WO2010070379A1 - Rxr-ppargamma agonist/growth factor inhibitor combination therapy for inducing apoptosis and the treatment of cancer - Google Patents

Abstract

The present invention relates to novel anti-cancer treatments and compositions using combinations of at least one agonist of the RXR-PPARγ heterodimer complex selected in the group consisting of a RXRα agonist, a RXRβ agonist, a RXRγ agonist, a PPARγ agonist or a RXR-PPARγ heterodimer agonist and a growth factor inhibitor.

Description

RXR-PP ARγ agonist/Growth Factor inhibitor combination therapy for inducing apoptosis and the treatment of cancer

The present invention relates to novel anti-cancer treatments and therapeutics using combinations of a RXR-PP ARγ agonist and a growth factor inhibitor. In particular the present invention relates to combinations of a rexinoid or PP ARγ agonist and a MAPK pathway inhibitor and their use for the prevention and/or the treatment of cancer or a tumour. In addition the present invention relates to a method to determine the effects of a RXR-PP ARγ agonist upon a cell line.

Retinoids bind to a heterodimer composed of retinoic acid (RARα, β or γ) and retinoid X receptors (RXRα, β or γ) and initiate the sequence of events leading to transcription activation by inducing allosteric alterations in a network of energetically coupled residues that ultimately alter the structure of interaction surfaces with secondary effectors, such as co-regulators (de Lera et al., 2007; Shulman et al., 2004). In the absence of a ligand or the presence of certain antagonists, the

RXR-RAR heterodimer interacts with co-repressor complexes which are released upon RAR agonist binding, thus allowing recruitment of co-activator containing complexes and the transcriptional machinery.

The anti-cancer activity of retinoids is well established and the induction of differentiation and/or apoptosis are hallmarks of there therapeutic potential (Altucci and Gronemeyer, 2001; Altucci et al., 2007; Minucci and Pelicci, 2007; Soprano et al., 2004). Clinical studies and experiments with the Ptchl^+/" mouse model indicate that retinoids can also act in a cancer-preventive manner in the cases of basal cell carcinoma (Orlandi et al., 2004; So et al., 2008; So et al., 2004) and oral cancers (Lippman et al., 2005; Sun and Lotan, 2002). Retinoids act through binding to the RAR partner of RAR-RXR heterodimers, but RXR is a promiscuous heterodimerization partner for multiple nuclear receptors.

In contrast RXR-selective agonists ("rexinoids") alone are unable to dissociate co-repressor and activate transcription, a phenomenon commonly referred to as RXR subordination, but a rexinoid can synergize with the retinoid to enhance the RXR-RAR-mediated transcription through cooperative recruitment of co-activators (Germain et al., 2002).

Rexinoids (de Lera et al., 2007) display lower toxicity than retinoids and are increasingly recognized for their cancer therapeutic and preventive potential (Gottardis et al., 1996; Li et al., 2007; Liby et al., 2007a; So et al., 2008; Wu et al., 2002). Data from cellular and animal models as well as clinical trials reveal the potent antitumor activity of rexinoids which are less toxic than retinoids, thus favouring their use in vivo (Altucci et al., 2007).

The rexinoid, bexarotene, has been approved for the treatment of refractory cutaneous T cell lymphoma (Altucci et al., 2007; Liby et al., 2007b). Rexinoids have attracted significant attention as recent molecular and mouse genetic studies continue to provide mechanistic insight.

Moreover, in recently completed phase III clinical trials combining bexarotene with chemotherapy revealed that a subgroup of non-small-cell lung cancer patients may benefit from the addition of bexarotene to the therapeutic scheme (Blumenschein et al., 2008; Ramlau et al., 2008).

In addition to there use in cancer chemotherapy, studies with transgenic MMTV-erbB2 and A/J mouse carcinogenesis models have revealed a potential for rexinoids to prevent cancer in both lung and in mammary gland tissue(Li et al., 2007; Liby et al., 2007a; Liby et al., 2007b).

In addition to interacting with RAR, RXR can heterodimerize with a variety of other nuclear receptors such as PPARs (peroxisome proliferator-activated receptors), VDR (Vitamin D Receptor), TR (Thyroid Hormone Receptor) or orphan receptors (an orphan receptor is an apparent receptor that has a similar structure to other identified receptors but whose endogenous ligand has not yet been identified).

In RXR-VDR or RXR-TR, RXR is also subordinated to its partner but in permissive heterodimers, such as PPAR-RXR, transcription can be activated by both liganded partners and rexinoids which operate autonomously (de Lera et al., 2007). The inventors have previously provided evidence that RXR is a therapeutic target for acute promyelocytic leukemias due to their suppression of RARα fusion-mediated transformation, possibly as consequence of rexinoid-induced apoptosis (Zeisig et al., 2007).

In order to understand the molecular basis of the apparent antitumor activity of rexinoids the inventors have studied signalling pathways by which rexinoids can act despite the subordination of its receptor and have uncovered two distinct mechanisms.

The first, involves the induction of differentiation and post- maturation apoptosis originally observed in acute promyelocytic leukemia (APL) cells in the presence of elevated cAMP levels (Benoit et al., 1999). This rexinoid-cAMP synergy was seen even in all-trans retinoic acid (ATRA)-resistant APL cells indicating a mechanism of action distinct from that of retinoids. Even more importantly, rexinoid-cAMP co-treatment induced post-maturation apoptosis also in non-APL acute myeloid leukaemia (AML) cells and patients' blasts that are entirely unresponsive to ATRA. Mechanistically, the increased cAMP level exerts a dual effect, it induces "desubordination" of RXR in the RXR-RAR heterodimer, thus activating the cognate gene program which itself synergizes with the cAMP- dependent program to induce AML differentiation and apoptosis. Interestingly, apoptosis involves the simultaneous induction of the tumor-selective TNF-related apoptosis — inducing ligand (TRAIL) and its cognate receptor (Altucci et al., 2005). The second signalling pathway was discovered when growth factor support of AML cells growing in the presence of rexinoid agonists was lowered by decreasing the serum concentration. Under such conditions the rexinoid alone induced rapid cell death referred to as "autonomous rexinoid apoptosis", as it did not involve co-activation of other pathways and resulted in apoptosis without prior induction of cell differentiation (Benoit et al., 2001 ).

While there are divergent reports from mouse genetic analyses on the possibility that PP ARy may act in some tissues as tumor suppressor (Cui et al., 2002; Koeffler, 2003; McAlpine et al., 2006; Saez et al., 2003), there is no controversy regarding the antiproliferative capacity of the activated RXR-PP ARγ heterodimer against a great number of transformed cells and the potential of cognate agonists for adjuvant and chemopreventive treatments of several human leukemias and cancers, including those of colon, prostate and breast (Elrod and Sun, 2008). The inventors have developed a new anti-cancer composition which acts via a previously unknown apoptotic mechanism involving the RXR-PP ARγ heterodimer and the inhibition of growth factor inhibition of this unknown apoptotic mechanism. Therefore according to a first aspect of the present invention there is provided a composition comprising: at least one agonist of the RXR-PP ARγ heterodimer complex selected in the group consisting of a RXRα agonist, a RXRβ agonist, a RXRy agonist, a PP ARγ agonist or RXR-PP ARγ heterodimer agonist; and a growth factor inhibitor; for the treatment of cancer or a tumour.

Hereafter, whatever the agonist, it is called RXR-PP ARγ agonist. The inventors have now found that rexinoid apoptosis is not mediated by (non-permissive) RXR-RAR but by permissive ligand-activated RXR- PP ARγ heterodimers and proceeds through the intrinsic death pathway upon NO production and amplification through an activation loop involving p53 and the TRAIL death receptor DR5. IGF (insulin like growth factor) and EGF (epidermal growth factor) antagonize rexinoid apoptosis through IGF receptor 1 whose signaling can be abrogated by IGF receptor kinase inhibitors or inhibitors of downstream MAP (mitogen-activated protein) kinase signaling. This mechanistic analysis provided the basis for the development of antiproliferative and pro-apoptogenic treatments of leukemic cells and acute myeloid leukemia patients blasts, as well as breast and colon cancer cells in vitro. Finally, the inventors demonstrate the efficacy of rexinoid apoptosis in the treatment of human colon carcinoma xenografts in mice in vivo. The inventors therefore have developed a new class of anticancer therapeutic composition comprising at least one agonist of one of the RXR isotypes α, β, γ and/or of PP ARγ and an inhibitor of at least one growth factor which is antagonistic to the RXR- PP ARγ mediated apoptosis mechanism they have discovered and which the composition acts through. A number of RXRα, RXRβ, RXRy agonists such as rexinoids,

PP ARγ agonists such as benzofibrate and growth factor inhibitors such as trastuzumab (herceptin®) are already in clinical trials or have been cleared for therapeutic use and therefore this should allow rapid testing of the novel combination compositions described in the present Patent Application.

This composition is intended for use in methods of treating cancer such as chemotherapy either as the principle agent or as part of a more complex multi- therapeutic scheme. Alternatively this composition could be used as a preventative measure to suppress cancer cell proliferation in at risk patients.

Preferably the agonist is a pan-agonist of RXR α, β and γ. Pan- agonists of RXR are agonists for all three RXR isotypes; examples include the rexinoid BMS749 (see below). A pan-agonist has the advantage of acting upon any heterodimeric RXR-PPARγ complex, irrespective of the recruited RXR isotype.

Alternatively the agonist is an exclusive agonist of RXR α, β and γ. Isotype specific modulators are known for Retinoic Acid Receptors and therefore isotype specific agonists are also expected to exist for each RXR isotype. Such isotype specific agonists to allow more specific modulatory/agonistic effects upon the specific

RXR isotype/ PPARγ complex.

Preferably the agonist is a rexinoid.

Rexinoids are agonistic ligands of the RXR (Retinoid X Receptor), but not the retinoic acid receptor. No general chemical formula exists representing rexinoids, as these molecules are defined functionally, although they are generally chemically related to vitamin A (retinol). Recent work has however better defined the ligand binding characteristics of the RXRα binding site (de Lera et al., 2007) and therefore rexinoids also must have the required chemical properties to interact with the RXR ligand binding site. The inventors have found that the apoptotic effect of rexinoid results from the agonistic effect of these molecules upon the RXR-PPARγ heterodimer. In particular the rexinoid is of the formula (i):

Rexinoid General Formula i

Wherein the X₀ — X₀ bond may be a single or double carbon bond; and in which groups Ri, R₂, R₃ and R₄ are:

Table I. List of Ri, R₂, R₃ and R₄ chemical groups

In particular the inventors have found that the rexinoid BMS749 (formula iv), also known as BMS230J49 (C₂₅H₂₈O₂ MW: 360,489), can result in rapid and specific apoptosis of cancer cells. Many other suitable rexinoids exist and are encompassed by the present invention such as SRl 1237(formula v), AGN194,204 (formula vi), LGD 100,268 (formula vii), CD3254 (formula viii), LGD 1069 (formula ix), Methoprene acid (formula x), Docosahexaenoic acid (formula xi), Phytanic acid (formula xii), DHA (formula xiii), SRl 1346 (formula xiv), CD3254 (formula xv), 3- Me-TTNPB (formula xvi), SRl 1251 (formula xvii), CD2624 (formula xviii), LG100754 (formula xix), LG100268 (formula xx), LG101506 (formula xxi), compounds of formula xxii with the chemical groups shown in Table I, or compounds of formula xxiii with the chemical groups shown in Table I. Representative chemical formulae are provided below for these compounds.

BMS749 (Formula iv)

SRl 1237 (BMS649) (Formula v)

AGN 194,204 (Formula vi)

LGD 100,268 (Formula vii)

CD3254 (Formula viii)

LGD 1069 (Formula ix)

Methoprene acid (Formula x)

Docosahexaenoic acid (Formula xi)

Phytanic acid (Formula xii)

DHA (Formula xiii)

SRl 1346 (Formula xiv)

CD3254 (Formula xv)

3-Me-TTNPB (Formula xvi)

SRl 1251 (Formula xvii)

CD2624 (Formula xviii)

LG 100754 (Formula xix)

LG 100268 (Formula xx)

LG101506 (Formula xxi)

Some RXR-selective compounds of formula xxii and of formula xxiii are described in Dawson et al. (2004). Alternatively the agonist is an agonist of PP ARγ.

Several classes of chemical molecules and compounds are known to be agonists of PP ARγ. These include, molecules comprising a thiazolidinedione group (formula xxiv) (known as thiazolidinediones and also known as glitazones), long chain fatty acids and prostanoids. In addition compounds are also known which are agonists of all

PPAR isotypes, such as perfluorononanoic acid. The use of such compounds in compositions or methods are also encompassed by the present invention.

Thiazolidinedione (Formula xxiv), wherein R₇ comprises a carbon chain as shown in formulae xxvi, xxvii, xxviii or xxix.

Examples of PP ARγ agonists include compounds such as benzofibrate (Bezalip/Difaterol/Bezatol/Bm 15075) (formula xxv); 9,10-dihydro-15- deoxy-Δ^12>14-PGJ₂ (9,10-dihydro-15-deoxy-Δ¹²'¹⁴-Prostaglandin J₂) (formula xxx); ciglitazone (formula xxvi); rosiglitazone (formula xxvii) and pioglitazone (formula xxviii).

Benzofibrate (Formula xxv)

Ciglitazone (Formula xxvi)

Rosiglitazone (Formula xxvii)

Pioglitazone (formula xxviii) PP ARγ agonists, represent a further new class of anti-cancer therapeutic agents when administered in the form of a composition comprising a growth factor inhibitor. In particular the inventors have also found that the PPARγ agonist troglitazone (a thiazolidinedione, formula xxix) can have an apoptotic effect upon cancer cells in the presence of a growth factor inhibitor.

Troglitazone (Formula xxix)

In particular the PPARγ agonist is a thiazolidinedione and is selected from the group comprising: Rosiglitazone; Pioglitazone; Troglitazone, Ciglitazone.

Alternatively the PPARγ agonist is 15-deoxy-Δ-12,14-prostaglandin J2 (formula xxvi) a prostanoid.

9,10-dihydro-15-deoxy-Δ¹²'¹⁴-PGJ₂ (Formula xxx) In particular the agonist, may be an agonist of RXR-PP ARγ- heterodimer, compounds have been reported which are RXR-PP ARγ heterodimer selective agonists for instance LGl 00754 (formula xvi). Unlike other rexinoids such as LGl 00268 (formula xvii) which are promiscuous agonists of any/multiple heterodimeric complexes comprising RXR, compounds such as LG 100754 have been demonstrated to exclusively activate the RXR-PP ARγ complex (Cesario et al, 2001). Such RXR-PP ARγ specific agonists would therefore be expected to have lower toxicity than more general RXR or PPARγ agonists which would act upon several undesired targets in addition to the RXR-PP ARγ complex. In particular the growth factor inhibitor included in the composition is a general inhibitor.

The inventors have found that general inhibitors of growth factor activity, such as UO 126 (Formula xxxi) which affects several growth factor activated signalling pathways such as the MEK 1/2 pathway as well as Raf, ERK, JNK, MEKK, MKK-3, MKK-4/SEK, MKK-6, AbI, Cdk2 and Cdk4 (Duncia, et al, 1998), when administered in combination with a RXR-PP ARγ agonist leads to apoptosis of cancer cells.

UO 126 (Formula xxxi)

A significant advantage of using a general growth factor inhibitor is that in combination with the RXR-PP ARγ agonist, such a composition would have a therapeutic activity against a large number of tumours and cancer types.

Alternatively the growth factor inhibitor is a specific inhibitor. Several specific growth factor inhibitors are now known and a few have been approved for clinical study or therapeutic use. An example of a specific growth inhibitor is trastuzumab (Herceptin®), which is a humanized monoclonal antibody that acts on the HER2/neu (erbB2) receptor. Such specific growth factor inhibitors are expected to have lower levels of toxicity than more general inhibitors and hence a composition comprising a combination of one of these together with a RXR-PP ARγ agonist is a particularly preferred embodiment of the present invention.

A growth factor inhibitor can directly affect the growth factor, for instance by binding thereto and preventing it acting upon its cognate receptor. Or alternatively, the growth factor inhibitor can act upon one or more components which are in turn activated either directly or as part of a cascade by the activity of the growth factor.

In particular the growth factor inhibitor is an antagonist to a receptor for the growth factor. Generally growth factors are secreted proteins which induce cellular growth and division. Growth factors normally act via one or more permissive cell surface receptors which in turn causes the activation of an intracellular signalling cascade which acts upon target intracellular components and has the desired effect. Recently much effort has been made to find antagonists of these growth factor receptors and several humanized antagonistic antibodies to the growth factor receptor have been generated such as trastuzumab (Herceptin ®) for ERBB2. With reference to Table 2 and 6 below, several other examples are provided. In addition to antagonistic antibodies growth factor mimetic proteins, peptides or other types have also been developed which antagonise the action of a growth factor upon its receptor and hence inhibit growth factor activity, several examples are listed in Table 3 below.

Alternatively the growth factor inhibitor is a inhibitor which blocks at least one enzymatic activity of the growth factor.

In addition to acting via a receptor/signalling cascade, several growth factors also act directly upon one or more specific targets or do so in combination with their receptor or another partner. They act upon their targets via enzymatic domains present in the growth factor/receptor/partner.

An example of this class of growth factor inhibitor is Gefitinib (IRESSA), which is a small molecule inhibitor which binds to the Epidermal Growth Factor Receptor tyrosine kinase domain at the adenosine triphosphate (ATP)-binding site of the enzymatic domain. This prevents the domain from functioning and hence stops the activation of the Ras signal transduction cascade. Other examples are given in Table II below.

Alternatively the growth factor inhibitor may inhibit at least one target of the growth factor.

As noted above as well as acting directly upon the growth factor or its receptor/partner, a growth factor inhibitor may also act upon one or more of the downstream factors activated by the growth factors, such as the MAP kinases pathway. An example of such an inhibitor is UO 126 (formula xxxi). In particular the growth factor inhibitor is one which inhibits the activities of at least one of the growth factors selected from the group comprising: IGFI; IGFII; EGF. From a non-exhaustive panel of growth factors tested including Insulin, G-CSF (granulocyte-colony stimulating factor), IL3 (interleukin 3), IL4 (interleukin 4), ILlO (interleukin 10), SCF (stem cell factor) and GM-CSF (granulocyte/Macrophage colony stimulating factors), the inventors have found that IGFI, IGFII and EGF can prevent the apoptotic activity of an RXR-PP ARγ agonist on otherwise sensitive cells. Inhibitors which affect these specific growth factors are preferred embodiments of the present invention.

UO 126 is known to affect the activity of IGFI, IGFII and EGF and therefore it is a preferred inhibitor of IGFI, IGFII and EGF activity. Other examples of IGF inhibitors include the following:

Table II. Examples of IGF inhibitors in preclinical/clinical trials

Table III. Examples of EGF inhibitors in preclinical/clinical trials

Tables II and III are adapted from 'Ηartog H, Wesseling J, Boezen HM, van der Graaf WT. 2007. The insulin-like growth factor 1 receptor in cancer: old focus, new future. Eur J Cancer. 43(13):1895-904" and "Grosios, K., Traxler, P. 2003. Tyrosine kinase targets in drug discovery. Drugs Fut, 28(7): 679". In particular the growth factor inhibitor inhibits the MAP kinase pathway.

The inventors have shown that the action of growth factors upon the MAP kinase pathway inhibit the newly discovered apoptotic mechanism and therefore growth factor inhibitors which affect the MAP kinase pathway and specifically components thereof constitute preferred embodiments of the present invention. UO 126 (formula xxxi) is known to affect the MAP Kinase pathway and therefore it is a preferred inhibitor of the MAP Kinase pathway.

Several other compounds are inhibitors of the MAP Kinase pathway such as Arctigenin (formula xxxii) a potent inhibitor of MKKl (MAP Kinase Kinase 1), PD 98059 (formula xxxiii) a potent inhibitor of MAP Kinase Kinase, SB202190 (formula xxxiv) an inhibitor of p38 Map Kinase, SL 327 (formula xxxv) an inhibitor ofMEKl and MEK2.

Arctigenin (Formula xxxii)

PD 98059 (Formula xxxiii)

SB202190 (Formula xxxiv)

SL 327 (Formula xxxv) In particular the growth factor inhibitor inhibits a component of the

MAP kinase pathway selected from the group comprising: MEK 1 ; MEK 2; p42; p44. As detailed above and in greater detail in this Patent Application, the newly discovered apoptotic mechanism appears to involve NO production in the targeted cell, the listed components of the MAP kinase pathway lead to NO production and hence the inhibition of one or more of these specific components is a preferred embodiment of the present invention.

Many other inhibitors of MAP Kinase and MAP Kinase targets/partners are known and further examples include:

- Apigenin, a plant flavonoid that acts as a potent inhibitor of p42/p44 MAP kinase;

- ERK Activation Inhibitor I, a cell-permeable, stearated 13-amino acid peptide corresponding to the N-terminus of MEKl that acts as a specific inhibitor of ERK activation, it selectively binds to ERK2 and prevents its interaction with MEK (IC₅₀ = 2.5 mM); - ERK Activation Inhibitor II, a cell-permeable, 13-amino acid peptide corresponding to the N-terminus of MEKl that is fused to the HIV-TAT membrane translocating peptide (MTP) sequence via a glycine linker, this acts as a specific inhibitor of ERK activation that binds to ERK2 and prevents its interaction with MEK (IC₅₀ = 210 nM);

- Hsp25 Kinase Inhibitor, a potent and selective inhibitor of mammalian HSP25 kinase (also known as mitogen-activated protein kinase activated protein kinase-2. Inhibition is competitive with respect to the substrate peptide (Kj = 8.1 mM) and non-competitively with respect to ATP (Kj = 134 mM). Other inhibitors of MAP Kinase are encompassed by the present invention.

In particular the composition may further comprise at least one component selected from: a RAR antagonist; a DR5 agonist such as TRAIL or a functional mimic of TRAIL, which activates the DR5 receptor and activates the formation of the death inducing signalling complex.

The inventors have found that the apoptotic mechanism they have discovered can be further enhanced by the administration of RAR antagonists and/or DR5 agonists such as TRAIL.

RAR antagonists act to prevent RAR-RXR dimer formation and hence promote instead RXR-PP ARγ dimer formation. This in turn allows the administered RXR-PP ARγ agonist to have a greater effect upon the targeted cell. Examples of RAR antagonists include Ro 41-5223 (formula xxxvi) and BMS 195,614 (formula xxxvii), CD2665 (formula xxxviii) (de Lera et al, 2008 (supplementary materials S4).

Ro 41-5223 (Formula xxxvi)

BMS 195,614 (Formula xxxvii)

CD2665 (Formula xxxviii)

DR5 (death receptor 5) is the cognate receptor of TRAIL and the inventors have shown that enhanced DR5 activity mediated by TRAIL increases the RXR-PPARγ apoptotic pathway. Additional DR5 agonists include staurosporine (formula xxxix) and CCCP (carbonyl cyanide m-chlorophenylhydrazone).

Staurosporine (Formula xxxix)

Hence in a preferred embodiment, the present invention involves the co-administration of a composition comprising a DR5 agonist such as TRAIL and/or an RAR antagonist such as CD2665 together with at least one agonist of a RXR isotype and/or an agonist of PPARγ agonist and/or an agonist of RXR-PP ARγ; and a growth factor inhibitor.

In particular the composition comprises both a rexinoid and a PPARγ agonist. The inventors have shown that compositions comprising both a rexinoid and a PPARγ agonist present a strong synergistic effect.

In particular the composition may be used to treat a cancer selected from the group: lung, ovarian, central nervous system, skin, colon or leukaemia.

In accordance with a second specific aspect of the present invention there is provided an in vitro method to determine the effects of administration of at least one agonist of RXRα, RXRβ, RXRy or PPARγ; upon apoptosis in at least one cell, comprising the steps: a) culturing said at least cell under conditions sufficient for said at least one cell to survive; b) supplementation of the cell culture of step a) with a growth factor inhibitor and the measurement of the cellular apoptosis level; c) supplementation of the cell culture of step b) with at least one agonist of RXRα, RXRβ, RXRγ or PPARγ; and the measurement of the cellular apoptosis level; d) comparing the measured apoptosis levels of step b) and step c) and determining the effects of said at least one agonist of RXRα, RXRβ, RXRγ or

PPARγ; upon said at least one cell.

The inventors have found that cancer cell line sensitivity to agonists of RXRα, RXRβ, RXRγ or PPARγ agonists can be assessed for the effects of such agonists, even for cells which do not survive on serum free (and hence growth factor deficient) media, by selectively removing the effects of some growth factors using growth factor inhibitors.

This method therefore allows a preliminary therapeutic evaluation of a specific agonist such as a rexinoid, upon a cancer cell line or an initial toxicity screen upon a healthy or cancer cell line. This method allows a worker to decide whether to proceed with a more detailed evaluation of the proposed anti-cancer therapeutic for use in a composition for instance as per the first aspect of the present invention.

Growth factor inhibition can be specific using agents such as trastuzumab or more general using an agent such as UO 126 or any of the other inhibitors described herein or specifically in Tables 2 and 3.

According to a third aspect of the present invention there is provided a kit to determine the effects of administration of at least one agonist of RXRα, RXRβ, RXRγ or PPARγ upon apoptosis at least one cell, comprising: a growth factor inhibitor; at least one agonist of RXRα, RXRβ, RXRγ or PPARγ; instructions.

The inventors also provide a kit to allow a user to evaluate the sensitivity of a cell line to at least one agonist of RXRα, RXRβ, RXRγ or PPARγ. In particular this kit can be used to perform the method according to the second specific aspect of the present invention. The invention also concerns products containing a RXR-PPARγ heterodimer complex agonist as defined above and a growth factor inhibitor as a combined preparation for simultaneous, separate or sequential use in cancer therapy.

The invention also concerns products containing a RXR-PPARγ heterodimer complex agonist as defined above, a growth factor inhibitor, and a RAR antagonist and/or TRAIL or a functional mimetic thereof as a combined preparation for simultaneous, separate or sequential use in cancer therapy.

In addition to a composition comprising both at least one agonist of

RXRα, RXRβ, RXRγ or PPARy or RXR-PPARγ heterodimer and a growth factor inhibitor, the present invention also encompasses compositions which can be used for the sequential administration of these substances to an individual in need thereof. Due to several clinical or logistical reasons, it can be better to administer components of a treatment regimen individually rather than as a single composition. Reasons include such factors as toxicity of the components and differences in absorption/diffusion rates of the components .

The present invention also relates to methods of treating a tumour or cancer using a composition comprising; at least one agonist of RXRα, RXRβ, RXRγ or PPARy; and a growth factor inhibitor. The present invention also relates to the preparation of medicaments for the treatment of a tumour or cancer using at least one agonist of RXRα, RXRβ, RXRγ or PPARγ; and a growth factor inhibitor.

Examples of pharmaceutical compositions include any solid (tablets, pills, capsules, granules, etc.) or liquid (solutions, suspensions or emulsions) with suitable composition or oral, topical or parenteral administration, and they may contain the pure compound or in combination with any carrier or other pharmacologically active compounds.

These compositions may need to be sterile when administered parenterally. Administration of the compounds or compositions of the present invention may be by any suitable method, such as intravenous infusion, oral preparations, intraperitoneal and intravenous administration. Infusion times of up to 24 hours are used, more preferably 2-12 hours, with 2-6 hours most preferred. Short infusion times which allow treatment to be carried out without an overnight stay in hospital are especially desirable. However, infusion may be 12 to 24 hours or even longer if required. Infusion may be carried out at suitable intervals of say 2 to 4 weeks.

Pharmaceutical compositions containing compounds of the invention may be delivered by liposome or nanosphere encapsulation, in sustained release formulations or by other standard delivery means.

The correct dosage of the compounds will vary according to the particular formulation, the mode of application, and the particular situs, host and tumour being treated. Other factors like age, body weight, sex, diet, time of administration, rate of excretion, condition of the host, drug combinations, reaction sensitivities and severity of the disease shall be taken into account. Administration can be carried out continuously or periodically within the maximum tolerated dose. Definitions

- Agonist is an endogenous substance or a drug that can interact with a receptor and initiate a physiological or a pharmacological response characteristic of that receptor.

- Antagonist is a drug or a compound that opposes the physiological effect of another. It is a chemical entity that opposes the receptor-associated response normally induced by another bioactive agent.

- Cancer is a class of diseases in which a group of cells display uncontrolled growth (division beyond the normal limits), invasion (intrusion on and destruction of adjacent tissues), and sometimes metastasis (spread to other locations in the body via lymph or blood). These three malignant features of cancers differentiate them from benign tumors, which are self-limited, do not invade or metastasize. Most cancers form a tumor but some, like leukemia, do not. In the present invention cancer is meant to refer to both malignant and benign tumoural growths and tumour free cancers. - Co-activators are proteins that cooperate with nuclear hormone receptors to activate transcription. Two classes are known: the pi 60 family, which recruits histone acetyltransferases, and the TRAP-DRIP-SMCC complex, which is thought to interact with the basal transcription machinery.

- Fas-Associated protein with Death Domain (FADD) is an adaptor molecule that bridges the Fas-receptor, and other death receptors, to caspase-8 through its death domain to form the death inducing signaling complex (DISC) during apoptosis.

- FCS, Fetal calf serum is serum taken from the fetuses of cattle. Fetal bovine serum (FBS) is the used serum in the culturing of eukaryotic cells. The globular protein, bovine serum albumin (BSA), is a major component of fetal bovine serum but in addition it also comprises growth factors, hormones and nutrients. The rich variety of proteins in fetal bovine serum maintains cultured cells in a medium in which they can survive, grow, and divide.

- General growth factor inhibitor, is a substance which inhibits the effects of more than one growth factor normally by interfering with the activity of more than one growth factor receptor/pathway. An example of a general growth factor inhibitor is U0126 which affects both MEK 1/2 as well as Raf, ERK, JNK, MEKK, MKK-3, MKK-4/SEK, MKK-6, AbI, Cdk2 and Cdk4.

- Trastuzumab is a humanized monoclonal antibody that acts on the HER2/neu (erbB2) receptor. Trastuzumab principal use is as an anti-cancer therapy in breast cancer in patients whose tumors over-express this receptor.

- Gefitinib is an epidermal growth factor receptor tyrosine kinase (EGFR-TK) inhibitor.

- Long chain fatty acids, are fatty acids with aliphatic tails of 16 carbons or more. - L-NMMA is L-N^G-monomethyl Arginine

- MAP kinase pathway: the MAPK/ERK pathway is a signal transduction pathway that couples intracellular responses to the binding of growth factors to cell surface receptors. This pathway is very complex and includes many protein components. - Prostanoids are a subclass of eicosanoids consisting of: the prostaglandins, the thromboxanes and the prostacyclins. - Retinoid/Rexinoid: a detailed analysis of retinoid/rexinoid structure and chemical properties can be found in de Lera et al., 2007.

- Stereoisomers, are different forms of molecule comprising the same chemical groups but arranged in a variety of combinations. This relative spatial arrangement of atoms within molecules is called the stererochemistry of the molecule. Alternatively arranged molecules of the various formulae provided herein, stereoisomers, explicitly form a part of the present invention irrespective of whether these are enatiomers or diastereomers.

- Specific growth factor inhibitor, is a substance which exclusively affects a single growth factor receptor or pathway. Examples of specific growth factor inhibitors are trastuzumab and gefitinib.

- tBid (truncated BH3 interacting domain death agonist) is a truncated form of the pro-apoptotic full-length BID is generated by Caspase-8 cleavage of BID. The truncated form of the protein translocates from the cytosol to mitochondria and transduces apoptotic signals.

- TRAIL (TNF-related apoptosis inducing ligand) is a homotrimeric type II membrane protein that rapidly induces oligomerization of receptor intracellular death domains and apoptosis in a variety of tumor cells.

- UO 126, is an inhibitor of both MEKl and MEK2, a type of MAPK/ERK kinase. UO 126 has been found to functionally antagonize AP-I transcriptional activity via noncompetitive inhibition of the dual specificity kinase MEK. U0126 is also a weak inhibitor of PKC, Raf, ERK, JNK, MEKK, MKK-3, MKK-4/SEK, MKK-6, AbI, Cdk2 and Cdk4.

For a better understanding of the invention and to show how the same may be carried into effect, there will now be shown by way of example only, specific embodiments, methods and processes according to the present invention with reference to the accompanying drawings in which:

- Figure 1. The rexinoid BMS749 induces PLB985 cells apoptosis via a caspase-dependent process. A. BMS749 (formula iv) induces a dose-dependent apoptosis while the RARα agonist BMS753 has no effect. The cells were treated for 24 hours with the 2 ligands at the indicated concentration and assayed for apoptosis by apo2.7 staining. On the right panel, FACScan analysis of Apo2.7 staining. B. The rexinoid treatment induces the cleavage of caspase-3, -9, -8, PARP and Bid as revealed by western blot experiments. The PLB985 cells were treated with BMS749 and collected at the indicated time for protein extraction. The full length (FL) or cleaved fragments are indicated. C. Caspase inhibitors reduces BMS749-induced apoptosis. The PLB985 cells were pretreated for 2 hours with lOOμM Z-VAD-FMK (pan inhibitor), Z-IETD-FMK (caspase 8 inhibitor), Z-AEVD-FMK (caspase 10 inhibitor) or a combination of them prior addition of BMS749. After 48 hours of treatment, the sub-Gl fractions representing apoptotic bodies were analysed by PI staining and FACS analysis. - Figure 2. The death receptor pathway contributes partially to rexinoid-induced apoptosis. A. RPA analysis of PLB985 cells treated for various periods of time with BMS749. B. Western blot analysis of TRAIL and DR5 protein expression. The PLB985 cells were treated for the indicated time with BMS749. The non-specific band (NS) confirms equal loading. C. and D. PLB985 cell lines stably expressing a GFP-tagged dominant negative FADD (GFP-F ADDdn) fusion protein or a β-galactosidase protein (LacZ) were generated by lentivirus-mediated infection of the PLB985 wild type (WT) cells. The expression of the fusion protein was verified by western blot (C.) or by GFP fluorescence using FACS analysis (B.). E. PLB985 cells expressing the GFP-F ADDdn protein are less sentitive to rexinoid-induced apoptosis. WT₅ Lac Z or GFP-F ADDdn PLB985 cells were treated for 24 hours with BMS749. TRAIL treatment (20ng/ml) was used as positive control for FADD dominant negative effect. Staurosporine (STS, 10OnM) and CCCP (2μM) are targeting the mitochondrial apoptotic pathway and were used as negative control.

- Figure 3. Crucial role of the mitochondial death pathway in rexinoid-induced PLB985 cells death. A. BMS749 induces loss of mitochondrial transmembrane potential. The PLB985 cells were treated for 24 hours with lμM BMS749 stained with DiOC6 and analysed by flow cytometry. B. Western blot analysis of cytochrome c and Smac/Diablo release from the mitochondria to the cytoplasm. The cells were treated with BMS749 and were collected at the indicated time for mitochondrial and cytoplasmic protein extraction. C. and D. PLB985 cells were stably transfected with pIRES EGFP or pIRES Bcl-2 EGFP vectors by using nucleofection technology. Western blot of bcl-2 (C.) and GFP detection by FACS analysis (D.) confirmed the expression of the transgene. E. Bcl-2 overexpression decreases rexinoid-induced apoptosis. WT, pIRES or pIRES Bcl-2 GFP expressing cells were exposed to BMS749 at the indicated concentrations for 30 hours and assessed for apoptois by Apo2.7 detection. - Figure 4. Nitric oxide (NO) production is required for rexinoid- induced apoptosis. A. Time course of NO production in PLB985 cells treated with BMS749 and detected by FACScan analysis. Right hand panel shows staining of the cells with DAF-FM, fluorescence is produced as a result of interaction of DAF-FM with NO. B. FACScan analysis revealing the production of NO. The cells were pre- treated or not with 500 μM of the pan NO synthase inhibitor L-NMMA, then exposed for 2 hours to BMS749 and prepared for flow cytometry. C. The pan NO Synthase inhibitor decreases rexinoid-induced apoptosis. PLB985 cells were pre-treated for 2 hours with L-NMMA (200 and 500μM) and 2 antioxidant GSH (reduced gluthatione at 1 and 5 mM) and L-NAC (N-Acetyl cysteine at 1 and 5 mM) prior exposition to BMS749 for 24 hours. The apoptosis was assessed by FACScan detection using Apo2.7 antibody labelling. D. NO production induces cytochrome c release as demonstrated by this western blot analysis of cytochrome c release. PLB985 cells were pre-treated for 2 hours with L-NMMA (200 and 500μM), or with 2mM of GSH or L-NAC. BMS749 was added to the culture for 4 hours before collecting the cells for fractionated protein extraction. E. BMS749 induces iNOS and the phosphorylation of eNOS enzymes as shown by western blotting. F. PLB985 cells were transfected with eNOS and iNOS specific siRNA and then exposed to BMS749. Apoptosis was measured by Apo2.7 staining by FACS. G. BMS749 causes the activation of p53 as detected by western blot of its Serl5 phosphorylation, which is blocked by treatment with L-NMMA.

- Figure 5. IGFI, IGFII and EGF rescue PLB985 cells from rexinoid-induced apoptosis. A. Serum factors inhibit rexinoid-induced cell death. The cells were treated for 24 hours with BMS749 and increasing amount of serum and assayed for apoptosis by Apo2.7 detection. B. PLB985 cells were co-treated with BMS749 and lμg/ml of several growth factors as indicated. After 48 hours, the cells were submitted to apoptosis detection by FACSscan analysis of Apo2.7 labelling. C. Blocking the IGF signalling by using IGF neutralising antibodies (anti-IGF) inhibits NO production. D. EGF and IGFI inhibit apoptosis by blocking cytochrome c release, as revealed by western blot. PLB985 cells were pre-treated for 2 hours with EGF and IGFI and exposed for 4 hours to BMS749. E. MEK-1/2 inhibition reverses EGF induced survival. PLB985 cells were pre-treated for 2 hours with lOμM SB203580 (p38 inhibitor), lOμM LY294002 (PI3 Kinase inhibitor) and 5 μM UO 126 (MEK inhibitor). The cells were then exposed to a combination of BMS749 and EGF or IGF and assessed for apoptosis by Apo2.7 detection after 24 hours of treatment. F. Kinetics of p42/44 phosphorylation under EGF exposure as assessed by western blot. G. In vitro kinase assay of p42/44 activity performed with PLB985 cell extracts treated or not for 1 hour with IGFI, EGF or 10% FCS. EIk-I is used as p42/44 target. H. Production of NO is controlled by the growth factor pathway. PLB985 cells were treated with IGFl, 10% FCS or 10% FCS +U0126 and NO production was assessed by flow cytometry.

- Figure 6. p42/44 MAP Kinase pathway mediates the survival signal counteracting rexinoid apoptosis through IGF-RI activation. A. Growth factors enhance the autophosphorylation of IGF-IR. PLB985 cells were treated with IGF-I, EGF and FCS for Ih, lysed and subjected to immunoprecitation with anti phosphor IGF-IR antibody. The upper panel indicates the levels of non-phosphorylated IGF-IR. B. p42/44 kinase activity assay performed on protein extract of PLB985 cells exposed to the indicated treatments. U0126 was used at 5μM and AG1024 at 2μM. EIk-I is used as p42/44 target. C. PLB985 were pre-treated with 2 tyrosine kinase inhibitors, the tyrphostin AG 1024 (IGFRI inhibitor) and the tyrphostin AGl 048 (EGFR inhibitor). The cells were then exposed to BMS749 in combination or not with EGF or IGFI and submitted to apoptosis detection by Apo2;7 staining. D. FCS-mediated survival is also triggered via IGFRI and p42/44 activation. PLB 985 cells were first pre-treated for 2 hours with 10 μM U0126, lOμM LY294002, lOμM SB203580 and 2μM of the tyrphostins AG1024 and AG1478. The cells were then exposed for 48 hours to BMS749 combined or not with 10% FCS and apoptosis was detected by FACScan detection of Apo2.7 antibody. - Figure 7. PPARγ agonists induce PLB985 cells apoptosis and synergize with rexinoid. A. PLB985 cells cultured in 1 or 10% FCS were treated for 24 hours with the indicated concentration of PPAR agonists. WY- 14643 is agonist for PP ARa, BMS990 for PPARβ and troglitazone (TZD) for PPARγ. Apoptosis was detected by Apo2.7 staining. B. PLB985 were exposed to 4 different rexinoids : BMS649 (lμM), 9-cisRA (lμM), LG1069 (lμM), and BMS749 (0.5μM), in combination with the synthetic PPARγ agonist troglitazone or the natural PPARγ agonist 15d prostaglandin J2. Apoptosis assessed by FACScan detection of Apo2.7 staining demonstrates a synergistic action of PPARγ and RXR agonists. C. to E. Inhibition of the NO producing enzymes iNOS and eNOS by siRNA reduces the number of NO positive cells (E.), phosphorylation of eNOS (western blot) (C.) and apoptosis (D.) induced by BMS749. - Figure 8. RXR-PP ARγ apoptotic pathway exists in many different cell types and in vivo. A. Apoptosis assay was performed with BMS 749 in different hematopoietic cultured in 1 or 10% FCS. B. Apoptosis assay was performed in blasts cultured from an AML patient. Blasts were cultured either in 20% FCS or 1% FCS+HY. Cells were treated with BMS 749, UO 126 and 2 HDAC inhibitors SAHA and MS275. C. Iressa potentiates rexinoid apoptosis. SKBR3 breast cancer cells. Apoptosis assays were performed in SKBR3 cells grown in 10% FCS and treated with BMS749, UO 126, PGJ₂ or Iressa. D. Apoptosis assay in HT-29 colon cancer cells. E. Rexinoid apoptosis can be demonstrated in vivo. Nude mice xenografted with HCT- 116 cells were administered BMS749, UO 126 or combination for 2 weeks. The panel depicts the photograph of the mice exhibiting differeing sized tumors at the end of the treatment. F. Graphical representation of the tumor weights in xenografted mice. G. In vivo imaging of the HCT-116 xenografted mice. Mice xenografted with HCT-116 cells stably expressing luciferase were imaged at the beginning and after 7 days of treatment. Tumor size can be quantified the total amount of light emitted (photons/second). H. Schematic representation of RXR-PP ARγ-mediated apoptosis in AML cells. The activation of RXR-PP ARγ heterodimer by RXR or PPARγ agonists induces NO production which in turn affects mitochondria integrity and triggers cytochrome c release. This step can be inhibited by Bcl-2 overexpression. The activation of intrinsic pathway leads to caspase 3 cleavage and the initiation of an amplification loop through caspase 8 and bid cleavage. A late overexpression of DR5, which could be attributed to a NO/p53 mediated transactivation, increases this caspase 8 activation and can be blocked by a dominant negative form of FADD. IGF-I like growth factors, through their binding to IGFRI activate an antagonizing MEKl/2/p42/44 survival pathway which could inhibit rexinoid apoptosis at several different steps going from initiation of transcription by the heterodimer to activation of intrinsic pathway.

- Figure 9. TRAIL or etoposide-induced apoptosis is not rescued by either IGF or EGF. a. PLB985 cells were treated with 20ng/ml TRAIL or 2μM etoposide. Addition of the 3 indicated growth factors does not inhibit apoptosis as assessed by APO2.7 labelling and FACScan detection, b. Effect of different growth factors on p38 or AKT activation was assessed. PLB985 cells were treated with IGF-I, IGF-II or EGF for either Ih or 24h and the cell protein extracts were revealed by western blot for p38 and AKT phosphorylation.

- Figure 10. PPARγ agonists induce PLB985 cells apoptosis and synergize with rexinoid. a, Western blot analysis of troglitazone (5μM) and BMS649 (lμM) treated PLB985 cells extracts reveals specific cleavage of caspase 3, 8,9 and PARP, as well as decrease of Bid expression, b, Western blot analysis of cytochrome c release using fractionated protein extracts of PLB985 cells treated as in (a), c, FACS analysis of Nitric Oxide production assessed on PLB985 cells treated for 2 hours with BMS649, troglitazone or both of them, d, EGF, IGFI or FCS treatment reduces PPARγ and PPARγ/rexinoid-induced cell death. PLB985 were co-treated for 24 hours with the indicated ligands in the presence or not of IGFI, EGF or 10% FCS. Troglitazone was used at 5 μM, BMS649 at 1 μM and 15d-PGJ2 at 2μM. The cells were then assayed for apoptosis by FACScan detection of Apo2.7 staining.

- Figure 11. Administration of BMS 749 or UO 126 either singly or in combination does not cause toxicity as observed by the lack of change in the body weight of animals in the different treatment groups.

- Figure 12. a. Shows that IGF (I and II) and EGF inhibition by neutralising antibodies significantly enhanced apoptosis (Figure 12). b. RAR antagonists (BMS 614) increased rexinoid apoptosis, showing that RAR-RXR heterodimer transactivation is not involved in rexinoid apoptosis.

There will now be described by way of example a specific mode contemplated by the Inventors. In the following description numerous specific details are set forth in order to provide a thorough understanding. It will be apparent however, to one skilled in the art, that the present invention may be practiced without limitation to these specific details. In other instances, well known methods and structures have not been described so as not to unnecessarily obscure the description. EXAMPLE 1: MATERIALS AND METHODS

1.1 Ligands, chemicals and antibodies.

Retinoids, rexinoid and PPARs agonists were provided by Bristol- Myers- Squibb. BMS749 was dissolved in ethanol and used at lμM unless specified. Caspases inhibitors (Z-VAD-FMK, Z-IETD-FMK, Z-AEVD-FMK, and Z-LEHD- FMK) as well as EGF, IGFI, IGFII, insulin, G-CSF, IL3, IL4, ILlO, SCF and GM- CSF were purchased from R&D Systems. TRAIL was kindly provided by Immunex. Kinase inhibitors UO 126, LY294002 came from Cell Signalling Technology whereas SB203580 and L-NMMA were from Calbiochem, and AG1024 and AG1478 from Alexis biochemicals. Antibodies directed against caspase 3, caspase 8 caspase 9, phospho

S15 p53, phospho Sl 177 eNOS, PARP, Bid, Bcl-2, EGFR, total and phospho IGFR, Akt, p42/44 and p38, were purchased from Cell Signalling technology. Anti- cytochrome c and anti-FADD antibodies were from BD Bioscience, anti-iNOS from Santa Cruz, anti-DR5 from Sigma and anti-TRAIL from R&D Systems. The Apo2.7 antibody was purchased from Beckman Coulter and DIOC6 and DAF-FM diacetate from Molecular Probes. SiRNA against PP ARy, RXRα, eNOS and iNOS were synthesised by Proligo, Sigma; luciferin was from Promega.

1.2 Cell culture, infection and transfection.

PLB985 were adapted to minimal serum media and were cultured in RPMI, 1% fetal calf serum, complemented with 1% HY (Biosepra), gentamycin, 25mM HEPES and 2mM glutamine.

Stable PLB985 cell lines were obtained as followed: PLB985 cells were transfected with pIRES GFP, the GFP ORF is provided as SEQ ID NO: 1, and pIRES-Bcl2 GFP, the Bcl2 ORF is provided as SEQ ID NO: 2 via nucleofection using the Amaxa technology and selected for transfection by 0.8 mg/ml G418 on the entire population. The GFP-F ADDdn, the FADDdn ORF is provided as SEQ ID NO: 3, was cloned into the pLenti6/v5 lentiviral vector (Invitrogen) following the manufacturers' instructions. The Virapower lentiviral expression system (Invitrogen) was used to produce lenti virus coding for GFP-F ADDdn or β-galactosidase protein following the provided instuctions using the pLenti/GFP-FADDdn/v5 and pLenti/LacZ constructs. Once collected, the lentivirus was concentrated by ultracentrifugation (25,000 rpm, 90 min, 4°c). 3x10⁵ PLB985 cells were infected with 2x10⁶ IU/ml of virus at an MOI of 20:1 and selected for stable integration with 1 Omg/ml blasticidin.

1.3 Flow cytometry.

PLB985 cell apoptosis was quantified by detection of the 7A6 mitochondrial antigen (which is selectively exposed in cells undergoing apoptosis) using the Apo2.7 antibody. 2x10⁵ cells were permeabilized for 20 min with 100 μl of cold (4°C) solution of lOOμg/ml digitonin(Sigma) in PBS + 2.5%FCS, washed with cold PBS + 2.5%FCS, incubated for 15 min in the dark with 30 μl of a 1/5 dilution of APO2.7 PE conjugated antibody in PBS and analysed by flow cytometry. An isotypic IgGl-PE antibody was used as a non-specific control.

For the quantification of the sub-Gl fraction, the cells were resuspended in 500 μl hypotonic buffer (0.1% Triton XlOO, 0.1% sodium citrate, 50 μg/ml PI), incubated in the dark over night at 4°C and subjected to FACScan analysis. Mitochondrial membrane potential was assayed by incubating 2.5x10⁵ cells with 50 nM DiOC₆(3) for 30 min in the dark and subsequent examination by flow cytometry.

Nitric Oxide production was quantified using the selective DAF-FM diacetate probe. Briefly, 2.5x10⁵ cells were incubated for 30 min at 37 °C with 5μM DAF-FM diacetate, washed in PBS and reincubated for 15 min at 37°C prior FACScan analysis with FL2 channel. 1.4 RNAse Protection Assay.

Total RNA was extracted with Trizol (Invitrogen-Gibco BRL, Carlsbad, CA) and RNase protection assays were performed according to the supplier's instructions (Pharmingen, San Diego, CA). Briefly, 4μg total RNA and 6- 8x10⁵ cpm of α- [³³P] -uridine triphosphate-labelled template sets were used; after RNAse treatment protected probes were resolved on 5% urea-polyacrylamide-bis- acrylamide gels. 1.5 Western blot.

Total cellular protein was prepared using high salt lysis buffer (420 mM KCl, 10 mM Tris-HCl pH 7.5, 0,5 mM EDTA, 0,5 mM EGTA, 0,3% NP-40, 1 raM DTT, 20 μg/ml leupeptin, 10 μg/ml pepstatin, 10 μg/ml aproptinin, 1 mM PMSF and 0,25μM okadaic acid). 50μg of total protein was separated by SDS acrylamide gel electrophoresis, transferred to nitrocellulose membranes and probed by antibodies as indicated.

The mitochondrial/cytoplasmic fractionated protein extraction was done as described (Adrain et al., 2001). Briefly, 10⁷ cells were permeabilised in 300μl of 200μg/ml digitonin buffer (25OmM sucrose, 7OmM KCl, 137 mM NaCl, 4.3 mM Na₂HPO₄, 1.4 mM KH₂PO₄, lOOμM PMSF, 10 μg/ml leupeptin, 2 μg/ml aproptinin) for 5 min at 4°C. Cells were centrifuged at lOOOg for 5 min, the supernatant representing the cytosolic fraction. The pellet was solubilized in 300 μl mitochondrial lysis buffer (5OmM Tris pH 7.4, 15OmM NaCl, 2mM EDTA, 2mM EGTA, 0.2% triton X-100, 0.3% NP-40, lOOμM PMSF, 10 μg/ml leupeptin, 2 μg/ml aproptinin) for 30min at 4°C and centrifuge at lOOOOg for lOmin. The supernatant was the mitochondrial fraction. 25 μl of each extract was used for western blot.

1.6 p42/44 MAP Kinase assay. p42/44 activity assay was measured in vitro, using the non- radioactive p42/44 MAP Kinase Assay Kit (Cell Signalling Technology) according to the manufacturer's protocol. Briefly, 200 μg of cell extract were immunoprecipitated with immobilized Phospho-p42/44 antibodies. The resulting immunoprecipitate was incubated with an EIk-I fusion protein in the presence of ATP. Phosphorylation of EIk-I was then verified by western blot analysis. 1.7 siRNA studies.

PLB cells were transfected with 50 pmoles of siRNA using Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol. Cells were used for analysis 36h after incubation. Western blots were performed to confirm the knockdown of gene expression. 1.S In vivo experiments.

The Xenograft study using HCT-116, colon carcinoma cell lines was performed in Crl:Nu(Ico)FoxnlNu mice (Swiss) mice (Charles River Laboratories). 4x10⁶ cells were injected subcutaneously to the left flank of each animal. Treatments were started when the tumours were approximately 4 mm in size (14 animals per group). Animals were randomised in the treatment groups, so that each group contains animals of approximately same size of tumours before treatment. BMS 740 (20 mg/kg bw) and UO 126 (30 mg/kg bw) were administered intraperitoneally for 2 weeks (11 doses, alternate days) and tumour size was determined using Vernier Calliper measurement using the formula: tumour volume = (a2 x b)/2 (in which a represents width and b the length of the tumour).

For in vivo imaging experiments, HCT-116 cells stably transfected with firefly luciferase gene were used. In vivo imaging was performed after the administration of luciferin (125 mg/kg bw) using a Xenogen IVIS 100 system (Caliper Life Sciences). EXAMPLE 2: RESULTS

2.1 Rexinoid apoptosis involves activation of intrinsic death pathway.

While RXR agonists did not affect growth or differentiation of myeloid PLB985 cells in normal culture conditions, cells adapted to growth in low serum undergo rapid and massive apoptosis upon exposure to the rexinoid BMS749; under identical conditions the RARα-selective agonist BMS753 was completely inactive (Figure IA).

To assess the contributions of the extrinsic and intrinsic death pathways (Debatin, 2004) to rexinoid apoptosis, the activation of caspases and the effects of caspase inhibitors were assessed.

Under apoptogenic (low serum) condition rexinoids activated pro- caspase 9 within 2h to 4h after induction, as revealed by generation of the p37/35 fragments, which was accompanied by activation of caspase 3 and PARP cleavage

(Figure IB). Efficient pro-caspase 8 activation ("pi 8" fragment) became apparent only after a significant lag phase, suggesting a sequential activation of the intrinsic and extrinsic death pathways. However, the extrinsic pathway contributed to rexinoid apoptosis, as selective inhibitors of caspase 8 and 10 decreased cell death by about

20% each but did not block it completely (Figure 1C). Cleavage of the pro-apoptotic bcl-2 family member Bid was observed between 3 and 4 hours of rexinoid treatment, indicating that the two apoptotic pathways are connected. Together these data reveal a contribution of both the intrinsic and extrinsic pathways to rexinoid-induced cell death.

RNAse protection assays did not show any regulation of FasL, TNFα or their cognate receptors by rexinoids (Figure 2A). While TRAIL expression decreased upon rexinoid treatment, its cognate receptor DR5 was strongly overexpressed (Figure 2B). However, this up-regulation did not temporally correspond to rexinoid apoptosis. To assess whether DR5 signaling may contribute to rexinoid apoptosis, the inventors generated PLB985 cell lines stably expressing a dominant negative FADD fused to GFP (GFP-F ADDdn) or β-galactosidase (LacZ) using lentiviral transduction.

Efficient transgene expression was confirmed by immunoblotting (Figure 2C) and FACS analysis of GFP fluorescence, revealing that 98% of the cells were infected (Figure 2D). Sensitivity to apoptosis induced by exogenous TRAIL was strongly reduced and close to control levels in GFP-F ADDdn but not in wild type (WT) or LacZ cells, thus demonstrating that the death receptor pathway is effectively inhibited in these cells (Figure 2E). As expected, staurosporine and CCCP, two drugs which target the mitochondrial death pathway induced apoptosis with the same efficiency in the various cell lines (Figure 2E). That rexinoid BMS749-induced cell death was about 20% to 25% lower in GFP-F ADDdn than in WT or LacZ cells indicates a contribution, albeit not a critical one, of the death receptor signalling to rexinoid-induced apoptosis, supporting the conclusions derived from figure 1.

Activation of the intrinsic pathway is initiated by permeabilisation of the outer mitochondrial membrane resulting in a loss of transmembrane potential (ΔΨm) (Green and Kroemer, 2004).

Rexinoids induce a loss of ΔΨm as demonstrated by DioC6 staining (Figure 3A) and the release of proapoptotic mitochondrial cytochrome c and Smac/Diablo in the cytoplasm 2 to 3 hours after rexinoid exposure (Figure 3B).

In order to validate the critical involvement of the intrinsic death pathway in rexinoid-induced apoptosis, the inventors generated cells lines stably expressing Bcl-2 (SEQ ID NO: 2) and GFP (SEQ ID NO: 1) or only GFP from bicistronic vectors by nucleofection. Immunoblotting (Figure 3C) and FACS analysis of GFP fluorescence (Figure 3D) confirmed transgene expression.

Notably, the sensitivity to rexinoids ("BMS749") as well a staurosporine ("STS") is dramatically reduced in Bcl-2 overexpressing cells compared to WT or vector (pIRES) nucleofected cells, demonstrating a dominant contribution of mitochondrial-dependent apoptosis in rexinoid-induced cell death (Figure 3E).

2.2 Rexinoid-induced nitric oxide (NO) production is required for apoptosis.

Death by mitochondrial damage and concomitant loss of membrane potential can be caused by multiple intracellular events, including the generation of reactive oxygen species like (ROS) or nitric oxide (NO).

In assaying multiple pathways the inventors detected the production of NO in BMS749-treated PLB985 cells (left panel) and confirmed by DAF-FM staining (right panel) the presence of increased levels of cytoplasmic NO in the treated cells (Figure 4A). This induction was blocked by L-NMMA, a pan-NO synthase inhibitor (Figures 4B, 4C). Interestingly, the kinetics of NO production correlated with that of cytochrome c release (Figure 3B).

In keeping with the observation that NO can activate the mitochondrial apoptosis pathway (Brune, 2003), L-NMMA decreased rexinoid apoptosis by 60% (Figure 4C) and abrogated cytochrome c release (Figure 4d).

Importantly, inhibitors of ROS, L-NAC and GSH, failed to inhibit rexinoid apoptosis indicating that only NO is involved in this process (Figures 4C and D).

Two main types of nitric oxide synthases are involved in the production of NO in normal cells, inducible NOS (iNOS) and endothelial NOS (eNOS). iNOS, generally, produces a rapid and large burst of NO, while eNOS, a constitutive enzyme, is involved in the maintenance of NO levels in the cell. Western blot analysis revealed strong iNOS induction within Ih of BMS749 exposure while eNOS levels remained unchanged (Figure 4E).

However, phosphorylation of specific residues of eNOS has been shown to modulate the activity of the enzyme; in particular Serπ₇₇ phosphorylation has been demonstrated to be an important positive regulator. Indeed, BMS749 induced within 2h the levels of pSeri₁₇₇ eNOS and RNAi-mediated knockdown confirmed the critical involvement of both NOSs in rexinoid apoptosis, as depletion of either NO synthase significantly decreased the apoptosis (Figure 4F; note that a combination of both siRNAs proved to be lethal and could not be tested). Together these data suggest that a synergy between the two NO synthases results in a large and sustained increase of NO levels.

As DR5 is a target of p53, the inventors wondered if its induction by a rexinoid was mediated by NO-activated p53, thereby linking the intrinsic with the extrinsic death pathway. Upon rexinoid exposure a strong increase of p53 activating phophorylation at serine 15 was observed (Figure 4G). That this phosphorylation was inhibited by L-NMMA reveals the involvement of NO. Therefore, rexinoid apoptosis connects the intrinsic and extrinsic death pathways by two loops, the generation of tBid (Figure IB) and NO-p53-activated expression of DR5. This suggests that TRAIL may efficiently synergize with rexinoids in apoptosis induction.

2.3 IGF Receptor I activation antagonizes rexinoid-apoptosis. That PLB 985 or NB4 cells maintained under normal serum conditions do not exhibit any apoptotic response to rexinoids and that progressive increase in the serum concentration inhibited rexinoid apoptosis (Figure 5A) (Benoit et al., 2001) prompted the inventors to investigate the possible contribution of growth factor signalling in this antagonism. Out of a battery of growth factors and cytokines, only IGFI, IGFII or EGF antagonized rexinoid apoptosis under low serum conditions (Figure 5B).

Importantly, the addition of IGFs (I and II) inhibited also NO induction, while depletion of residual IGFs in the medium by neutralising antibodies significantly enhanced this process (Figure 5C) as well as subsequent apoptosis (Figure 12). Moreover, IGFs also blocked cytochrome c release (Figure 5D). Note that apoptosis induced by TRAIL (extrinsic death pathway) or etoposide (intrinsic death pathway) was not rescued by these growth factors (Figure 9). These results confirm the ability of these growth factors to selectively counteract apoptosis proceeding along the rexinoid-NO signalling axis. Analysis of the various signalling cascades that are triggered by these growth factors revealed that MAP kinase activation was the cause for the block of rexinoid apoptosis. The MEK 1/2 inhibitor UO 126 completely blocked EGF or IGF- mediated survival, whereas p38 MAP kinase (SB) or PI3 kinase (LY) inhibitors were ineffective (Figure 5E).

In keeping with these results, addition of EGF to the culture medium induced a rapid and sustained phosphorylation of the MEK 1/2 substrate p42/44 MAP kinase as shown by Western blot analysis (Figure 5F). The induction of p42/44 activity by IGFI, EGF, as well as by FCS, was confirmed by in vitro kinase assays using EIk-I as substrate (Figure 5G). No significant changes were observed in the phosphorylation status of p38 or Akt (Figure SIb). While IGF receptor 1 (IGFRl) is well expressed, in agreement with previous observations (Chen et al., 1993; Stegmaier et al., 2005), no EGF receptor could be detected in PLB985 (data not shown; note that PLB985 is an HL60 subclone). Notably, both IGFl and EGF, as well as FCS, stimulated IGFRl autophosphorylation (Figure 6A) and Elkl phosphorylation (Figure 6B). Moreover, the IGFRl tyrosine kinase inhibitor AG1024 antagonized IGFl as well as EGF- induced survival, while the EGFR inhibitor AG1478 had no effect (Figure 6C and D).

Thus, as has been observed for hepatocytes (Hallak et al., 2002), also in PLB958 cells, EGF operates through IGFRl. Altogether, these results demonstrate that the MEK/p42/44 MAPK pathway acts upstream of the NO production step to mediate the anti-apoptotic action of IGFs and EGF against rexinoid-induced cell death. It is important pointing out that while the rexinoid was ineffective in the presence of 10% FCS, the addition of the MEK 1/2 inhibitor U0126 turned it into a powerful death inducer (Figure 5H), as this permits the study of rexinoid apoptosis under culture conditions without the need to adapt cells to low serum. 2.4 Rexinoid-induced apoptosis is mediated by the RXR-PPARγ heterodimer.

The involvement of RAR-RXR heterodimer transactivation in rexinoid apoptosis can be excluded from the fact that RAR agonists (e.g., BMS753 in Figure IA) by itself did not induce, but rather inhibited death induction, while RAR antagonists increased rexinoid apoptosis (Figure 12b) (Benoit et al., 2001).

While screening a panel of ligands for multiple permissive and non- permissive RXR heterodimer partners, the inventors observed that the PPARγ agonist troglitazone (TZD) induced PLB985 cell apoptosis. Most importantly, this apoptosis occurred under the same conditions where rexinoid induced apoptosis also occured, i.e. when serum concentration is limiting (Figure 7A, compare lane 7 and 11).

The pattern of activation of various caspases, NO production, cytochrome c release were indistinguishable from that seen with rexinoids (Figure

S2a-c). PPARγ agonists troglitazone and 15-deoxy-Δ-12,14-prostaglandin J2 (15d-

PGJ2) strongly synergized with various rexinoids for apoptosis induction while no effect was seen under identical conditions with PP ARa (WY) or PPARβ (BMS990) agonists (Figure 7A and B), thus revealing the specific involvement of RXR-PPARγ heterodimers.

The addition of rexinoids enhanced TZD-induced apoptosis, which was attenuated by IGFl, EGF and FCS (Figure 10).

Finally, siRNA-mediated depletion of either PPARγ or RXR demonstrated there critical implication in both production of NO (Figure 7E), phosphorylation of eNOS and inhibition of apoptosis (Figure 7D). Taken together, these data fully support a model in which the permissive RXR-PPARγ heterodimer is the functional unit mediating the rexinoid signal which activates the intrinsic death pathway though NO production.

2.5 Growth factor-antagonized RXR apoptosis is a general phenomenon in different cell types.

The inventors believe that rexinoid apoptosis may correspond to a general phenomenon which allows the eradication of cells that are not growth factor supported. Indeed, analysis of distinct cell types supported this hypothesis.

Within the hematopoietic lineage several cell lines displayed apoptogenic sensitivity towards rexinoids exclusively in low serum (Figure 8A and data not shown; the inventors noted a weak effect, such as for K562, for some cell lines). In cases where it was possible to adapt the blasts to low serum conditions, some primary blast cultures of acute myeloid leukaemia patients went into dramatic apoptosis when rexinoid, low serum and MAPkinase inhibitor were combined (Figure 8B, lane 10). It is worth noting that the extent of death was similar to that seen with the histone deacetylase inhibitors MS275 (lane 3) and SAHA (lane 2). Also non-hematopoietic cell lines, including SKBR3, MCF7 and ZR75 breast cancer and HCTl 16 and HT29 colon cancer cell lines displayed rexinoid apoptosis (Figure 8C, D and data not shown). In the case of SKBR3, the additional blocking of the signalling emanating from HER2 by gefitinib (IRESSA) resulted in a further increase of apoptosis (Figure 8C, lane 9). Rexinoid apoptosis in SKBR3 was sensitive to the pan-NOS inhibitor L-NMMA, thus supporting the implication of NO signalling similar to PLB985 (data not shown).

2.6 Anti-cancer action of retinoid apoptosis in vivo.

To assess anti-tumor efficacy of rexinoid apoptosis in vivo the inventors treated nude mice bearing HCTl 16 tumor xenografts with BMS749 alone and in combination with UO 126 for 2 weeks. Using 20mpk BMS749 and 30mpk UO 126 the inventors did not observe any toxicity, as supported by the normalized weight profile (Fig. 11). While BMS749 and UO 126 alone had a moderate effect of tumor growth (21 +/-13%) and 44+/- 10% growth inhibition, respectively), their combination dramatically impaired tumor cell expansion in vivo (83+/-7%; see Figs. 8E, F). Using luciferase-expressing HCTl 16 the combined treatment halted tumor growth in the first week of treatment nearly quantitatively (Fig. 8G). These results indicate that the combination of rexinoids with agents that interfere with growth factor signalling correspond to an efficient novel therapeutic paradigm. 3. DISCUSSION

The inventors have discovered a new anti-cancer signalling pathway for rexinoids, which is based on the growth factor-antagonized induction of "rexinoid apoptosis" mediated by the putative tumor suppressor PPARγ-RXR through NO- dependent activation of the intrinsic death pathway (schematically summarized in Fig. 8H) and reveal its therapeutic potential in vitro, in leukemic blasts ex vivo and in xenograft assays in vivo.

Rexinoid apoptosis in the absence of growth factor support is entirely different from the rexinoid - cAMP crosstalk mentioned above (Altucci et al., 2005) because (i) it is not inhibited but rather increased by RAR antagonists, (ii) it does not involve the RXR-RAR but the RXR-PP ARγ permissive heterodimer, (iii) it does not require prior differentiation and (iv) apoptosis induction involves different apoptotic pathways. The inventors work suggests the existence of a regulatory paradigm by which a cell (compartment/tissue) that is dependent on growth factor support, particularly IGFs or EGF, will be eliminated by rexinoid apoptosis as soon as GF support ceases. The simplicity of this concept is that it is solely GF presence (in cases of constant rexinoid levels) which suffices to promote survival and its absence automatically leads to death.

Apart from GF modulation, this model implies the existence of endogenous RXR ligands which may potentially represent a second trigger, allowing non-supported cells to survive in the absence of (sufficient) rexinoid. Indeed, the existence of endogenous retinoids has been reported (de Urquiza et al., 2000; Forman et al., 1995; Solomin et al., 1998). It is worth noting that data from RXR mutant mice adds credence to such a concept. Interdigital webbing (soft tissue syndactyly) has been observed in RXR AFl domain mutants, which is an accepted marker for a defect in the apoptotic process leaving the interdigital mesenchyme intact during embryonic development (Mascrez et al., 2001). Therefore, it is possible that rexinoid apoptosis is a general mechanism actively involved in tissue development, repair and homeostasis.

The inventors results reveal that the intrinsic death pathway has a predominant role in rexinoid apoptosis, as there is an early release of cytochrome c, a subsequent depolarisation of the mitochondrial membrane potential and the inhibitory role of bcl-2 overexpression in rexinoid-induced cell death. This apoptosis signalling is further enhanced by a caspase 3/ caspase 8/ Bid amplification loop, p53 activation and a late induction of DR5 overexpression. A similar Bid-mediated amplification loop leading to, synergy between TRAIL and DNA damage -induced apoptosis has been observed previously (Broaddus et al., 2005). As the early trigger of rexinoid apoptosis, the inventors have identified NO. This is supported by the fact that selective NO synthase inhibitor L-NMMA dramatically decreases rexinoid-induced NO production and apoptosis. NO has been described to induce a MAPK-mediated phosphorylation and activation of p53 (Brune, 2003; Schneiderhan et al., 2003). NO- mediated p53 activation has been observed to be responsible for the induction of DR5, a known p53 target (Wu et al., 1997) and shown in Fig. 4G. The role of NO in the modulation of apoptosis is controversial. Low amounts of NO appear to protect the cell while significantly large amounts cause apoptosis (Li and Wogan, 2005; Ohshima et al., 2003; Patel et al., 1999). Other factors which weigh into this discussion are the intracellular distribution and the effect of other co-factors such as calcium.

The two main types of nitric oxide synthases are the inducible (iNOS) and the endothelial (eNOS) form. iNOS has been shown to rapidly and strongly induce the production of NO during inflammatory response while eNOS has been assigned the function of producing low amounts of NO for basic cell homeostasis (Li and Wogan, 2005). The role of eNOS as a protective agent has been demonstrated in endothelial cells where it is able to prevent TNFα, LPS or reactive oxygen species induced apoptosis (Dimmeler et al., 1997). Activation of eNOS or even the addition of NO donors has been observed to increase the levels of Bcl-2 and inhibit cytochrome c release (Kim et al., 1998). However, this anti-apoptotic function seems to be restricted to endothelial cells and a few other cell types. iNOS induction and resultant NO increase plays an important role in septic shock and has an apoptotic role in most cell types including macrophages and monocytes. However, in eNOS knockout mice, induction in iNOS activity due to LPS was significantly diminished (Connelly et al., 2005). In fact, NO produced by eNOS has been shown to be essential for enhancement of iNOS activity in macrophages, thereby pointing to a pro- inflammatory role for eNOS (Connelly et al., 2003). Upon RXR ligand stimulation, a strong and sustained induction of NO was observed (see Fig. 4A). The kinetics of NOS activation and siRNA studies suggest collaboration between the two types of NOS. It remains to be established how rexinoids, in the absence of growth factor, cause the transcriptional induction of iNOS and phosphorylation of eNOS. A number of mechanisms have been proposed to explain NO-mediated apoptosis through induction of the intrinsic pathway, ROS production, formation of peroxynitrite and protein nitrosylation. Upon the exclusion of ROS, it is possible that protein nitrosylation, peroxynitrite formation, which leads to the opening of the mitochondrial transition pore or inhibition of cytochrome c oxidase play a critical role in NO- mediated apoptosis (Beltran et al., 2000; Halestrap et al., 1997; Moncada and Erusalimsky, 2002) . A apoptogenic function of PPARγ is well documented (Grommes et al., 2004) and its ligands may represent a promising, novel therapeutic approach for certain malignancies (Wang et al., 2006). Indeed, loss of function mutations in PPARγ have been associated with colon cancer and could argue for a tumor suppressor function of PPARγ (Sarraf et al., 1999). The discovery that the PAX8-PPARγ fusion protein plays a critical role in thyroid follicular oncogenesis through a dominant negative function of PPARγ has raised the possibility that abrogation of normal PPARγ function may play a role in the development of cancer (Kroll et al., 2000). Notably, synergy between rexinoids and PPARγ agonist prevent the development of chemically-induced preneoplastic mammary lesions in a mouse model of organ culture (Mehta et al., 2000) and induce apoptosis in breast cancer, thyroid carcinoma, or multiple myeloma cells (Crowe and Chandraratna, 2004; Klopper et al., 2004; Ray et al., 2004). Together these reports are in keeping with the concept of rexinoid apoptosis through RXR-PP ARy and it will be interesting to assess the possible implication of GFs in these various models.

Growth factors play an essential role in cell survival both in vivo and in vitro. Serum deprivation in culture media results in apoptosis but growth factors such as IGF-I and II can support cell survival, at least temporarily (Kurmasheva and Houghton, 2006). IL-3 dependent haematopoietic cells can avoid apoptosis in the absence of IL-3 upon treatment with IGF and such an effect has been extended to diverse cells such as cortical neuronal cells, cerebral granule cells and a variety of malignant cells (Rodriguez-Tarduchy et al., 1992). IGF can activate a number of anti- apoptotic genes such as Bcl2 (Minshall et al., 1997) and induce phosphorylation of BAD, a crucial regulator of mitochondrial apoptosis (Datta et al., 1992). IGF effectively inhibits apoptosis by stress inducing agents such as ROS and NO (Matsuzaki et al., 1999). Activation of AKT by IGF-I has been proposed as a central mechanism for the inhibition of NO-induced apoptosis. Thus, it could be intuited, inhibition of growth factor signalling would leave cells particularly susceptible to apoptotic agents, their default state. Indeed, inhibition of growth factors by receptor antibodies or small molecules against downstream signalling, sensitizes tumour cells to chemotherapeutic agents and such a strategy is now widely in use. Receptor tyrosine kinase inhibitors and receptor antibodies are used to treat a variety of tumours such as breast, head and neck, lung, colon (Wanebo et al., 2006). Together these studies suggest the hypothesis that RXR, along with PPARγ, could control an endogenous default apoptotic pathway kept in silent state by the action of growth factors.

The link between growth factor support and rexinoid apoptosis prompted the inventors to investigate the possible cancer therapeutic potential of rexinoid apoptosis. It was particularly interesting that a variety of cell types, both hematopoietic and solid cancer cell lines would undergo rexinoid apoptosis either in low serum or under condition where MAPK signalling is impaired (see Fig. 8).

Interestingly, blocking GF support by combining the MAPK inhibitor and IRESSA superactivated rexinoid apoptosis. In one case the inventors could establish conditions for primary blasts of an AML patient that allowed them to demonstrate the induction of rexinoid apoptosis which was similar to that seen with two prototypic HDAC inhibitors.

A number of growth factor signalling inhibitors, like gefitinib, herceptin etc, have been developed and are in therapeutic use as adjuvants to chemotherapy for the treatment of cancer. Their combination with RXR and/or

PPARγ agonists would be an attractive therapeutic strategy. Mouse xenograft experiments confirmed the therapeutic potential of rexinoid apoptosis in vivo and revealed the absence of any major toxicity. Interestingly, the experiments demonstrated that even a general growth factor signalling inhibitor like UO 126 can be highly effective in vivo in combination with a RXR ligand, BMS749. Combination with specific inhibitors like gefitinib would further augment the specificity and thereby reduce the drug toxicity.

REFERENCES

Adrain, C, Creagh, E. M., and Martin, S. J. (2001). Apoptosis- associated release of Smac/DIABLO from mitochondria requires active caspases and is blocked by Bcl-2. Embo J 20, 6627-6636.

Altucci, L., and Gronemeyer, H. (2001). The promise of retinoids to fight against cancer. Nature Reviews Cancer 1, 181-193.

Altucci, L., Leibowitz, M. D., Ogilvie, K. M., de Lera, A. R., and Gronemeyer, H. (2007). RAR and RXR modulation in cancer and metabolic disease.

Nat Rev Drug Discov 6, 793-810. Altucci, L., Rossin, A., Hirsch, O., Nebbioso, A., Vitoux, D.,

Wilhelm, D., Guidez, F., De Simone, M., Schiavone, E. M., Grimwade, D., et al.

(2005). Rexinoid-triggered differentiation and tumor-selective apoptosis of acute myeloid leukemia by protein kinase A-mediated desubordination of retinoid X receptor. Cancer Res 65, 8754-8765.

Beltran, B., Mathur, A., Duchen, M. R., Erusalimsky, J. D., and Moncada, S. (2000). The effect of nitric oxide on cell respiration: A key to understanding its role in cell survival or death. Proc Natl Acad Sci U S A 97, 14602- 14607. Benoit, G., Altucci, L., Flexor, M., Ruchaud, S., Lillehaug, J.,

Raffelsberger, W., Gronemeyer, H., and Lanotte, M. (1999). RAR-independent RXR signaling induces t(15;17) leukemia cell maturation. Embo J 18, 7011-7018.

Benoit, G. R., Flexor, M., Besancon, F., Altucci, L., Rossin, A.,

Hillion, J., Balajthy, Z., Legres, L., Segal-Bendirdjian, E., Gronemeyer, H., and Lanotte, M. (2001). Autonomous rexinoid death signaling is suppressed by converging signaling pathways in immature leukemia cells. MoI Endocrinol 15, 1154-

1169.

Blumenschein, G. R., Jr., Khuri, F. R., von Pawel, J., Gatzemeier, U., Miller, W. H., Jr., Jotte, R. M., Le Treut, J., Sun, S. L., Zhang, J. K., Dziewanowska, Z. E., and Negro-Vilar, A. (2008). Phase HI trial comparing carboplatin, paclitaxel, and bexarotene with carboplatin and paclitaxel in chemotherapy-naive patients with advanced or metastatic non-small-cell lung cancer: SPIRIT II. J Clin Oncol 26, 1879-1885.

Broaddus, V. C, Dansen, T. B., Abayasiriwardana, K. S., Wilson, S. M., Finch, A. J., Swigart, L. B., Hunt, A. E., and Evan, G. I. (2005). Bid mediates apoptotic synergy between tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and DNA damage. J Biol Chem 280, 12486-12493.

Brune, B. (2003). Nitric oxide: NO apoptosis or turning it ON? Cell Death Differ 10, 864-869. Cesario RM, Klausing K, Razzaghi H, Crombie D, Rungta D,

Heyman RA, LaIa DS. The rexinoid LG 100754 is a novel RXR:PPAR gamma agonist and decreases glucose levels in vivo. MoI Endocrinol. 2001 Aug; 15(8): 1360-9. Chen, L. L., Gansbacher, B., Gilboa, E., Taetle, R., Oval, J., Hibbs,

M. S., Huang, C. K., Clawson, M. L., Bilgrami, S., Schlessinger, J., and et al. (1993).

Retroviral gene transfer of epidermal growth factor receptor into HL60 cells results in a partial block of retinoic acid-induced granulocytic differentiation. Cell Growth Differ 4, 769-776.

Connelly, L., Jacobs, A. T., Palacios-Callender, M., Moncada, S., and Hobbs, A. J. (2003). Macrophage endothelial nitric-oxide synthase autoregulates cellular activation and pro-inflammatory protein expression. J Biol Chem 278, 26480- 26487. Connelly, L., Madhani, M., and Hobbs, A. J. (2005). Resistance to endotoxic shock in endothelial nitric-oxide synthase (eNOS) knock-out mice: a pro- inflammatory role for eNOS-derived no in vivo. J Biol Chem 280, 10040-10046.

Crowe, D. L., and Chandraratna, R. A. (2004). A retinoid X receptor (RXR)-selective retinoid reveals that RXR-alpha is potentially a therapeutic target in breast cancer cell lines, and that it potentiates antiproliferative and apoptotic responses to peroxisome proliferator-activated receptor ligands. Breast Cancer Res 6, R546-555.

Cui, Y., Miyoshi, K., Claudio, E., Siebenlist, U. K., Gonzalez, F. J.,

Flaws, J., Wagner, K. U., and Hennighausen, L. (2002). Loss of the peroxisome proliferation-activated receptor gamma (PPARgamma ) does not affect mammary development and propensity for tumor formation but leads to reduced fertility. J Biol

Chem 277, 17830-17835.

Datta, S., Magge, S. N., Madison, L. D., and Jameson, J. L. (1992). Thyroid hormone receptor mediates transcriptional activation and repression of different promoters in vitro. MoI Endocrinol 6, 815-825. Dawson M.I., Curr. Med. Chem.-Anti-Cancer Agents, 2004, 4, 199-

230. de Lera, A. R., Bourguet, W., Altucci, L., and Gronemeyer, H. (2007). Design of selective nuclear receptor modulators: RAR and RXR as a case study. Nat Rev Drug Discov 6, 811-820. de Urquiza, A. M., Liu, S., Sjoberg, M., Zetterstrom, R. H., Griffiths,

W., Sjovall, J., and Perlmann, T. (2000). Docosahexaenoic acid, a ligand for the retinoid X receptor in mouse brain. Science 290, 2140-2144. Debatin, K. M. (2004). Apoptosis pathways in cancer and cancer therapy. Cancer Immunol Immunother53, 153-159.

Dimmeler, S., Haendeler, J., Nehls, M., and Zeiher, A. M. (1997). Suppression of apoptosis by nitric oxide via inhibition of interleukin-1 beta-converting enzyme (ICE)-like and cysteine protease protein (CPP)-32-like proteases. J Exp Med 185, 601-607.

Duncia, J.V., et al., MEK inhibitors: The chemistry and biological activity of UO 126, its analogs, and cyclization products. Bioorg. Med. Chem. Lett. 8, 2839-2844, (1998). Elrod, H. A., and Sun, S. Y. (2008). PPARgamma and Apoptosis in

Cancer. PPAR Res 2008, 704165.

Forman, B. M., Tontonoz, P., Chen, J., Brun, R. P., Spiegelman, B. M., and Evans, R. M. (1995). 15-Deoxy-delta 12, 14-prostaglandin J2 is a ligand for the adipocyte determination factor PPAR gamma. Cell 83, 803-812. Germain, P., Iyer, J., Zechel, C, and Gronemeyer, H. (2002).

Coregulator recruitment and the mechanism of retinoic acid receptor synergy. Nature 415, 187-192.

Gottardis, M. M., Bischoff, E. D., Shirley, M. A., Wagoner, M. A., Lamph, W. W., and Heyman, R. A. (1996). Chemoprevention of mammary carcinoma by LGD 1069 (Targretin): an RXR-selective ligand. Cancer Res 56, 5566-5570.

Green, D. R., and Kroemer, G. (2004). The pathophysiology of mitochondrial cell death. Science 305, 626-629.

Grommes, C, Landreth, G. E., and Heneka, M. T. (2004). Antineoplastic effects of peroxisome proliferator-activated receptor gamma agonists. Lancet Oncol 5, 419-429.

Grosios, K., Traxler, P. 2003. Tyrosine kinase targets in drug discovery. Drugs Fut, 28(7): 679.

Halestrap, A. P., Woodfield, K. Y., and Connern, C. P. (1997). Oxidative stress, thiol reagents, and membrane potential modulate the mitochondrial permeability transition by affecting nucleotide binding to the adenine nucleotide translocase. J Biol Chem 272, 3346-3354. Hallak, H., Moehren, G., Tang, J., Kaou, M., Addas, M., Hoek, J. B., and Rubin, R. (2002). Epidermal growth factor-induced activation of the insulin-like growth factor I receptor in rat hepatocytes. Hepatology 36, 1509-1518.

Hartog H, Wesseling J, Boezen HM, van der Graaf WT. 2007. The insulin-like growth factor 1 receptor in cancer: old focus, new future. Eur J Cancer. 43(13): 1895-904.

Indra, A. K., Castaneda, E., Antal, M. C, Jiang, M., Messaddeq, N.,

Meng, X., Loehr, C. V., Gariglio, P., Kato, S., Wahli, W., et al. (2007). Malignant transformation of DMBA/TPA-induced papillomas and nevi in the skin of mice selectively lacking retinoid-X-receptor alpha in epidermal keratinocytes. J Invest

Dermatol 127, 1250-1260.

Kim, Y. M., Kim, T. H., Seol, D. W., Talanian, R. V., and Billiar, T. R. (1998). Nitric oxide suppression of apoptosis occurs in association with an inhibition of Bcl-2 cleavage and cytochrome c release. J Biol Chem 273, 31437- 31441.

Klopper, J. P., Hays, W. R., Sharma, V., Baumbusch, M. A.,

Hershman, J. M., and Haugen, B. R. (2004). Retinoid X receptor-gamma and peroxisome proliferator-activated receptor-gamma expression predicts thyroid carcinoma cell response to retinoid and thiazolidinedione treatment. MoI Cancer Ther 5, 1011-1020.

Koeffler, H. P. (2003). Peroxisome proliferator-activated receptor gamma and cancers. Clin Cancer Res 9, 1-9.

Kroll, T. G., Sarraf, P., Pecciarini, L., Chen, C. J., Mueller, E., Spiegelman, B. M., and Fletcher, J. A. (2000). PAX8-PPARgammal fusion oncogene in human thyroid carcinoma [corrected]. Science 289, 1357-1360.

Kurmasheva, R. T., and Houghton, P. J. (2006). IGF-I mediated survival pathways in normal and malignant cells. Biochim Biophys Acta 1766, 1-22.

Li, C. Q., and Wogan, G. N. (2005). Nitric oxide as a modulator of apoptosis. Cancer Lett 226, 1-15. Li, Y., Zhang, Y., Hill, J., Shen, Q., Kim, H. T., Xu, X., Hilsenbeck,

S. G., Bissonnette, R. P., Lamph, W. W., and Brown, P. H. (2007). The Rexinoid LGl 00268 prevents the development of preinvasive and invasive estrogen receptor negative tumors in MMTV-erbB2 mice. Clin Cancer Res 13, 6224-6231.

Liby, K., Royce, D. B., Risingsong, R., Williams, C. R., Wood, M. D., Chandraratna, R. A., and Sporn, M. B. (2007a). A new rexinoid, NRXl 94204, prevents carcinogenesis in both the lung and mammary gland. Clin Cancer Res 13, 6237-6243.

Liby, K. T., Yore, M. M., and Sporn, M. B. (2007b). Triterpenoids and rexinoids as multifunctional agents for the prevention and treatment of cancer. Nat Rev Cancer 7, 357-369. Lippman, S. M., Sudbo, J., and Hong, W. K. (2005). Oral cancer prevention and the evolution of molecular-targeted drug development. J Clin Oncol 23, 346-356.

Mascrez, B., Mark, M., Krezel, W., Dupe, V., LeMeur, M., Ghyselinck, N. B., and Chambon, P. (2001). Differential contributions of AF-I and AF-2 activities to the developmental functions of RXR alpha. Development 725, 2049-2062.

Matsuzaki, H., Tamatani, M., Mitsuda, N., Namikawa, K., Kiyama, H., Miyake, S., and Tohyama, M. (1999). Activation of Akt kinase inhibits apoptosis and changes in Bcl-2 and Bax expression induced by nitric oxide in primary hippocampal neurons. J Neurochem 73, 2037-2046.

McAlpine, C. A., Barak, Y., Matise, L, and Cormier, R. T. (2006). Intestinal-specific PPARgamma deficiency enhances tumorigenesis in ApcMin/+ mice. Int J Cancer 119, 2339-2346.

Mehta, R. G., Williamson, E., Patel, M. K., and Koeffler, H. P. (2000). A ligand of peroxisome proliferator-activated receptor gamma, retinoids, and prevention of preneoplastic mammary lesions. J Natl Cancer Inst 92, 418-423.

Minshall, C, Arkins, S., Straza, J., Conners, J., Dantzer, R., Freund, G. G., and Kelley, K. W. (1997). IL-4 and insulin-like growth factor-I inhibit the decline in Bcl-2 and promote the survival of IL-3-deprived myeloid progenitors. J Immunol 159, 1225-1232. Minucci, S., and Pelicci, P. G. (2007). Determinants of oncogenic transformation in acute promyelocytic leukemia: the hetero-union makes the force. Cancer Cell 12, 1-3.

Moncada, S., and Erusalimsky, J. D. (2002). Does nitric oxide modulate mitochondrial energy generation and apoptosis? Nat Rev MoI Cell Biol 3, 214-220.

Ocadiz-Delgado, R., Castaneda-Saucedo, E., Indra, A. K., Hernandez-Pando, R., and Gariglio, P. (2008). Impaired cervical homeostasis upon selective ablation of RXRalpha in epithelial cells. Genesis 46, 19-28. Ohshima, H., Tatemichi, M., and Sawa, T. (2003). Chemical basis of inflammation-induced carcinogenesis. Arch Biochem Biophys 417, 3-11.

Orlandi, A., Bianchi, L., Costanzo, A., Campione, E., Giusto Spagnoli, L., and Chimenti, S. (2004). Evidence of increased apoptosis and reduced proliferation in basal cell carcinomas treated with tazarotene. J Invest Dermatol 122, 1037-1041.

Patel, R. P., McAndrew, J., Sellak, H., White, C. R., Jo, H., Freeman, B. A., and Darley-Usmar, V. M. (1999). Biological aspects of reactive nitrogen species. Biochim Biophys Acta 1411, 385-400.

Ramlau, R., Zatloukal, P., Jassem, J., Schwarzenberger, P., Orlov, S. V., Gottfried, M., Pereira, J. R., Temperley, G., Negro- Vilar, R., Rahal, S., et al. (2008). Randomized phase III trial comparing bexarotene (L 1069- 49)/cisplatin/vinorelbine with cisplatin/vinorelbine in chemotherapy-naive patients with advanced or metastatic non-small-cell lung cancer: SPIRIT I. J CHn Oncol 26, 1886-1892. Ray, D. M., Bernstein, S. H., and Phipps, R. P. (2004). Human multiple myeloma cells express peroxisome proliferator-activated receptor gamma and undergo apoptosis upon exposure to PPARgamma ligands. Clin Immunol 113, 203- 213.

Rodriguez-Tarduchy, G., Collins, M. K., Garcia, L, and Lopez- Rivas, A. (1992). Insulin-like growth factor-I inhibits apoptosis in IL-3 -dependent hemopoietic cells. J Immunol 149, 535-540. Saez, E., Olson, P., and Evans, R. M. (2003). Genetic deficiency in Pparg does not alter development of experimental prostate cancer. Nat Med 9, 1265- 1266.

Sarraf, P., Mueller, E., Smith, W. M., Wright, H. M., Kum, J. B., Aaltonen, L. A., de Ia Chapelle, A., Spiegelman, B. M., and Eng, C. (1999). Loss-of- function mutations in PPAR gamma associated with human colon cancer. MoI Cell 3, 799-804.

Schneiderhan, N., Budde, A., Zhang, Y., and Brune, B. (2003). Nitric oxide induces phosphorylation of p53 and impairs nuclear export. Oncogene 22, 2857-2868.

Shulman, A. L, Larson, C, Mangelsdorf, D. J., and Ranganathan, R. (2004). Structural determinants of allosteric ligand activation in RXR heterodimers. Cell 11 «5, 417-429.

So, P. L., Fujimoto, M. A., and Epstein, E. H., Jr. (2008). Pharmacologic retinoid signaling and physiologic retinoic acid receptor signaling inhibit basal cell carcinoma tumorigenesis. MoI Cancer Ther 7, 1275-1284.

So, P. L., Lee, K., Hebert, J., Walker, P., Lu, Y., Hwang, J.,

Kopelovich, L., Athar, M., Bickers, D., Aszterbaum, M., and Epstein, E. H., Jr.

(2004). Topical tazarotene chemoprevention reduces Basal cell carcinoma number and size in Ptchl+/- mice exposed to ultraviolet or ionizing radiation. Cancer Res 64,

4385-4389.

Solomin, L., Johansson, C. B., Zetterstrom, R. H., Bissonnette, R. P., Heyman, R. A., Olson, L., Lendahl, U., Frisen, J., and Perlmann, T. (1998). Retinoid- X receptor signalling in the developing spinal cord. Nature 395, 398-402. Soprano, D. R., Qin, P., and Soprano, K. J. (2004). Retinoic acid receptors and cancers. Annu Rev Nutr 24, 201-221.

Stegmaier, K., Corsello, S. M., Ross, K. N., Wong, J. S., Deangelo, D. J., and Golub, T. R. (2005). Gefitinib induces myeloid differentiation of acute myeloid leukemia. Blood 106, 2841-2848. Sun, S. Y., and Lotan, R. (2002). Retinoids and their receptors in cancer development and chemoprevention. Crit Rev Oncol Hematol 41, 41-55. Wanebo, H. J., Argiris, A., Bergsland, E., Agarwala, S., and Rugo, H. (2006). Targeting growth factors and angiogenesis; using small molecules in malignancy. Cancer Metastasis Rev 25, 279-292.

Wang, T., Xu, J., Yu, X., Yang, R., and Han, Z. C. (2006). Peroxisome proliferator-activated receptor gamma in malignant diseases. Crit Rev Oncol Hematol 58, 1-14.

Wu, G. S., Burns, T. F., McDonald, E. R., 3rd, Jiang, W., Meng, R., Krantz, I. D., Kao, G., Gan, D. D., Zhou, J. Y., Muschel, R., et al (1997). KILLER/DR5 is a DNA damage-inducible p53-regulated death receptor gene. Nat Genet 77, 141-143.

Wu, K., Zhang, Y., Xu, X. C, Hill, J., Celestino, J., Kim, H. T.,

Mohsin, S. K., Hilsenbeck, S. G., Lamph, W. W., Bissonette, R., and Brown, P. H.

(2002). The retinoid X receptor-selective retinoid, LGD 1069, prevents the development of estrogen receptor-negative mammary tumors in transgenic mice. Cancer Res 62, 6376-6380.

Zeisig, B. B., Kwok, C, Zelent, A., Shankaranarayanan, P., Gronemeyer, H., Dong, S., and So, C. W. (2007). Recruitment of RXR by homotetrameric RARalpha fusion proteins is essential for transformation. Cancer Cell 12, 36-51.

Claims

1. A composition comprising: at least one agonist of the RXR-PP ARγ heterodimer complex selected in the group consisting of a RXRα agonist, a RXRβ agonist, a RXRγ agonist, a PP ARγ agonist or RXR-PP ARγ heterodimer agonist; and a growth factor inhibitor; for the treatment of cancer or a tumour.

2. The composition of claim 1, wherein said at least one agonist is a pan-agonist of RXRα, RXRβ and RXRγ.

3. The composition of claim 1, wherein said at least one agonist is an exclusive agonist of RXRα, RXRβ or RXRγ.

4. The composition according to claims 2 or 3, wherein said at least one agonist is a rexinoid.

5. The composition according to claim 1, wherein said at least one agonist is an agonist of PP ARγ.

6. The composition according to claim 5, wherein said PP ARγ agonist is selected from the group comprising: thiazolidinediones, prostanoids and long chain fatty acids.

7. The composition according to claim 6, wherein said thiazolidinedione is selected from the group comprising: rosiglitazone; pioglitazone; troglitazone, ciglitazone.

8. The composition according to claim 6, wherein said prostanoid is 15-deoxy-Δ-12,14-prostaglandin J2.

9. The composition according to claim 1 , wherein said at least one agonist is a RXR-PP ARγ agonist.

10. The composition according to any one of claims 1 to 9, wherein said growth factor inhibitor is a general inhibitor.

11. The composition according to any one of claims 1 to 9, wherein said growth factor inhibitor is a specific inhibitor.

12. The composition according to any one of claims 1 to 11, wherein said growth factor inhibitor is an antagonist to a receptor for said growth factor.

13. The composition according to any one of claims 1 to 11, wherein said growth factor inhibitor is a inhibitor which blocks at least one enzymatic activity of said growth factor.

14. The composition according to any one of claims 1 to 11, wherein said growth factor inhibitor inhibits at least one target of said growth factor.

15. The composition according to any one of claims 1 to 14, wherein said growth factor inhibitor affects at least one of the growth factors selected from the group comprising: IGFI; IGFII; EGF.

16. The composition according to any one of claims 1 to 15, wherein said growth factor inhibitor inhibits the MAP kinase pathway.

17. The composition according to claim 16, wherein said growth factor inhibitor inhibits a component of the MAP kinase pathway selected from the group comprising: MEK 1; MEK 2; p42; p44.

18. The composition according to any one of claims 1 to 17, further comprising at least one component selected from: a) a RAR antagonist; and b) TRAIL or a functional mimic which activates the DR5 receptor and activates the formation of the death inducing signalling complex.

19. The composition according to any one of claims 1 to 18, wherein said composition comprises both at least one RXR agonist and at least one PPARγ agonist.

20. The composition according to any one of claims 1 to 19, wherein said cancer is selected from the group: lung, ovarian, central nervous system, skin, colon or leukaemia.

21. Products containing a RXR-PP ARγ heterodimer complex agonist as defined in claim 1 and a growth factor inhibitor as a combined preparation for simultaneous, separate or sequential use in cancer therapy.

22. Products containing a RXR-PP ARγ heterodimer complex agonist as defined in claim 1 , a growth factor inhibitor, and a RAR antagonist and/or TRAIL or a functional mimetic thereof as a combined preparation for simultaneous, separate or sequential use in cancer therapy.

23. An in vitro method to determine the effects of administration of at least one agonist of RXRα, RXRβ, RXRγ or PPARγ, upon apoptosis in at least one cell, comprising the steps: a) culturing said at least cell under conditions sufficient for said at least one cell to survive; b) supplementation of the cell culture of step a) with a growth factor inhibitor and the measurement of the cellular apoptosis level; c) supplementation of the cell culture of step b) with at least one agonist of RXRα, RXRβ, RXRγ or PPARγ,and the measurement of the cellular apoptosis level; d) comparing the measured apoptosis levels of step b) and step c) and determining the effects of said at least one agonist of RXRα, RXRβ, RXRγ or PPARγ,upon said at least one cell.

24. A kit to determine the effects of administration of at least one agonist of RXRα, RXRβ, RXRγ or PPARγ, upon apoptosis in at least one cell, comprising: a) a growth factor inhibitor; b) at least one agonist of RXRα, RXRβ, RXRγ or PPARγ,; c) instructions.