damage
(with UG1 & UG2).
Tumor-derived eicosanoids and MMPs. Tumor-derived eicosanoids and matrix metalloproteinases
(MMPs) are crucial regulators of inflammation, modulate sensitivity of tumor cells to death-inducing
insults and are deeply implicated at the tumor invasion/metastatic front. These molecules are
promising targets in many disease pathways and are explored at the molecular and in vivo levels by
several laboratories of this consortium.
UG1 has shown that overexpression of the antioxidant enzyme glutathione peroxidase-4 (GPx4) result
in pronounced suppression of solid tumor growth in weakly malignant tumors, whereas no effect is
observed in modestly and strongly malignant tumors (Heirman et al. 2006). This differential effect of
GPx4 on solid tumor growth correlated with the expression levels of cyclooxygenase-2 (COX-2) and
VEGF in the tumor models (e.g. COX-2high/VEGFlow in weakly malignant tumor, COX-2low/VEGFhigh in
modestly/strongly malignant tumors). Furthermore, tumor growth-inhibiting effect of GPx4 in the weakly
malignant tumor model was due to the inhibition of COX-2 expression and COX-2 activity in the tumor
cells. UG2 will now concentrate on the clarification of the mechanisms and signalling pathways that
define the malignant potential of tumors. As COX-2 expression is mainly dependent on p38 MAPK
signalling pathways and NF-B activity, the constitutive activity of these signalling molecules in low
malignancy tumors will be verified and compared to the activity levels in high malignancy tumors. As
VEGF expression is controlled by the transcription factor HIF-1, also HIF-1 basal expression levels
will be verified in the respective tumor models. As GPx4 inhibits basal COX-2 expression and COX-2
induction by TNF treatment, the interference of the phospholipid hydroperoxide scavenger with p38
and NF-B pathways will be verified in order to look for regulatory feedback loops between COX–
derived prostanoids and p38/NF-B pathways (with KUL & UG2). To establish the relevance for human
malignancies, the tumor-sensitizing effect of GPx4 on angiostatic therapy applying drugs presently in
clinical use (thalidomide, others), will be evaluated. Also, the malignancy predictive value of the
polarized COX-2/VEGF expression ratio will be verified by in silico analysis of public databases along
with expression analysis on cancer biopsies.
KUL and ULG1 will focus on the molecular pathways modulating the expression/activity of COX-2
(Hendrickx et al. 2003; 2005; Volanti et al. 2005) and MMPs in PDT treated cells using
photosensitizers targeted to different subcellular localization (with UG2). The significance of MMP-1,
MMP-10 and MMP-13, as major MMP members induced by hypericin-PDT in promoting motility,
invasion and metastasis of the surviving cancer cells will be challenged in specific MMP knockout
genetic models (with UG2). As recent transcriptomic analysis of bladder cancer cells treated with
hypericin-PDT either alone or in combination with a p38 MAPK antagonist discloses that the
expression of key pro-inflammatory, proangiogenic and metastastic genes is dependent on the p38
MAPK pathway, KUL and ULG1 will analyse the underlying mechanisms (e.g. transcriptional, mRNA
stability) leading to their up-regulation and validate these targets at the functional level. A transgenic
zebrafish (fli1:EGFP) model will be used to visualize and validate the effect of the inhibition
(pharmacological or morpholino knockdown) of proangiogenic signalling pathways in PDT at the
organism level.
WP6.4 Cell death signalling molecules as a therapeutic agent
Defects in apoptosis programs may contribute to tumor progression and treatment resistance and may
be caused by deregulated expression and/or function of anti-apoptotic or pro-apoptotic molecules. The
recent success using therapeutics that specifically target deregulated signal transduction pathways in
cancer cells, underscores the importance of identifying deranged pathways in cancer therapy and
evidences that the most adroit strategy to increase therapeutic potential of conventional anticancer
therapies relies in the simultaneous targeting of key elements of (pathogenic) signal transduction
cascades.
Targeted genes with strong immunomodulatory effects include the caspase-1 derived COP, INCA and
ICEBERG, which lead to CARD-only proteins that modulate the activation of caspase-1 by interfering
with the recruitment of caspase-1 in inflammasome complexes and/or preventing the recruitment of
other immunomodulatory pathways (RIP2, TRAF6). Moreover, in human also caspase-12 has also
evolved to a shortened and catalytically inactive version of caspase-12 that would modulate immune
responses during sepsis (Saleh et al., 2004). In view of these strong modulatory effects of CARD-only
proteins UG2 is identifying mutant forms of caspase-1 and caspase-2 that have very potent antiinflammatory properties and which are able to interfere with inflammasome and stressosome
formation. On the other hand, certain CARD-only proteins derived from caspase-1 and caspase-2, are
also strong activators of NF-B through the recruitment of RIP2/TRAF6 (Lamkanfi et al. 2004b) and
RIP1/TRAF2 (Lamkanfi et al. 2005), respectively.
Signal transduction-based therapy in cancer. Inhibitor of apoptosis proteins (IAPs) are expressed at
high levels in many tumors and have been associated with refractory disease and poor prognosis.
Since IAPs block apoptosis at the effector phase strategies targeting IAPs may prove to be especially
effective to overcome resistance. To target IAP expression in cancers, EU3 has developed cellpermeable Smac peptides, which strongly sensitized even resistant tumor cells for apoptosis induced
by death receptor ligation, chemotherapy or -irradiation in vitro and also in vivo in an intracranial
mouse model of malignant glioma (Fulda et al., 2002). Moreover, ectopic expression of mitochondrial
or cytosolic Smac sensitize various cancer types for -irradiation-induced apoptosis (Giagkousiklidis et
al., 2005) and exert an inhibitory effect on cell proliferation under conditions where cell-cell contact is
lost (Vogler et al., 2005). With the primary aim of evaluating IAP antagonists as novel cancer
therapeutics to potentiate the efficacy of cytotoxic therapies selectively in cancer cells, the following
questions will be addressed: what is the impact of different IAP antagonists (Smac peptides, full-length
and cytosolic Smac, small molecule IAP antagonists) on cancer cell sensitivity towards cytotoxic
therapies (chemotherapy, cytotoxic ligands,-irradiation)? Do IAP antagonists selectively sensitize
malignant cells compared to nonmalignant cells for apoptosis? Which signalling pathways, e.g. NF-B,
are modulated by IAP antagonists? Finally, we will evaluate the therapeutic potential of IAP
antagonists in combination with cytotoxic therapies in an orthotopic mouse model of human malignant
glioma in vivo.
KUL is investigating the therapeutic efficacy of hypericin-based PDT for superficial bladder tumors
(Kamuhabwa et al. 2004). In this WP6.4 an orthotopic rat bladder tumor model, which is currently used
as a preclinical in vivo model in our laboratory, will be used to validate the knowledge obtained from in
vitro studies. The small intravesical volume (300 µl) of the rat bladder and the readily accessibility of
the urothelium tumor lesions, makes it possible to instillate hypericin in the presence of cell permeable
pharmacological inhibitors of known (e.g. p38 MAPK/COX-2, MMPs) or newly identified molecular
targets revealed by the approaches described in WP6.1-3, prior to light delivery. IAP antagonists (e.g.
cell permeable Smac peptides) will be also explored in combination with hypericin-PDT (with EU3).
This preclinical model will reveal the in vivo significance of the signalling pathways associated with
hypericin-based PDT and whether their inhibition is of therapeutic benefit and leads to an improvement
of the tumoricidial effects of PDT. In addition the potential participation of autophagy as in vivo
response to PDT will be assessed (UG1, UG2, ULG1).