Potentiating a cell death-inducing compound via disruption of CXCR4 clusters on tumor cells

Abstract

The invention relates to methods of treating a tumor comprising CXCR4 clusters in a subject, comprising simultaneous or sequential administering a CXCR4 cluster inhibitor and a cell death-inducing compound to the subject. The invention further relates to a composition and to a kit of parts, comprising a CXCR4 cluster inhibitor and a cell death-inducing compound.

Description

[0001] P135924PC00 Title: Potentiating a cell death-inducing compound via disruption of CXCR4 clusters on tumor cells FIELD The invention relates to methods for the treatment of a tumor, especially tumors expressing CXCR4 clusters. INTRODUCTION G protein-coupled receptor (GPCR)-mediated signal transduction is involved in a plethora of cellular processes. Because of their relevance in the control of (patho- )physiology, the superfamily of GPCRs constitutes the largest class of targets for FDA- approved drugs (Hauser et al., 2017. Nat Rev Drug Discov 16: 829-842). Although classically being thought of as monomeric units, the concept of GPCRs existing and signaling as dimers or as higher-order oligomers (referred to as clusters) has been increasingly recognized (Milligan et al., 2019. Curr Opin Cell Biol 57: 40-47). GPCRs can be grouped into 6 classes based on sequence homology and functional similarities. For the functionality of class C family of GPCRs (metabotropic glutamate / pheromone), formation of homo- and heterodimers has been shown to be essential (Moustaine et al., 2012. Proc Natl Acad Sci USA 109:16342-16347; Kniazeff et al., 2011. Pharmacology & therapeutics 130: 9-25). In contrast, the existence of such receptor oligomers and the potential functional relevance thereof for the larger class A family of GPCRs (rhodopsin-like) is still a topic of debate (Bouvier et al., 2014. J Physiol 592: 2439-2441; Lambert et al., 2014. J Physiol 592: 2443-2445). The chemokine receptor CXCR4 is a class A GPCR that is overexpressed in many human cancers, including hematological malignancies. In these cancers, CXCR4 activates several signaling pathways that promote pro-tumorigenic phenotypes such as survival and metastasis (Chatterjee et al., 2014. Adv Cancer Res 124: 31-82). In chronic lymphocytic leukemia (CLL), CXCR4 creates a protective tumor microenvironment (TME) by allowing migration into the lymph node and subverting adjacent cells (Smit et al., 2021. Annu Rev Pharmacol Toxicol 61: 541-563). Moreover, CXCR4 contributes to retention of malignant cells in the bone marrow, which protects these cells from chemotherapy (Peled et al., 2018. Cytokine 109: 11-16). Given the central role of CXCR4 in the pathophysiology of hematological malignancies, CXCR4 is considered an important drug target (Cancilla et al., 2020. Front Oncol 10:1672). Upon recombinant overexpression to levels mimicking an oncogenic setting, CXCR4 was reported to exist almost exclusively as dimer or higher-order oligomer (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154; Ward et al., 2021. J Biol Chem 296: 100139). In primary T-cells, CXCR4 exists predominantly as monomer whereas in the Jurkat T-cell line, CXCR4 resides mainly in nanoclusters (i.e. clusters of three or more receptors) (Martinez-Munoz et al., 2018. Mol Cell 70: 106-119). Oligomerization of CXCR4 has also been differentially implicated in receptor function. The capacity of CXCL12, a ligand of CXCR4, to induce formation of CXCR4 nanoclusters was reported as essential for migration of CXCR4-expressing T-cells (Martinez-Munoz et al., 2018. Mol Cell 70: 106-119). In heterologous expression systems, CXCR4 dimerization has been proposed to be a consequence of basal receptor activity (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154). Whether CXCR4 receptors extensively oligomerize in a cancer setting, such as a hematological malignancy, and whether pro-tumorigenic effects of CXCR4 can potentially be attributed to receptor cluster-specific signaling, is still elusive. It has now been found that clustering of native CXCR4 instigates basal signaling activity towards cell migration and cell survival. Said oligomerization-driven basal signaling was found to overcome the death of tumor cells by a cell death-inducing compound. As most cell-death-inducing compounds produce side effects, reducing CXCR4 basal signaling may provide a more effective anti-tumor therapy and, in particular, an effective anti-tumor therapy with fewer side effects. BRIEF DESCRIPTION OF THE INVENTION It was found that CXCR4 receptors oligomerize constitutively in cells and primary cultures of hematological malignancies. Said CXCR4 clusters instigate basal signaling activity toward cell migration and survival. Furthermore, it was found that different types of molecules, including antibodies and small molecules are able to disrupt said CXCR4 oligomers and reduce said basal signaling. Reduction of the size of CXCR4 clusters induced cell death and increased the sensitivity to a cell death-inducing agent such as venetoclax in high CXCR4-expressing cells. Such effect was not observed for non-cluster-reducing small CXCR4 antagonists such as AMD3100 (plerixafor) and a non-cluster disrupting CXCR4 inverse agonist such as TG-0054 (burixafor). Accordingly, this invention relates to methods of potentiating a cell death-inducing compound, comprising providing a cell with a CXCR4 cluster inhibitor and the cell death-inducing compound, wherein the cell is a tumor cell comprising CXCR4 clusters. An advantage of potentiating a cell death-inducing compound is that a smaller dose of the compound is needed for the same therapeutic effect, resulting in lower treatment costs, fewer side effects caused by the cell death-inducing compound and reducing the chance of developing resistance against the cell death-inducing compound. The invention therefore provides a CXCR4 cluster inhibitor, for use in a method of treating a tumor comprising CXCR4 clusters in a subject, the method further comprising simultaneous or sequential administering a cell death-inducing compound to the subject. Said CXCR4 cluster inhibitor may be a small molecule, an antibody, an antibody fragment, or a combination thereof. Said CXCR4 cluster inhibitor preferably is selected from IT1t, 2D01, VUN401, VUN411, VUN418, VUN419 or VUN421, or any combination thereof. In embodiments, said CXCR4 cluster inhibitor for use according to the invention is an antibody conjugated to IT1t, or to AMD070. Said cell death-inducing compound preferably is a TNF-related ligand and / or a BCL-2 inhibitor. In embodiments, said cell death-inducing compound is venetoclax. In embodiments, a CXCR4 cluster inhibitor is administered before the cell death- inducing compound. The invention further provides a composition comprising a CXCR4 cluster inhibitor and a cell death-inducing compound, preferably a pharmaceutical composition. The CXCR4 cluster inhibitor in a composition according to the invention may be a small molecule, an antibody, antibody fragment, or a combination thereof, preferably selected from IT1t, 2D01, VUN401, VUN411, VUN418, VUN419, VUN421, or a combination thereof. In embodiments, the CXCR4 cluster inhibitor is an antibody fragment conjugated to a small molecule. A cell death-inducing compound in a composition according to the invention may be a TNF-related ligand and / or a BCL-2 inhibitor, preferably venetoclax. The invention further provides a kit of parts, comprising a CXCR4 cluster inhibitor, a cell death-inducing compound, and instructions to provide an individual having a tumor comprising CXCR4 clustering with the CXCR4 cluster inhibitor and the cell death-inducing compound. The invention further provides a method of treating a tumor comprising CXCR4 clusters in a subject, comprising simultaneous or sequential administering a CXCR4 cluster inhibitor and a cell death-inducing compound to the subject. Said tumor may be a lymphoma, myeloma, leukemia, glioma, rhabdomyosarcoma or prostate cancer. In embodiments, said tumor is a non-Hodgkin lymphoma. The invention further provides a method of reducing basal motility of a cell by administering a CXCR4 cluster inhibitor, whereby the CXCR4 cluster inhibitor is IT1t, VUN401, 2D01, VUN411, VUN418, VUN419, VUN421, or a combination thereof. Said CXCR4 cluster inhibitor preferably is IT1t and / or VUN401. BRIEF DESCRIPTION OF THE FIGURES Figure 1. CXCR4-targeting ligands differentially modulate receptoroligomerization. (A) Schematics and levels of BRET-based measurement of CXCR4oligomerization in HEK293T cells expressing CXCR4-Rluc and increasing amounts of CXCR4-YFP, unstimulated or stimulated with CXCL12 (100 nM). Data are the mean ±SD and represent three independent experiments performed in triplicate. (B) BRET-based measurement of CXCR4 oligomerization using tagged receptors in HEK293T cells after stimulation with increasing concentrations of indicated ligands. Data are normalized to the vehicle and are the pooled mean ± SEM of three independentexperiments performed in triplicate. (C-E) BRET-based measurement of CXCR4oligomerization using tagged receptors in HEK293T cells after stimulation with increasing concentrations of indicated single-domain antibodies (C), small molecules (D) and IgG Fab antibody fragment or IgG antibody (E). Data are normalized to vehicle (100%) and 1 µM VUN401 (0%) conditions and are the pooled mean ± SEM of three independent experiments performed in triplicate. (F). SpIDA analysis of HEK293AD cells expressing monomeric control ^1AR-EYFPica and unstimulated CXCR4-EYFP orCXCR4-EYFP after stimulation with ligands as indicated. Each data point representsthe average brightness value from three regions in one cell, normalized to the monomer control, given with mean ± SD as error bars. Data were obtained from three independent experiments per condition. N.S. non-significant, ** P <0.01, *** P < 0.001, **** P < 0.0001 compared to unstimulated CXCR4-EYFP, according to a one-way ANOVAajggjr`_ ]t ^d_zf^n kjno cj^ o`no (^d_zf, 1967. J Am Statist Assoc 62: 626^633).Figure 2. Modulation of CXCR4 oligomerization inversely correlates with CXCR4-JAK2 association. (A) Schematics and results of BRET-based measurement of basal orCXCL12-stimulated association of CXCR4-NanoLuc (CXCR4-Nluc) with increasing amounts of mVenus-JAK2 in HEK293T cells. (B, C) BRET-based measurement of CXCR4-JAK2 association in HEK293T cells after stimulation with increasing concentrations of CXCL12, in the absence or presence of PTx (100 ng / mL) (B) or increasing concentrations of indicated ligands which are either single-domain antibodies (VUN codes), small molecules (IT1t or AMD3100) or single-domain-antibody-Fc fusionmolecules (C). (D) Linear correlation plot of BRETmax values in CXCR4 oligomerization(x-axis) vs. CXCR4-JAK2 association (y-axis) obtained for the panel of ligands, as indicated. Data are pooled mean ± SEM of at least two independent experiments performed in duplicate. Figure 3. Cluster sizes of endogenous CXCR4 receptors in hematological cancer cell lines. (A) Schematics explaining assay format and single-domain antibody-based bioluminescence resonance energy transfer (BRET) measurement of CXCR4 clusters in hematological cancer cell lines MEC-1, RPCI-WM1 and Z-138. Data are representative of at least three independent experiments and depicted as mean ± SD. (B) Normalized CXCR4 cluster BRETmax vs flow cytometry surface receptor expression levels for full hematological cancer cell line panel. Data are pooled mean ± SEM of at least three independent experiments, each performed in duplicate. (C) Effects of silencing of CXCR4 using scramble shRNA or CXCR4-targeting siRNA in Z-138 cells on oligomerization BRET measured by single-domain antibody-based sensor. Data normalized to the BRETmin value of each individual cell line are representative of at least three independent experiments and depicted as mean ± SD. (D) Data of dSTORM (direct stochastic optical reconstruction microscopy) imaging and spatial point distribution\i\gtndn pndib Pdkg`t^n I api^odji ji PNAG-WM1 (CXCR4low) and Z-138 (CXCR4high)cells. Analyzed region of reconstructed dSTORM image is visualized. Correspondingpseudo-colored cluster heat map, thresholded ]di\mt h\k \i_ Pdkg`t^n F api^odji kgjo(gdi`\m om\inajmh`_ Pdkg`t^n I) \m` ncjri. Gi Pdkg`t^n F api^odji kgjo, ]gp` gdi` (])di_d^\o`n oc` Pdkg`t^n F api^odji ajm oc` ocm`ncjg_`_ ]di\mt h\k, rc`m`\n 99% AG jaMonte Carlo simulation is indicated by red (r) and green (g) lines. Representative analysis of two independent experiments is shown for both cell lines. (E) Cluster diameter comparison between RPCI-WM1 and Z-138 cell lines based on the spatial point distribution analysis. Data (violin plot) are pooled from eight analyzed areas, obtained from two independent experiments per cell line. (F) Cluster stoichiometry comparison between the RPCI-WM1 and Z-138 cell lines. Data are pooled mean ± SEM of eight analyzed areas, obtained from two independent experiments per cell line. Figure 4. Pharmacological disruption of endogenous CXCR4 clusters by small molecules. (A) Schematics and levels of bioluminescence resonance energy transfer (BRET)-based measurement of VUN415-ATTO565 (1 nM) displacement by increasing concentrations of indicated CXCR4 small molecule antagonists using membrane extracts from NanoLuc-CXCR4-expressing HEK293T cells. Data are pooled mean ± SEM of three independent experiments. (B) Schematics and displacement of indicated antibody- ATTO565 (1 nM) from NanoLuc-CXCR4 in HEK293T cell membrane extracts by increasing concentrations of IT1t, measured by BRET. Data are pooled mean ± SEM of three independent experiments. (C) CXCR4 monomerization by 10 ^M AMD070 and IT1t in Z-138 cells, measured by single-domain antibody-based BRET. Increasing equimolar concentrations of detection single-domain antibodies (VUN415-NanoLuc / -ATTO565 for AMD070, VUN416-NanoLuc / -ATTO565 for IT1t) were used. Data are mean ± SD and are representative of three independent experiments, each performed in triplicate. Figure 5. Inhibition of basal migration of Z-138 MCL cells by cluster disrupting small molecule IT1t, single-domain antibody VUN401 and Fab 2D01. Corresponding non-monomerizing molecules bivalent 2D01-Fc, single-domain antibody VUN415 and small molecule AMD3100 are also included. Basal migration of Z-138 MCL cells was assessed in a Transwell® migration assay. Data are pooled mean ± SEM of at least three independent experiments. Figure 6. Disruption of CXCR4 clusters inhibits spheroid growth in primary CLL and MCL cultures. (A) Effects of 1 and 10 ^M of indicated CXCR4 antagonists on IL- 2 / IL-15 / IL-21 / CpG cocktail-induced growth curve in CLL patient-derived spheroid model (Haselager et al., 2023. Hemasphere 7: e938). Data are mean ± SEM of cultures from four (TG-0054) or five individual patients. (B) Effects of indicated CXCR4 antagonists (10 ^M) on cell viability in primary CLL spheroid model. Data are mean ± SEM of cultures from four (TG-0054) or five individual patients. (C,D) Effect of 10 ^M of indicated CXCR4 antagonists on CLL activation marker CD95 (C) and CD4 / CD8 activation marker CD25 (D) in primary CLL spheroid model. Data are mean ± SEM of cultures from four (TG-0054) or five individual patients. (D,C) Effects of 10 ^M of indicated CXCR4 antagonists on CLL activation marker CD95 (C) and CD4 / CD8 activation marker CD25 (D) in primary CLL spheroid model. Data are mean ± SEM of cultures from four (TG-0054) or five individual patients. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001 compared to vehicle according to one-way ANOVAs followed byDpii`o^n kjno cj^ o`no (A,B). No significant difference compared to vehicle, according toone-r\t ?LMT?n ajggjr`_ ]t Bpii`o^n kjno cj^ o`no (B). Figure 7. CXCR4 oligomers drive anti-apoptotic signaling in MCL cells. (A) Phosphoproteomics study setup and workflow. (B) Venn diagram depicting overlapping and non-overlapping GO biological processes of IT1t and VUN40160 min treatment samples. (C) Protein network of downregulated phosphosites in apoptosis sensitization triggered by IT1t and VUN401. The phosphoproteins depicted with bold text are known functional phosphosites. Figure 8. CXCR4 oligomers drive anti-apoptotic signaling and phenotype in MCL cells. (A) Resazurin-based measurement of metabolic activity in Z-138, JEKO-1 and MAVER-1 MCL cells after 48 h treatment with increasing concentrations venetoclax in absence (vehicle) or presence of 10 ^M AMD3100, IT1t or VUN401 (Z-138 only). Data,ijmh\gdu`_ oj oc` ^ij q`i`oj^g\s^ ^ji_dodji, r`m` kjjg`_ h`\i ^ QCK ja \o g`\no ocm``independent experiments, each performed in triplicate. (B) Assessment of dead cell population by staining Z-138, JEKO-1, and MAVER-1 MCL cells with the viability dyes TOPRO-3 and MitoTracker Orange following 48 h treatment with increasing concentrations of venetoclax in the absence (vehicle) or presence of 10 ^M AMD3100, IT1t, AMD070, TG-0054 or VUN401 (Z-138 only). (C) Resazurin-based assessment of dose-dependent potentiation of venetoclax by IT1t in Z-138 MCL cells. Data normalizedoj oc` ^ij q`i`oj^g\s^ ^ji_dodji were pooled mean ± SEM of at least three independentexperiments, each performed in duplicate. (D) Synergy assessment between 10 ^M AMD3100, IT1t, AMD070, TG-0054 or VUN401 and venetoclax to induce cell death based on the bliss independence model with data obtained in (B). (E) Cytotoxicity assessment of IT1t only (10 ^M), venetoclax only (3.16 nM), and co-treatment of IT1t and venetoclax in the absence and presence of the pan-caspase inhibitor qVD (20 µM). (F, G) Effects of AMD3100 and IT1t on venetoclax-induced cell death in primary cultures of CLL patients. Full concentration-response curves for one patient (F) or pIC50 for four patients (G) are shown. * P < 0.05 according to unpaired t-tests (G). Figure 9. Dose-dependent potentiation of venetoclax by 10 µM of IT1t in DMS 79lpib nh\gg ^`gg ^\i^`m KAJ ^`ggn. B\o\ ijmh\gdu`_ oj oc` ^ij q`i`oj^g\s^ ^ji_dodji (h`\i± SD, N=2 technical replicates). Figure 10. Migration trajectory plots of MCL Z-138 cells. Trajectories are representative of at least three independent experiments. Shown are (A): the average constitutive cell migration; (B): the average migration speed; and (C) the averagetraveled distance, ajggjrdib om`\oh`io rdoc 1 ^K ja ?KB3100, GR1o, TSL415 jmVUN401 derived from average trajectory information. Data are pooled mean ± SEM of at least three independent experiments. * P < 0.05, ** P < 0.01 compared to vehicle, according to unpaired t-tests. DETAILED DESCRIPTION OF THE INVENTION Definitions As are used herein, the singular forms "a", "an" and "the", are intended to include the plural forms as well. As is used herein, the term "or" includes any and all combinations of one or more of the associated listed items, unless the context clearly indicates otherwise (e.g. if an^`doc`m ^.jm^ ^jinomp^odji dn pn`_).As are used herein, the terms "comprise" and "comprising", and conjugations thereof, are open language and specify the presence of stated features but do not preclude the presence or addition of one or more other features. It will be understood that when a particular step of a method is referred to as subsequent to another step, it can directly follow said other step or one or more intermediate steps may be carried out before carrying out the particular step, unless specified otherwise. ?n dn pn`_ c`m`di, oc` o`mh ^h`ocj_ ja kjo`iod\odib \ ^jhkjpi_^, m`a`mn oj \method wherein the potency of a compound is increased. An increased potency of a compound may mean that the compound has an increased activity, e.g. an increased cell death-inducing activity. For example, a compound may have a certain activity when administered at a certain dose; if this compound is potentiated, e.g. by a method of the invention, it means that the compound will have a higher activity at that certain dose, or has that certain activity at a lower dose. An increased potency of a compound may, for example, be due to an increased bioavailability, and / or a reduced degradation. ?n dn pn`_ c`m`di, oc` o`mh ^^`gg _`\oc-di_p^dib ^jhkjpi_^, m`a`mn oj \ ^jhkjpi_that induces or causes cellular death in a cell such as, e.g., a tumor cell. Cell death may occur via necrosis or apoptosis. Examples of cell death-inducing compounds are classical anticancer agents such as anthracyclines, antimetabolites and platinum drugs; topoisomerase inhibitors, e.g. camptothecin; cell death-inducing cytokines, such as IFN- ^, IFN-^, TNF-^, or TRAIL; taxanes, such as paclitaxel and fluorinated taxanes. A preferred cell death-inducing compound is provided by a cytokine, especially a TNF- related ligand such as TNF-e and TRAIL, and / or a BCL-2 inhibitor such as venetoclax. ?n dn pn`_ c`m`di, oc` o`mh ^CXCR4^, m`a`mn oj a G protein-coupled, seven-transmembrane chemokine receptor. CXCR4 is an important regulator of the tumor- microenvironment of various malignancies such as e.g. hematological malignancies, inducing pro-survival signaling and tumor homing to lymphoid organs. CXCR4 is known to organize into multi-receptor clusters at the plasma membrane, a process referred to as oligomerization. CXCR4 may have the amino acid sequence of human CXCR4 with GenBank accession number P61073. A CXCR4 sequence may differ from human CXCR4 having GenBank accession number P61073, by having e.g., mutations. Said mutations preferably do not influence the biological function of CXCR4. CXCR4 is also known in the art as leukocyte-derived seven-transmembrane domain receptor (LESTR), Fusin, and CD184. ?n dn pn`_ c`m`di, oc` o`mh ^CXCR4 cluster^, refers to a group of two or moreCXCR4 receptors organized as a dimer or higher-order oligomer. ?n dn pn`_ c`m`di, oc` o`mh ^CXCR4 cluster dicd]dojm^ jm ^CXCR4 clusterhjijh`mdudib \b`io^,AVAP4 ^gpno`m _dnmpkojm^ jm ^AVAP4 hjijh`mdudib \b`io^, refersto a molecule that is able to monomerize a CXCR4 cluster. Examples of known CXCR4 cluster disrupters include small molecules IT1t and AMD070, as well as antibodies such as 2D01, VUN401, VUN411, VUN418, VUN419 and VUN421. Preferred CXCR4 cluster disrupters are selected from the group consisting of IT1t, VUN401, 2D01, VUN411, VUN418, VUN419 and VUN421. CXCR4 monomerization of a CXCR4 cluster can be determined by, e.g., bioluminescence resonance energy transfer (BRET) method, time- resolved fluorescence resonance energy transfer (TR-FRET), localization microscopy method such as direct stochastic optical reconstruction microscopy (dSTORM), or spatial- intensity Distribution Analysis (SpiDA), as explained herein below. ?n dn pn`_ c`m`di, oc` o`mh ^tumor^, m`a`mn oj a neoplastic cell growth, whethermalignant or benign, and pre-cancerous and cancerous cells and tissues. As used herein, the term "cancer" refers to a malignant tumor, i.e., an uncontrolled cellular growth that can spread into, or invade, nearby tissues and can travel to distant places in the body to form new tumors (a process called metastasis). ?n dn pn`_ c`m`di, oc` o`mh ^nh\gg hjg`^pg`^, m`a`mn oj \ hjg`^pg` with a molecularweight less than 10,000 Dalton (D), preferably less than 5,000 D. Said molecule preferably is an organic molecule. Said molecule may bind to a specific target molecule. Said binding to a target molecule may alter the activity or function of the target molecule. A small molecule can have a variety of biological functions or applications, serving as a cell signaling molecule, as a drug in medicine, or as a pesticide in farming. A small molecule can be a natural molecule (such as a secondary metabolite) or an artificial molecule (such as an antiviral drug). A small molecule may have a beneficial effect in the treatment of a disease. ?n dn pn`_ c`m`di, oc` o`mh ^\iod]j_t^ refers to a molecule that specifically binds toan antigen. Antibodies can be intact immunoglobulins derived from natural sources or from recombinant sources, and can be immunoreactive fragments of said immunoglobulins. Classical antibodies are often tetramers of immunoglobulin molecules.?iod]j_d`n `sdno di \ q\md`ot ja ajmhn. Rc` o`mh ^\iod]j_t^ di^gp_`n i\opm\ggt j^^pmmdiband non-naturally occurring binding molecules such as a designed ankyrin repeat protein, a binding protein that is based on a Z domain of protein A, a binding protein that is based on a fibronectin type III domain, engineered lipocalin, and a binding protein that is based on a human Fyn SH3 domain (Skerra, 2007. Current OpinionBiotechnol 18: 295-304; ^fmg`^ et al., 2015. Trends Biotechnol 33: 408-418), maxibodies,minibodies, intrabodies, diabodies, triabodies, tetrabodies, v-NAR and bis-scFv (see, e.g., Hollinger & Hudson, 2005. Nature Biotech. 23:1126-1136), polyclonal and monoclonal antibodies, one-armed antibodies, chimeric antibodies, wholly synthetic antibodies and fragments thereof, such as for example, the Fab', F(ab')2, Fv or Fab fragments, or other antigen recognizing immunoglobulin fragments. ?n dn pn`_ c`m`di, oc` o`mh ^\iod]j_t am\bh`io^, m`a`mn oj \ kjmodji ja \i intactantibody and more specifically refers to the antigenic determining variable regions of anantibody. Examples of antibody fragments include D\], D\]^, D(\]^)2, and Fv fragments,linear antibodies, scFv antibodies, single-domain human antibodies (sdAb) and multivalent and / or multispecific antibody fragments such as a bivalent fragment comprising two Fab fragments linked by a disulfide bridge. ?n dn pn`_ c`m`di, oc` o`mh ^single-domain antibody^, m`a`mn oj \ VHH antibody, ascan be obtained from a camelid. A VHH comprises a single heavy chain monomeric variable antibody region. A VHH is able to bind selectively to a specific antigen. With a molecular weight of only 12^15 kDa, a single-domain antibody is much smaller than common antibodies (150^160 kDa) which are composed of two heavy protein chains and two light chains, and even smaller than a Fab fragment (~50 kDa, one light chain and half a heavy chain) and single-chain variable fragments (~25 kDa, one light chain and one heavy chain variable region. ?n dn pn`_ c`m`di, oc` o`mh ^^jh]di\odji^ m`a`mn oj oc` \_hdidnom\odji ja `aa`^odq`amounts of a CXCR4 cluster inhibitor and a cell death-inducing compound to a patient in need thereof. Said CXCR4 cluster inhibitor and cell death-inducing compound may be provided in one pharmaceutical preparation, or as two distinct pharmaceutical preparations. In methods of the invention, a CXCR4 cluster inhibitor may be administrated simultaneously with, or sequentially to the cell death-inducing compound. When administered as two distinct pharmaceutical preparations, they may be administered on the same day or on different days to a patient in need thereof, and using a similar or dissimilar administration protocol, e.g. daily, twice daily, biweekly, orally and / or by infusion. Said combination is preferably administered repeatedly according to a protocol that depends on the patient to be treated (age, weight, treatment history, etc.), which can be determined by a skilled physician. Said protocol may include daily administration for 1-30 days, such as 2 days, 10 days, or 21 days, followed by period of 1- 14 days, such as 7 days, in which no compound is administered. ?n dn pn`_ c`m`di, oc` o`mh ^pharmaceutically \^^`ko\]g` ^jhkjndodji^, m`a`mn oj \composition that is suitable for administration to an individual such as an animal or a human. CXCR4 cluster inhibitors In methods of the invention, a cell is provided with a CXCR4 cluster inhibitor and a cell death-inducing compound. Said cell is a cell comprising a CXCR4 cluster, meaning that the cell expresses a group of CXCR4 receptors organized as a dimer or higher-order oligomer. In methods of the invention, said cell may be a tumor cell comprising a CXCR4 cluster. In particular, tumor cells that are known to comprise CXCR4 clusters include tumor cells from lymphoma, such as Non-Hodgkin Lymphoma (NHL), myeloma such as multiple myeloma (MM), breast cancer, prostate cancer, ovarian cancer, lung cancer, gastro-intestinal cancer, renal cell carcinoma, melanoma, brain cancer and soft tissue sarcoma. In a method according to the invention, the CXCR4 cluster inhibitor may be any compound that is able to monomerize a CXCR4 cluster. In a method according to the invention, the CXCR4 cluster inhibitor may be a small molecule, an antibody, antibody fragment, or a combination thereof. Examples of suitable small compound molecules that act as CXCR4 cluster inhibitors include 6,6-dimethyl-5H-imidazo[2,1-b][1,3]thiazol-3-yl)methyl N,N'- dicyclohexylcarbamimidothioate; dihydrochloride (IT1t). A further CXCR4 inhibitor is (S) N'-(1H-benzimidazol-2-ylmethyl)-N'-[(8S)-5,6,7,8-tetrahydroquinolin-8-yl]butane-1,4- diamine; AMD070; also termed Mavorixafor). Examples of a suitable antibody that acts as a CXCR4 cluster inhibitor include VUN401, VUN411, VUN418, VUN419 and VUN421, and 2D01. The amino acid sequences of these antibodies are shown in Table 1. These molecules are able to monomerize CXCR4 and thereby inhibit clustering-mediated signaling. A bivalent version of these single domain antibodies or Fab, for example a homodimer generated by fusing 2D01 to an antibody fragment crystallizable (Fc) region, did not disturb CXCR4 clustering. A CXCR4 cluster inhibitor may also be an antibody or antibody fragment conjugated to a small molecule. Preferred conjugates of an antibody fragment and a small molecule for use in a composition or method of the invention are conjugates of VUN401, VUN411, VUN416, VUN418, VUN419, VUN421, or 2D01 with IT1t, or conjugates of VUN401, VUN411, VUN416, VUN418, VUN419 and VUN421, or Fab fragment 2D01 with and AMD070. Methods to provide conjugates of antibodies are known in the art, including maleimide-thiol chemistry. For this, an antibody may be reduced by tris(2-carboxyethyl)phosphine (TCEP), followed by incubating the antibody with a small molecule. Non-reacted maleimide groups may be quenched using L-cysteine and an conjugated antibody may be isolated, for example by dialysis. Oligomerization of membrane receptors is common, also among GPCRs, and can be studied using a variety of techniques, including, for example, 2-photon fluctuation microscopy with scanning number and brightness analysis (Møller et al., 2018. Sci Rep 8: 10414), and proximity-ligation assay (PLA) (Laufer et al., 2018. Front Immunol 9: 3115). In addition, resonance energy transfer (RET) methods may be used to detect oligomerization of GPCRs, such as the use of antibody-based time-resolved Förster resonance energy transfer (TR-FRET) sensors by Meng et al., 2022 (Meng et al., 2022. Nature Chem Biol 18: 894-903), and can be used for determining relative differences in cluster sizes. Furthermore, bioluminescence resonance energy transfer (BRET) can be used to study oligomerization of GPCRs, including CXCR4. BRET is based on energy transfer between a light-emitting molecule (typically a luciferase) and a light-sensitive molecule (typically a fluorescent protein). A standard luciferase is, or is derived from the firefly, or the sea pansy Renilla reniformis. Said luciferase preferably is a recently discovered luciferase from a shrimp, Oplophorus gracilirostris, which is a 19 kDa protein that uses coelenterazine as a substrate. Modification of this enzyme resulted in an enzyme termed NanoLuc, which uses furimazine as substrate, and shows a more than 150-fold increase in luminescence, when compared to a standard luciferase. A further improved small luciferases includes GLuc, isolated from Gaussia princeps (Verhaegen and Christopoulos, 2002. Anal Chem 74: 4378^4385). A BRET method that is based on NanoLuc has been termed NanoBRET (Mo and Fu, 2016. Methods Mol Biol 1439: 263-271), while a NanoBRET method involving camelid single-domain antibodies was termed NanoB2(van den Bor et al., 2023. Cell Rep Methods 3: 100422). Any one of these methods may be employed to study the oligomerization status of membrane receptors such as CXCR4. As such, an antibody-based BRET method comprising NanoLuc- and fluorophore-labeled non-competitive antibodies is a suitable method to demonstrate the presence or disruption of CXCR4 clusters, e.g., in tumor cells. For example, in order to determine if a tumor cell comprises a CXCR4 cluster, an antibody-based BRET method may be used. In a BRET assay bioluminescence resonance energy transfer will occur when donor and acceptor are in close-proximity of each other, for example when two or more CXCR4 receptors are clustered. Such BRET signal is typically measured as a ratio between light emitted by the acceptor and light emitted by the donor. A BRET signal will increase as two or more CXCR4-receptors are present in a cluster. In contrast, no BRET signal will be apparent if a cell does not comprise a CXCR4 cluster. For such BRET assay, an antibody may be used that binds to CXCR4, while not changing the oligomeric state of the CXCR4 receptor, such as antibody VUN415. A suitable antibody such as VUN415 may be labeled with a luciferase, preferably NanoLuc, as a suitable donor, while antibody VUN415 conjugated with fluorescent dye ATTO565 or with AlexaFluor 488 may be used as a suitable acceptor for the determination of a CXCR4 cluster in a cell using BRET technology. BRET may be determined in the presence and absence of a known CXCR4 cluster inhibitor, such as IT1t or single-domain antibody VUN401. A high BRET signal in the absence of the CXCR4 cluster inhibitor, and / or a reduction of the signal in the presence of the CXCR4 cluster inhibitor is indicative of CXCR4 clustering. Although BRET methods are suitable for determining relative differences in cluster sizes, they do not allow for absolute oligomeric stoichiometry measurements. To obtain stoichiometric insights, for example a single-molecule localization microscopy method such as direct stochastic optical reconstruction microscopy (dSTORM) may be used, using a CXCR4-targeting antibody. A further method that may be used is spatial- intensity distribution analysis (SpiDA) to study the oligomerization status of GPCRs (Pediani et al., 2018. Trends in Pharmacological Sciences 39: 175-186). Table 1: Amino-acid sequences of antibodies that function as CXCR4 cluster inhibitor. Antibody Sequence fragment name VUN401 QVQLQESGGGLVQAGGSLRLSCAASGIRFSLYDMGWYRQAPGNQRELVATISTLSTRYADSVKGRFTISRDNAKNTVYLQMNSLKPEDTAVYY CNAKRYRTNYWGQGTQVTVSS VUN411 QLQLVESGGGLVQAGGSLRLSCAASGSTFSNTAMGWYRQAPGKQRELVVAISSGGSTYYVDSVKGRFTISRDSAKNTVYLQMNSLKPEDTAAYY CNSGSWARRSRTYLTGSYWGQGTQVTVSS VUN418 QLQLVESGGGSVQAGGSLRLSCAASGITFSSYRMGWYRQAPGKVRQLVASISNGGMTNYADSVKGRFTISRDNAKNTVYLQMNSLKPEDTAVY YCNAGTSTAARYMGRYWGQGTQVTVSS VUN419 QLQLVESGGGLVQAGGSLRLSCAASGRTFSSYTMGWFRQAPGKEREFVAYISWSSTYYADIVKGRFTISRDNAKNTVFLQMDYLKPEDTAVYY CAAGSLGGRSRYFRGDYWGQGTQVTVSS VUN421 ELQLVESGGGLVQAGGSLRLSCTASGRAFSRYAMGWFRQAPGQERERVAAIGWSPSKTYYADSVKGRFTVSRDNNKNTVYLQMNSLKPEDTA VYYCAAKYSNSDATFRSDYNYWGQGTQVTVSS 2D01 VHQVQLVESGGGLVQPGGSLTLSCAASGFTFRNYAMSWVRQAPGKGLE WLSVIHSSGSSTRYSDSVKGRFTVSRDNAQNTLYLQMSSLKPEDTAL YYCARGQRYIASWGQGTQVTVSS VL ATMLTQSPGSLSVVPGESVTISCKASQSLVHTDGKTYLSWLLQKPGQ RPQLLIYQVSNRASGVPDRFTGSGSGTDFTLKISGVKAEDAGVYYCA QATYYPLTFGQGTKVELK Table 2: Amino-acid sequences of further antibodies Antibody Sequence fragment name VUN415 QLQLVESGGGLVQPGGSHRLSCAASGSILSRNAMGWYRQAPGKQRELVATITGPITSGGRTNYAESIKGRFTISRDNAKNTVYLQMNSLKPEDT AVYYCTIGRASRSGSYYPDYWGQGTQVTVSS VUN416 QLQLVESGGGLVQAGGSLRLSCAASTSTSSITAIGWYRQVPGKQRELVAGISSSGSLRTNYADSVKGRFTISRDNAKNTVYLQMNSLKPEDTAV YYCNAGAMARRSGAYYPGAYWGQGTQVTVSS Cell death inducing compounds Examples of cell death-inducing compounds include classical anticancer agents such as anthracyclines, antimetabolites and platinum drugs; topoisomerase inhibitors, e.g., camptothecin; cell death-inducing cytokines, such as IFN-^, IFN-^, TNF-^, and TRAIL; and taxanes, such as paclitaxel and fluorinated taxanes. A broad variety of synthetic and natural apoptosis inducers have been described in Billard et al. (Billard et al., 2014. Oncotarget. 5: 309^325) and include inhibitors of prosurvival protein expression such as cyclin-dependent kinase inhibitors e.g. flavopiridol, roscovitine, dinaciclib and SNS-032, translational inhibitors e.g. homoharringtonine and silvestrol, small interfering RNA and antisense oligonucleotides, e.g., oblimersen; inhibitors of prosurvival protein activity such as BH3 mimetics and second mitochondria-derived activator of caspase (SMAC) mimetics; enhancers of proapoptotic protein expression such as proteasome inhibitors, e.g., bortezomib, lactacystin, MG-132 and carfilzomib, plant- derived proteasome inhibitors, e.g., epigallocatechin gallate (EGCG), quercetin, apigenin and xanthohumol; histone deacetylase inhibitors such as depsipetide, valproic acid, MGCD0103 (N-(2-aminophenyl)-4-({[4-(pyridin-3-yl)pyrimidin-2-yl]amino} methyl)benzamide), and vorinostat; activators of apoptotic pathways such as p53 activators such as nutlin-3a and PARP inhibitor; inhibitors of survival pathways such as nuclear factor-kB inhibitors including BAY-117082 ((2E)-3-[(4- methylphenyl)sulfonyl]acrylonitrile), phosphoinositol-3 kinase / AKT inhibitors e.g. CAL- 101 ((S)-2-(1-(9H-Purin-6-ylamino)ethyl)-5-fluoro-3-phenylquinazolin-4(3H)-one) and Akt-1 / 2 kinase inhibitor (3-[1-[[4-(6-phenyl-8H-imidazo[4,5-g]quinoxalin-7- yl)phenyl]methyl]piperidin-4-yl]-1H-benzimidazol-2-one;2,2,2-trifluoroacetic acid; hydrate), and inhibitors of microenvironment signals such as lenalidomide and plerixafor; and modulators of other signaling pathways, including kinase inhibitors such as . fostamatinib and dasatinib, sorafenib, imatinib and fenretinide. Preferably, in a method or composition according to the invention, the cell death- inducing compound is a TNF-related ligand and / or a BCL-2 inhibitor. Tumor necrosis factor (TNF)-alpha related ligands such as TNF-alpha, Fas Ligand (FasL), and TNF-related apoptosis-inducing ligand (TRAIL) can induce apoptosis, a regulated form of cell death that keeps the cell population balanced in an organism. It is normally used to prevent the accumulation of damaged or infected cells. TNF-alpha related ligands are being used for cancer treatment because of their capability of specific apoptosis induction in transformed cells (Walczak, 2013. Cold Spring Harb Perspect Biol 5: 1^18). TNF-alpha related ligands normally act as trimeric soluble proteins. These proteins can interact with their respective receptors but sometimes cannot activate them. Studies have shown that the re-enforcement or improvement of the trimeric conformation may improve apoptosis-inducing activity. Common approaches use leucine zippers, His-tagged proteins, and covalent trimerization domains (Berg et al., 2007. Cell Death Differ 14: 2021^2034). For example, APO010 (MegaFasL) is a hexameric fusion protein created by the fusion of the collagen domain adiponectin to two FasL extracellular domain trimers (Eisele et al., 2011. Neuro Oncol 13: 155^164). Viral vectors may be employed to deliver a TNF-alpha related ligand and to avoid repeated administration. Suitable TNF-alpha related ligands further include trail receptor agonistic antibodies such as anti-death receptor 4 (DR4) or anti-DR5 agonistic antibodies (Qiu et al., 2012. IUBMB Life 64: 757^765; Dai et al., 2015. Exp Biol Med 240: 760^773), and multivalent anti-DR4 and anti-DR5 antibodies (Dubuisson and Micheau, 2017. Antibodies 6: 16). Suitable BCL-2 inhibitors include the small molecules venetoclax (ABT-199; 4-[4- [[2-(4-chlorophenyl)-4,4-dimethylcyclohexen-1-yl]methyl]piperazin-1-yl]-N-[3-nitro-4- (oxan-4-ylmethylamino)phenyl]sulfonyl-2-(1H-pyrrolo[2,3-b]pyridin-5-yloxy)benzamide), navitoclax (ABT-263; 4-[4-[[2-(4-chlorophenyl)-5,5-dimethylcyclohexen-1- yl]methyl]piperazin-1-yl]-N-[4-[[(2R)-4-morpholin-4-yl-1-phenylsulfanylbutan-2- yl]amino]-3-(trifluoromethylsulfonyl)phenyl]sulfonylbenzamide), ABT737 (4-[4-[[2-(4- chlorophenyl)phenyl]methyl]piperazin-1-yl]-N-[4-[[(2R)-4-(dimethylamino)-1- phenylsulfanylbutan-2-yl]amino]-3-nitrophenyl]sulfonylbenzamide), maritoclax ([4,5- dichloro-1-[4,5-dichloro-2-(2-hydroxybenzoyl)-1H-pyrrol-3-yl]pyrrol-2-yl]-(2- hydroxyphenyl)methanone), clitocine ((2R,3R,4S,5R)-2-[(6-amino-5-nitropyrimidin-4- yl)amino]-5-(hydroxymethyl)oxolane-3,4-diol), obatoclax ((2Z)-2-[(5Z)-5-[(3,5-dimethyl- 1H-pyrrol-2-yl)methylidene]-4-methoxypyrrol-2-ylidene]indole), subatoclax (2-(2-((3,5- dimethyl-1H-pyrrol-2-yl)methylene)-3-methoxy-2H-pyrrol-5-yl)-1H-indole), gossypol (7- (8-formyl-1,6,7-trihydroxy-3-methyl-5-propan-2-ylnaphthalen-2-yl)-2,3,8-trihydroxy-6- methyl-4-propan-2-ylnaphthalene-1-carbaldehyde), apogossypol (3-methyl-5-propan-2-yl- 2-(1,6,7-trihydroxy-3-methyl-5-propan-2-ylnaphthalen-2-yl)naphthalene-1,6,7-triol), TW- 37 (N-[4-(2-tert-butylphenyl)sulfonylphenyl]-2,3,4-trihydroxy-5-[(2-propan-2- ylphenyl)methyl]benzamide), UMI-77 (2-[4-[(4-bromophenyl)sulfonylamino]-1- hydroxynaphthalen-2-yl]sulfanylacetic acid), BDA-366 (1-[[(2S)-3-(diethylamino)-2- hydroxypropyl]amino]-4-[[(2S)-oxiran-2-yl]methylamino]anthracene-9,10-dione) and AT- 101 ((3S,3aR,6R,6aR)-2,3,3a,5,6,6a-hexahydrofuro[3,2-b]furan-3,6-diol), and the antisense compound oblimersen (heptadecasodium), which is complementary to the first 6 codons of Bcl-2 mRNA. Most cell-death inducing compounds are being used, or are being developed, for treating cancer. However, they may induce cell death in normal tissues, such as in epithelial cells, which raises the concern for side effects during tumor treatment. Cisplatin, for example, may also trigger cell death in various normal cells (Wang et al., 2017. Nature 547: 99^103). In addition, many tumors are resistant to therapeutic agents, or become resistant during treatment with a cell-death inducing compound, which may be related to downregulation of cell death pathways. A combination of a CXCR4 cluster inhibitor with a cell-death inducing compound may enhance the cell- death inducing activity of a cell death-inducing compound. In addition, a combination of a CXCR4 cluster inhibitor with a cell-death inducing compound may reduce intrinsic or acquired resistance to a cell death inducing compound. A preferred cell death-inducing compound is a BCL-2 inhibitor such as venetoclax. Venetoclax is a BCL-2 protein antagonist that induces the activation of pro-apoptotic proteins, such as BAX and BAK, and is clinically used to treat cancers such as chronic lymphocytic leukemia (CLL) and small lymphocytic lymphoma (SLL). Long-term treatment is often associated with acquired resistance and serious side effects. It is now found that a CXCR4 cluster inhibitor, by disrupting basal signaling of a CXCR4 cluster, enhances the cell death-inducing activity of a cell death-inducing compound. This means that a lower dose of a cell death-inducing compound can be used to reach the same or similar cell death-inducing effect, which will result in a reduction of treatment costs and side-effects. The described co-treatment may be effective for all tumors comprising CXCR4 clustering. In addition, a sensitizing CXCR4 cluster inhibitor may reduce the chance of tumor cells becoming resistant to a cell death-inducing compound such as venetoclax. Methods of potentiating a cell death-inducing compound In a method of potentiating a cell death-inducing compound, the potency of said compound is increased. An increased potency of a compound can be measured as an increased activity of the compound. As shown in the examples below, the potency of a cell death inducing compound such as venetoclax may be determined by measuring the metabolic activity of cells after treatment with the cell death inducing compound. Methods to determine metabolic activity of a cell are known in the art, including a cell viability assay. For example, mitochondrial respiration and function may be examined by measuring key biomarkers of mitochondrial activity such as succinate dehydrogenase subunit A (SDHA) and / or cytochrome c oxidase subunit I (COX1). In addition, mitochondrial integrity can be measured by examining biomarkers including cytochrome C, which is released from mitochondria into the cytosol during the initiation of apoptosis. Furthermore, suitable cell viability assays include tetrazolium reduction, resazurin reduction, and protease activity assays. These assays include incubation of a cells with a reagent that is converted to a colored or fluorescent product that can be detected. A suitable assay is provided by a Resazurin-based method, in the absence and presence of compounds. Compounds that in combination with a cell-death inducing compound such as venetoclax were found to reduce the metabolic activity of certain cells compared to the metabolic activity of those cells when venetoclax is administered alone, can be considered compounds that potentiate the cytotoxic activity of venetoclax. The invention relates to methods of identifying a CXCR4 cluster inhibitor that potentiates a cell-death inducing compound. For this, a cell comprising a CXCR4 cluster may be identified as described herein above. Said cell is incubated with a cell-death inducing compound in the presence and absence of a potential CXCR4 cluster inhibitor. Preferably, said cell is incubated with different concentrations of a cell-death inducing compound, and with different concentrations of a potential CXCR4 cluster inhibitor. Following this method, a CXCR4 cluster inhibitor may be identified that results in killing of cells at a lower concentration of a cell-death inducing compound, when compared to a control, for example in the absence of said CXCR4 cluster inhibitor. The invention relates to a method of potentiating a cell death-inducing compound, comprising providing a cell with a combination of a CXCR4 cluster inhibitor and the cell death-inducing compound, wherein the cell is a tumor cell comprising a CXCR4 cluster. The invention relates to a method of potentiating a cell death-inducing compound, comprising providing a tumor cell comprising a CXCR4 cluster with a CXCR4 cluster inhibitor and the cell death-inducing compound. The invention further relates to a CXCR4 cluster inhibitor and a cell death- inducing compound for use in a method of potentiating a cell death-inducing compound, wherein the CXCR4 cluster inhibitor and the cell death-inducing compound are provided to a tumor cell comprising a CXCR4 cluster. The invention further relates to a CXCR4 cluster inhibitor for use in a method of potentiating a cell death-inducing compound, wherein the CXCR4 cluster inhibitor is provided to a tumor cell comprising a CXCR4 cluster, in combination with the cell death- inducing compound. The invention further relates to the use of a CXCR4 cluster inhibitor and a cell death-inducing compound in the preparation of a medicament for potentiating a cell death-inducing compound, wherein the CXCR4 cluster inhibitor and the cell death- inducing compound are provided to a tumor cell comprising a CXCR4 cluster. The invention further relates to the use of a CXCR4 cluster inhibitor in the preparation of a medicament for potentiating a cell death-inducing compound, wherein the CXCR4 cluster inhibitor is provided to a tumor cell comprising a CXCR4 cluster, in combination with the cell death-inducing compound. In embodiments, said CXCR4 cluster inhibitor may be conjugated to the cell death- inducing compound. Compositions The invention further provides a composition comprising a CXCR4 cluster inhibitor and a cell death-inducing compound, preferably a pharmaceutical composition comprising a CXCR4 cluster inhibitor and a cell death-inducing compound. A pharmaceutical composition according to the invention preferably comprises between 1 microgram and 1 gram of a CXCR4 cluster inhibitor, and between 1 microgram and 1 gram of cell death-inducing compound. The exact dosage of a CXCR4 cluster inhibitor and a cell death-inducing compound will be determined by the individual to which the dose is administered, in light of factors related to the individual. Said dosage preferably is between 0.1 microgram and 10 milligram per kg. Dosage and administration may be adjusted to provide sufficient levels of the CXCR4 cluster inhibitor and the cell death- inducing compound, or to maintain the desired therapeutic effect for these compounds. Factors that can be taken into account include the general health of the subject, age, weight and gender of the subject, diet, time and frequency of administration, drug combination(s), reaction sensitivities and tolerance / response to therapy, as is known to a person skilled in the art. A pharmaceutical composition according to the invention may be provided to an individual by systemic or by local administration, including parenteral administration such as injection and / or infusion. A preferred route of administration is oral administration. However, for example antibodies are generally administered intravenously, but other routes such as subcutaneous or aerosol delivery could be preferred for some applications. A pharmaceutical composition according to the invention may comprise a CXCR4 cluster inhibitor and a cell death-inducing compound, or comprises a kit of parts comprising a composition comprising the CXCR4 cluster inhibitor and a composition comprising the cell death-inducing compound. A kit of part may be convenient in case the CXCR4 cluster inhibitor is to be administered through a different route when compared to the cell death-inducing compound. A pharmaceutical composition according to the invention may comprise a pharmaceutically acceptable excipient. Said pharmaceutically acceptable excipient preferably is selected from a diluent, a binder, a granulating ingredient, a carbohydrate such as starch, a starch derivative such as starch acetate and / or maltodextrin, a polyol such as xylitol, sorbitol and / or mannitol, lactose such as ^-lactose monohydrate, anhydrous ^-lactose, anhydrous ^-lactose, spray-dried lactose, and / or agglomerated lactose, a sugar such as dextrose, maltose, dextrate and / or inulin, a glidant, and a lubricant, and combinations thereof. Said pharmaceutical composition preferably is for oral administration. A pharmaceutical composition for parenteral administration preferably is a sterile isotonic solution. Said solution preferably is buffered. A preferred buffer is a citrate- based buffer, preferably lithium-, sodium-, potassium-, or calcium- citrate monohydrate, citrate trihydrate, citrate tetrahydrate, citrate pentahydrate, or citrate heptahydrate; lithium, sodium, potassium, or calcium lactate; lithium, sodium, potassium, or calcium phosphate; lithium, sodium, potassium, or calcium maleate; lithium, sodium, potassium, or calcium tartarate; lithium, sodium, potassium, or calcium succinate; or lithium, sodium, potassium, or calcium acetate, or a combination of two or more of the above. The pH of said buffer may be adjusted, preferably to a pH of 7.2 ^ 7.5 by hydrochloric acid, sodium hydroxide, citric acid, phosphoric acid, lactic acid, tartaric acid, succinic acid, or a combination of two or more of the above. The volume may range from 0.5 ml to 5 ml. An excipient preferably is selected from, but not limited to, urea, L-histidine, L-threonine, L- asparagine, L-serine, L-glutamine, polysorbate, polyethylene glycol, propylene glycol, polypropylene glycol, or a combination of two or more of the above. Methods of treating a tumor The invention relates to a method of treating a tumor comprising CXCR4 clusters in a subject, comprising simultaneous or sequential administering a CXCR4 cluster inhibitor and a cell death-inducing compound to the subject. Methods of treating a tumor according to the invention may include a step of analyzing the presence of CXCR4 clusters in cells of the tumor, and simultaneous or sequential administering a CXCR4 cluster inhibitor and a cell death-inducing compound to a subject that comprises CXCR4 clusters in cells of the tumor. Methods of determining or analyzing the presence of CXCR4 cluster may include bioluminescence resonance energy transfer (BRET) method, time-resolved fluorescence resonance energy transfer (TR-FRET), localization microscopy method such as direct stochastic optical reconstruction microscopy (dSTORM), or spatial-intensity Distribution Analysis (SpiDA), as explained herein. The invention further relates to a CXCR4 cluster inhibitor and a cell death- inducing compound for use in a method of treating a tumor comprising CXCR4 clusters in a subject, wherein the CXCR4 cluster inhibitor and the cell death-inducing compound are administered simultaneous or sequential. The invention further relates to the use of a CXCR4 cluster inhibitor and a cell death-inducing compound for the preparation of a medicament for treating a tumor comprising CXCR4 clusters in a subject, wherein the CXCR4 cluster inhibitor and the cell death-inducing compound are administered simultaneous or sequential. The invention further relates to a CXCR4 cluster inhibitor for use in a method of treating a tumor comprising CXCR4 clusters in a subject, wherein the CXCR4 cluster inhibitor is administered simultaneous or sequential with a cell death-inducing compound. The invention further relates to the use of a CXCR4 cluster inhibitor for the preparation of a medicament for treating a tumor comprising CXCR4 clusters in a subject, w wherein the CXCR4 cluster inhibitor is administered simultaneous or sequential with a cell death-inducing compound. The invention further relates to a cell death-inducing compound for use in a method of treating a tumor comprising CXCR4 clusters in a subject, wherein the cell death-inducing compound is administered simultaneous or sequential with a CXCR4 cluster inhibitor. The invention further relates to the use of a cell death-inducing compound for the preparation of a medicament for treating a tumor comprising CXCR4 clusters in a subject, wherein the cell death-inducing compound is administered simultaneous or sequential with a CXCR4 cluster inhibitor. A CXCR4 cluster inhibitor preferably is IT1t, VUN401, 2D01, VUN411, VUN418, VUN419, VUN421, or any combination thereof, more preferably IT1t and / or VUN401. As is shown herein below, IT1t and VUN401 surprisingly increased the sensitivity to a cell death-inducing agent such as venetoclax to a much higher level, when compared to other CXCR4 inhibitors such as AMD070. Simultaneous administration refers to administration of the CXCR4 cluster inhibitor and the cell death-inducing compound together, for example as a pharmaceutical composition containing both agents, or immediately after each other and optionally via the same route of administration, e.g., oral or parenteral administration. . Sequential administration refers to administration of one of the CXCR4 cluster inhibitor or the cell death-inducing compound, followed after a period of time by separate administration of the other agent. It is not required that the two agents are administered by the same route, although this may be the case. The time interval may be any time interval. Preferably, when the CXCR4 cluster inhibitor or the cell death- inducing compound are administered sequentially, the time interval is selected such that the CXCR4 cluster inhibitor has taken effect prior to the administration of the cell death-inducing compound. For example, said time interval may be 5 minutes or less, 10 minutes or less, 15 minutes or less, 20 minutes or less, 25 minutes or less, 30 minutes or less, 45 minutes or less, 60 minutes or less, 90 minutes or less, 120 minutes or less, 180 minutes or less, 240 minutes or less, 300 minutes or less, 360 minutes or less, or 720 minutes or less, 1 day or less, or 2 days or more. Administration of the CXCR4 cluster inhibitor and the cell death-inducing compound for use according to the invention is preferably in a therapeutically effective amount. The actual amount administered, and rate and time-course of administration, will depend on the nature and severity of the disease being treated. A person skilled in the art knows well the determination of the dosage or route of appropriate administration, and typically takes account of the disorder to be treated, the condition of the individual subject, the site of delivery, the method of administration and other factors. A tumor cell to be treated according to the invention can be any histological type of cancer. The tumor to be treated can be a carcinoma e.g. an adenocarcinoma, or squamous cell carcinoma; a sarcoma such as osteosarcoma or osteogenic sarcoma, a chondrosarcoma, a leiomyosarcoma, a rhabdomyosarcoma, a mesothelial sarcoma or mesothelioma, a fibrosarcoma, an angiosarcoma, a hemangioendothelioma, a liposarcoma, a glioma. an astrocytoma, a myxosarcoma, a mesenchymous or mixed mesodermal tumor; a myeloma, a leukemia, such as a lymphatic, lymphocytic, or lymphoblastic leukemia, an erythremia, a lymphoma, e.g., Hodgkin lymphoma and non- Hodgkin lymphoma; and / or a mixed type cancer such as an adenosquamous carcinoma, a mixed mesodermal tumor, a carcinosarcoma and a teratocarcinoma. A tumor to be treated according to the invention can be any topological type of cancer, in particular a tumor selected from the group of bladder cancer, bone cancer such as osteosarcoma, brain cancer, breast cancer, cervical cancer, colorectal cancer, eye cancer, head and neck cancer, kidney cancer, liver cancer, lung cancer, ovarian cancer, pancreas cancer, prostate cancer, skin cancer such as melanoma, testicular cancer, thyroid cancer, uterine cancer, vaginal cancer, vulvar cancer, mouth cancer and throat cancer. In particular, tumors that are known to present CXCR4 clusters include Non- Hodgkin Lymphoma (NHL), multiple myeloma (MM), breast cancer, prostate cancer, ovarian cancer, lung cancer, gastrointestinal cancers, renal cell carcinoma, melanoma, brain cancer and soft tissue sarcoma, in particular a lymphoma, myeloma, leukemia, glioma, rhabdomyosarcoma or prostate cancer, especially a non-Hodgkin lymphoma. A tumor that is not known to present CXCR4 clusters may be analyzed for the presence of CXCR4 clustering by, for example, an antibody-based BRET method as indicated herein above. For the purpose of clarity and a concise description, features are described herein as part of the same or separate aspects and preferred embodiments thereof, however, it will be appreciated that the scope of the invention may include embodiments having combinations of all or some of the features described. The invention will now be illustrated by the following examples, which are provided by way of illustration and not of limitation and it will be understood that many variations in the methods described and the amounts indicated can be made without departing from the spirit of the invention and the scope of the appended claims. EXAMPLES Materials and methods used in the examples DNA constructs and molecular cloning The pcDEF3 vector was obtained from Langer (Goldman et al., 1996. BioTechniques 21: 1013-1015). cDNA encoding the BRET-based cAMP biosensor His- CAMYEL pcDNA3.1(L) was purchased from ATCC (#ATCC-MBA-277). pLKO.1 puro CXCR4 siRNA-1 and siRNA-2 were obtained from Bob Weinberg (Addgene plasmid #12271). plKO.1 scramble shRNA was a gift from David Sabatini (Addgene plasmid #1864). Myc-CXCR4-Rluc pIRES, HA-CXCR4-YFP pIRES (Paradis et al.¸ 2022. Nat Commun 13: 6826), HA-CXCR4 WT pcDEF3 (Watts et al., 2013. Br J Pharmacol 168: 1662- 1674), NanoLuc-CXCR4 pcDNA3.1 (van den Bor et al., 2023. Cell Rep Methods 3: 100422) and CXCR4-EYFP pcDNA3 (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154) were described previously. The HA-CXCR4-N119S pcDEF3 plasmid was generated by replacing CXCR4 WT for the CXCR4 N119S sequence using the Acc65I and the internal BamHI restriction sites. HA-CXCR4 pLenti6.3 / TO / V5-DEST was generated by exchanging US28 for CXCR4 in the previously described HA-US28 pLenti6.3 / TO / V5-DEST plasmid (de Wit et al., 2016. Oncotarget 7: 67966-67985; van Senten et al., 2019. J Biol Chem 294: 16297-1630). HA-CXCR4 pEUI was generated by exchanging VUN103-FLAG for HA- CXCR4 in the previously described VUN103-FLAG pEUI plasmid (de Groof et al., 2021. Nat Commun 12: 4357). Patient material After written informed consent, patient blood samples were obtained during diagnostic or follow-up procedures at the Departments of Hematology and Pathology of the Academic Medical Center Amsterdam. This study was approved by the AMC Ethical Review Biobank Board under the number METC 2013 / 159 and conducted in accordance with the Declaration of Helsinki. Peripheral blood mononuclear cells (PMBCs) of patients with CLL, obtained after Ficoll density gradient centrifugation (Pharmacia Biotech), were cryopreserved and stored as previously described (Hallaerts et al., 2008. Blood 112: 5141- 5149). On the day of the experiment, the PBMCs were thawed in a water bath at 37°C.Rc\rdib h`_dph, ^jindnodib ja Gn^jq`^n Kj_dad`_ Bpg]`^^j^n K`_dph (GKBK, Ed]^j)supplemented with 100 units of penicillin, 100 g / mL streptomycin (P / S, Gibco) and 20% (v / v) Fetal Bovine Serum (FBS, Bodinco), was added and cells were rested in the dark for 20 min at RT. Next, cells thawing medium was removed and cells were washed by centrifuging for 5 min at 300 x g with the deacceleration rate set at 7. Cells were then resuspended in assay buffer, consisting of IMDM supplemented with 10% FBS and 1%N / Q. A`ggn r`m` ^jpio`_ \i_ \ qd\]dgdot ja ^ 70% r\n `inpm`_ ]t ^ji_p^odib \ omtk\i ]gp`staining using a LUNA-II^ automated brightfield cell counter (Logos Biosystems). Cell lines and cell culture Human embryonic kidney 293T (HEK293T), CHO-K1, Jeko-1, CCRF-CEM, Namalwa, Maver-1 and Z-138 were obtained from American Type Culture Collection (ATCC), and MEC-1, PGA-1, CII and L363 from DSMZ. RPCI-WM1 and TMD8 were kindly provided to Dr. M. Spaargaren by Dr. S.P. Treon and Dr. G. Lenz, respectively.FCI293R ^`ggn r`m` ^pgopm`_ di Bpg]`^^j^n Kj_dad`_ C\bg`^n K`_dph (BKCK, Ed]^j)supplemented with 10% FBS and 1% P / S. CHO-K1 cells were cultured in DMEM / F-12, supplemented with 10% FBS and 1% P / S. MEC-1, RPCI-WM1, TMD-8, PGA-1, L363, CCRF-CEM, Jeko-1, CII, Namalwa, Maver-1 and Z-138 cells were cultured in IMDM supplemented with 10% FBS and 1% P / S. One day prior to experiments, suspension cells were prepared at a concentration of 1 x 106cells / mL and adherent cells were maintained in culture under non-confluent conditions. On the day of the experiment, cells werem`^jpio`_ \i_ qd\]dgdot ja ^ 90% r\n `inpm`_ ]t ^ji_p^odib \ omtk\i ]gp` no\didib pndiba LUNA-II^ automated brightfield cell counter (Logos Biosystems). Antibody generation and production Previously described single-domain antibodies were cloned into the pMEK222 bacterial expression vector with C-terminal FLAG-6xHis tag (Van Hout et al., 2018. Biochem Pharmacol 158: 402-412; Godar et al., 2016. Sci Rep 6: 31621). VUN415- NanoLuciferase (VUN415-NanoLuc) and VUN416- NanoLuc were generated by subcloning VUN415 and VUN416 into a modified version of the pMEK222 vector, with a C-terminal upper Hinge linker^NanoLuciferase^6xHis tag (Ren et al., 2019. J Agric Food Chem 67: 5221-5229). BL21 Codon+ bacteria (Agilent) transformed with these pMEK222 plasmids were grown O / N in 10 mL of 2xYT medium, supplemented with glucose (2%) and ampicillin (1 ^g / mL). Next day, this O / N culture was inoculated (1:100) in Terrific Broth with ampicillin (1 ^g / mL). After bacteria grew at 37 °C to OD600 of 0.5, antibody production was initiated by adding isopropyl ^-D-1-thiogalactopyranoside (1 mM) and incubation took place for 4 h at 37 °C. After centrifugation, pellets were frozen O / N at -20 °C. After thawing and dissolving the pellet in phosphate-buffered saline (PBS, pH 7.4), periplasmic extracts were incubated head-over-head for 1.5 h at 4 °C. Antibodies were purified from the periplasm using immobilized affinity chromatography (IMAC) via 6x- His tags. Antibodies bound to ROTI®-Garose cobalt agarose beads (Carl Roth) were eluted with 150 mM imidazole (Sigma-Aldrich). Afterwards, the buffer of antibody eluates was exchanged for PBS by O / N dialysis using Snakeskin Dialysis Tubing (Thermo Fisher Scientific). Dialyzed fractions were combined and stored at -20 °C until experiments. Fluorescent labeling of single-domain antibodies The labeling of CXCR4 single-domain antibodies with ATTO565 fluorescent dyes (ATTO-TEC, #AD565-41, #AD565-31) using thiol-maleimide coupling and N-hydroxy- succinimide (NHS) chemistry was described previously (van den Bor et al., 2023. Cell Rep Methods 3: 100422). UV-VIS spectrometry was performed to ensure degree of labeling (DOL) > 0.5. Free dye of <5% was assessed by SDS-PAGE, followed by a fluorescence scan using an Odyssey imager (LI-COR, at suboptimal wavelength to prevent detector saturation) or Azure400 imager (Azure Biosystems, at 524 nm excitation). In a similar fashion, VUN415-Cys was conjugated using thiol-maleimide coupling with an excess of Alexa Fluor 647 C2-maleimide (Invitrogen, A20347), to ensure a DOL of 1. Transfection HEK293T cells HEK293T cells were transfected with a total of 17^g DNA and 67^g 257kDa linear polyethyleimine (PEI, Polysciences Inc.) in 1507mM NaCl solution per 1 × 106 cells. DNA encoding receptors and biosensors was, if necessary, supplemented with empty pcDEF3 to obtain a total DNA amount of 17^g. The DNA-PEI mixture was vortexed for 3 s and incubated for 157min at room temperature (RT). HEK293T cells were detached with Trypsin (Gibco) and resuspended in DMEM. The HEK293T cell suspension was added to the DNA-PEI mixture and cells were seeded at 3.5 × 104 per well in white flat-bottom 96- well plates (Greiner Bio-One). Receptor oligomerization CXCR4-Rluc and CXCR4-YFP For receptor oligomerization experiments using tagged receptors, HEK293T cells were transfected with 40 ng Myc-CXCR4-Rluc and 400 ng HA-CXCR4-YFP. After 48 hdi^p]\odji, ^`ggn r`m` r\nc`_ ji^` pndib N@Q \i_ h\dio\di`_ di F\if^n @paa`m`_ Q\gdi`Solution (HBSS), supplemented with 0.1% BSA, 1 mM MgCl2 and 2 mM CaCl2. Cells were stimulated with increasing concentrations of CXCL12, small molecules or single-domain antibodies for 15 min before BRET measurements. After incubating cells for 10 min with 5 ^M coelenterazine-h substrate (Promega), bioluminescence was measured at 535 / 307nm and 475 / 307nm using a PHERAstar plate reader (BMG). BRET signals were determined as the ratio of luminescence in the acceptor channel divided by the donor channel. The ligand-promoted BRET signal was calculated by dividing the pre-read-normalized BRET values of each ligand concentration by the BRET ratio obtained for the vehicle condition. Membrane extract preparation Two million HEK293T cells were plated in a 107cm² dish (Greiner Bio-One). The next day, cells were transfected with 250 ng NanoLuc-CXCR4, supplemented to a total of 57^g DNA with empty pcDEF3 vector, and 307^g PEI in 1507mM NaCl solution. The DNA- PEI mixture was vortexed for 37s and incubated for 157min at RT. Subsequently, the mixture was added dropwise to the adherent HEK293T cells. Protein expression was allowed to proceed for 487h. Media was then removed and cells were washed once with cold PBS. Next, cells were detached and resuspended in cold PBS. Cells were centrifuged at 15007×7g at 47°C, resuspended in cold PBS, and again centrifuged at 15007×7g at 47°C. The pellet was resuspended in membrane buffer (157mM Tris-Cl, 0.37mM EDTA, 27mM MgCl2, pH 7.5) and disrupted by the homogenizer Potter-Elvehjem at 12007rpm. Next, membranes were freeze-thawed using liquid nitrogen, pelleted by ultracentrifugation (25 min, 40000 7×7g, 4°C), carefully washed with Tris-Sucrose buffer (20 mM Tris, 250 mM Sucrose, pH = 7.4 at 4°C) and resuspended in Tris-Sucrose buffer. The membranes were homogenized using a 23G needle (10 strokes), aliquoted, snap-frozen using liquid nitrogen and protein concentrations were determined using a bicinchoninic acid assay (Pierce^ BCA Protein?nn\t; Rc`mhj Ddnc`m Q^d`iodad^). Qp]nlp`iogt, oc` h`h]m\i`n r`m` nojm`_ \o ^80~A piodguse in NanoBRET assays. Saturation binding of VUN415-ATTO565 Approximately 0.25 ^g per well of membrane extracts from HEK293T cells expressing NanoLuc-CXCR4 was added to a white flat-bottom 96-well plate. Subsequently, increasing concentrations of ATTO565-labeled VUN415 in HBSS, supplemented with 0.1% BSA, 1 mM MgCl2 and 2 mM CaCl2, were added. The plate was spun down and incubated for 1 h at RT. Next, 15 ^M furimazine substrate was added and luminescence was measured using a PHERAstar plate reader with 610 nm / LP and 460 / 80 nm filters until the luminescence signal stabilized. Displacement of fluorescent single-domain antibodies and CXCL12 Approximately 0.25 ^g per well of membrane extracts from NanoLuc-CXCR4- expressing HEK293T cells was added to a white flat-bottom 96-well plate. Subsequently, increasing concentrations of unlabeled ligands in HBSS, supplemented with 0.1% BSA, 1 mM MgCl2 and 2 mM CaCl2. The plate was spun down and incubated for 30 min at RT. Next, 316 pM antibody-ATTO565 or 10 nM CXCL12-AZ647 (Protein Foundry) was added and incubated for 1 h at RT. Next, 15 ^M furimazine substrate was added and luminescence was measured using a PHERAstar plate reader with 610 nm / LP and 460 / 80 nm filters until the luminescence signal stabilized. Flow cytometry for CXCR4 surface expression determination For each sample, 5 x 105cells were washed with ice-cold FACS buffer (0.5% BSA (PanReac AppliChem, A6588,0100) in PBS) and resuspended in ice-cold FACS buffer containing 3 µg / mL mouse anti-CXCR4 antibody 12G5 (Thermo Fisher Scientific, 35-8800) in polypropylene 5-mL tubes (Falcon). Following incubation on ice for 1 h, samples were washed three times with excess ice-cold FACS buffer to remove unbound antibody. Subsequently, samples were resuspended in ice-cold FACS buffer containing 2 µg / mL goat anti-mouse IgG (H+L) AlexaFluor^ 488 (Thermo Fisher Scientific, A-11001). After incubation and washing as described before, samples were resuspended in ice-cold FACS buffer. Subsequently, samples were analyzed utilizing an Attune Nxt Flow Cytometer (Thermo Fisher Scientific) at the AUMC Microscopy Cytometry Core Facility (MCCF), with flow rates not exceeding 500 µL / min. Sample analysis was conducted using FlowJo version 10 (BD Biosciences) to determine CXCR4 surface expression levels. Oligomer detection using antibody-based BRET in hematological malignancy cell lines 1 × 106 hematological malignancy cells were seeded in a white flat-bottom 96-well plate. In case of small molecule disruption, 10 ^M of IT1t or AMD070 was added to the cells. Subsequently, cells were stimulated with increasing equimolar concentrations of VUN415 / VUN416-NanoLuc and VUN415 / VUN416-ATTO565 or ITGB1-Nb-HL555 (QvQ) in assay buffer (HBSS, supplemented with 0.1% BSA, 1 mM MgCl2 and 2 mM CaCl2). For oligomer detection on PBMCs derived from CLL patients, 31.6 nM of VUN416-NanoLuc / - ATTO565 detection single-domain antibodies were added with a ATTO565:NanoLuc ratio of 0.25 (BRETmin) or 19 (BRETmax). After incubation for 2 h at RT, cells were washed twice with PBS and assay buffer was added. Subsequently, fluorescence of fluorescently labeled single-domain antibodies was measured using a CLARIOstar plate reader at 563 / 30 nm excitation and 592 / 30 nm emission. After addition of 15 ^M furimazine substrate, luminescence was measured using a PHERAstar plate reader with 610 nm / LP and 460 / 80 nm filters until the luminescence signal stabilized. Lentivirus production and transduction MEC-1 and RPCI-WM1 cell lines with inducible HA-CXCR4 expression and Namalwa and Z-138 cell lines with constitutive siRNA CXCR4 or scramble shRNA expression were generated by lentiviral transduction, as previously described (de Wit et al., 2016. Oncotarget 7: 67966-67985; Heukers et al., 2018. Oncogene 37: 4110^4121). Briefly, lentivirus was produced for 487h after co-transfecting four dishes of 2 x 106HEK293T cells with HA-CXCR4 pLenti6.3 / To / V5-DEST, pLKO.1 puro CXCR4 siRNA-1 / 2 or plKO.1 scramble shRNA together with pRSV-REV, pMDLg / pRRE and pMD2.g packaging vectors, using PEI as transfection reagent. Lentivirus solution from four dishes was pooled, cleared by centrifugation for 10 min at 500 x g and filter-sterilized. Subsequently, lentivirus was ultra-centrifuged for 1 h at 70.000 x g and supernatant was discarded until approximately 1 mL concentrated lentivirus solution was remaining. This lentivirus solution was then aliquoted and stored at -80 °C until lentiviral transduction. At the day of lentiviral transduction, 100 µL of concentrated lentivirus solution was added to 1 x 106cells in 1 mL. Subsequently, cells were incubated for three days before addition of the appropriate antibacterial selection agent. Knockdown efficiency and enhanced CXCR4 surface expression in the different cell lines was validated by determining CXCR4 surface expression levels as described before. CXCR4 expression in the doxycycline- inducible cell lines was induced using 1 µg / mL doxycycline (Sigma-Aldrich). dSTORM microscopy RPCI-WM1, Z-138 and CHO-K1 cells were fixated using 4% paraformaldehyde (PFA) in PBS for 15 min at 37 C. Next, cells were washed once and resuspended in FACS buffer (0.05% BSA in PBS). The fixated cells were then subjected to staining with VUN415-AlexaFluor647 at RT for 1 h. Unbound VUN415-AF647 was removed through a series of three consecutive washing steps using FACS buffer. 1 x 106cells were added to a poly-l-lysine (Sigma)-coated coverslip (VWR) in a 6-well plate (Greiner Bio-One). The coverslip was subjected to centrifugation in the 6-well plate at 500 x g for 15 min using a plate centrifuge (Eppendorf). Following this, the samples were stored in suspension in a dark environment at 4 °C until the time of readout. Before imaging, samples were mounted in oxygen scavenger-containing Glox-buffer to facilitate blinking conditions. Glox-buffer was prepared as described previously (28). Briefly, the following stock solutions were prepared and stored at -80 oC: 1M cysteamine (^-mercaptoethylamine, MEA) in 250 mM (Sigma, in 250 mM HCl), 70 mg / mL glucose- oxidase (Sigma-Aldrich) and 4 mg / mL catalase (Sigma-Aldrich). When mounting the sample, the final buffer was prepared freshly by diluting stock solutions MEA, glucose- oxidase plus catalase and glucose solution in 50 mM Tris pH 8.0 (final concentrations: 100 mM MEA, 7007^g / mL glucose oxidase, 407^g / mL catalase, 5% w / v glucose). To prevent oxygen from entering the sample during imaging, coverslips were mounted on cavity slides (Sigma-Aldrich) filled with imaging buffer. By removing surplus buffer from the sides of the coverslip a vacuum seal was created. Imaging was performed on a Ti-E microscope (Nikon) equipped with a 100x Apo TIRF oil immersion objective (NA. 1.49) and Perfect Focus System 3 (Nikon). A Lighthub- 6 laser combiner (Omicron) containing a 647 nm laser (LuxX 140 mW, Omicron) and a 405 nm diode laser (Power technology, 15 mW) together with optics allowing for a tunable angle of incidence were used for excitation. Illumination was adjusted for (pseudo-) total internal reflection fluorescence (TIRF) microscopy to remove out-of-focus signal. To separate emission light from excitation light, a quad-band polychroic mirror (ZT405 / 488 / 561 / 640rpc, Chroma) and a quad-band emission filter (ZET405 / 488 / 561 / 640m, Chroma) were used. Detection of the emission signal was done using a Hamamatsu Flash 4.0v2 sCMOS camera. Image stacks were acquired with a 30 ms exposure time, 50-100% laser power of 647 laser, 3-5% laser power of the 405 laser which was increased during imaging, and 5000 images per field of view. Components were controlled using MicroManager (Edelstein et al., 2014. J Biol Methods 1: e10). Acquired stacks were analyzed using v.1.2.1 of a custom ImageJ plugin called DoM (Detection of Molecules) (https: / / github.com / ekatrukha / DoM_Utrecht), as previously described (Chazeau et al., 2016. Methods Cell Biol 131: 127-149). Briefly, each image in an acquired stack was convoluted with a two-dimensional Mexican hat kernel whichh\o^c`n oc` hd^mjn^jk`^n kjdio nkm`\_ api^odji (NQD) ndu`. Rc` m`npgodib dio`indothistogram was utilized to create a thresholded mask that was used to calculate the centroids on the original image. These centroids were used as initial values to perform unweighted nonlinear least squares fitting with a Levenberg-Marquardt algorithm to an asymmetric two-dimensional Gaussian point spread function (PSF), allowing for the sub- pixel localization of particles. The acquired localization output by DoM was imported into the application ClusterViSu (https: / / github.com / andronovl / SharpViSu) that conducts astatistical cluster-\i\gtndn ]\n`_ ji Pdkg`t^n I-function as previously described (Andronovet al., 2016. Sci Rep 6: 24084). Eight areas per sample, ranging from 9-25 µm², were examined for the RPCI-WM1 and Z-138. Four areas per sample were examined for the CHO-K1 and non-specificity control samples (i.e. displacement by an excess of CXCR4 antagonist AMD3100). Selected areas did not overlap or came in contact with the edges ofoc` ^jmm`nkji_dib \i\gtu`_ ^`gg. Pdkg`t^n I-function was calculated for the indicated areasand during the analysis concentric circles were drawn around each localized point. The number of additional localizations enclosed within these circles were determined and statistically compared with a random distribution based on a similar surface area and number of localized points by conducting Monte-Carlo simulations. Cluster maps wereb`i`m\o`_ rdoc oc` ijmh\gdu`_ Pdkg`t^n I-function L(r), for r = 20 nm as the z-axis of everylocalization. Segmentation was conducted subsequently by thresholding the cluster map by setting a cut-off value for L(r) of 40-50 nm. Quantitative output, including cluster area, diameter and stoichiometry, were determined. Spatial-intensity Distribution Analysis (SpiDA) For spatial-intensity Distribution Analysis (SpiDA) analysis, 2.5 × 105 HEK293AD cells were grown on glass coverslips in six-well plates. Next day, cells were transfected with 600 ng of CXCR4-EYFP using Effectene transfection reagent (Qiagen) according tooc` h\ipa\^opm`m^n kmjoj^jg. Rc` i`so, oc` ^jq`mngdk r\n gj\_`_ dioj oc` ?oojagpjm dh\bdibchamber (Thermo Fisher Scientific). Prior to imaging, cells were stimulated for 30 min with 10 ^M IT1t, AMD0070 or TG-0054 in HBSS supplemented with 0.1% BSA. Imaging was performed using a commercial laser-scanning confocal microscope (Leica SP8) equipped with a 63× / 1.40 NA oil immersion objective, a white light laser (WLL), and photon-counting hybrid detectors. For excitation, 514 nm lines of the WLL were used, and for the detection of EYFP, emission bands of 520 nm to 600 nm were used. Images were acquired using 15% laser power. The image format was xy and image size was set to 512 × 512 pixels with 50-nm pixel size. For image analysis, the open-source custom-made code (https: / / github.com / PaoloAnnibale / MolecularBrightness) was loaded onto the Igor Pro software (WaveMetrics). Polygonal region of interest (ROI) selection was performed to avoid regions with non-homogenous fluorescence distribution (e.g. membrane raffles, clusters). Phosphoproteomics lysed in 8 M urea with 50 mM ammonium bicarbonate (pH 8, Sigma-Aldrich) with 1× Protease inhibitor cocktail EDTA (Roche) and 1× PhosSTOP (Roche). Sonication was performed with a Bioruptor (Diagenode) sonicator for 5 cycles (307s on, 307s off) at 47°C. Spun down for 1 h at 14,000 rpm at 16 °C to pellet cell debris and DNA. Protein concentration was determined by a microplate Bradford assay (Sigma- Aldrich).17mg aliquot of each sample was taken for further digestion and phosphopeptide enrichment. Protein samples were reduced in 10 mM dithiothreitol (DTT, Sigma-Aldrich) at 20 °C for 60 min, and alkylated in the dark with 20 mM iodoacetamide (IAA, Sigma-Aldrich) at 20 °C for 30 min. An additional final concentration of 10 mM DTT was added to quench the excess IAA. 50 mM ammonium bicarbonate was used to dilute to reach a final concentration of 2 M Urea. The alkylated proteins were sequentially digested using Lys- C (Wako) and trypsin (Sigma-Aldrich) at a 1:75 enzyme-to-protein ratio, and carried out at 37 °C. The Lys-C digestion lasted for 4 h, and the trypsin digestion was performed overnight. 3% formic acid was used to quench the digestion, and digested peptides were desalted by Sep-Pak C181 cc Vac cartridges (Waters), dried using a vacuum centrifuge, and stored at -80 °C for further use. Phosphopeptides were enriched by using Fe(III)-NTA 5 ^l (Agilent Technologies) in an automated AssayMAP Bravo Platform (Agilent Technologies). Fe(III)-NTA (nitrilotriacetic acid) cartridges were first primed with 250 ^L of priming buffer (99% acetonitrile (ACN), 0.1% TFA) at a flow rate of 100 ^L / min and equilibrated with 250 ^L of loading buffer (80% ACN, 0.1% TFA) at a flow rate of 50 ^L / min. Dried peptides were dissolved in 210 ^L of loading buffer and centrifuged at 14000 rpm for 10 min. Samples were then loaded at a flow rate of 3 ^L / min onto the cartridge, the flowthrough was collected into a separate plate. Cartridges were washed with 250 ^L of loading buffer at a flow rate of 20 ^L / min, and the phosphopeptides were eluted with 50 ^L of 10% ammonia at a flow rate of 5 ^L / min directly into 50 ^L of 10% formic acid. Flowthroughs and elutions were dried and injected directly on a liquid chromatography-coupled mass spectrometer. The phosphoproteome measurement was performed on an Orbitrap Exploris 480 mass spectrometer (Thermo Fisher Scientific) coupled with an UltiMate 3000 UHPLC system (Thermo Fisher Scientific) fitted with a µ-precolumn (C18 PepMap100, 5 µm, 100 Å, 5 mm × 300 µm; Thermo Fisher Scientific). Samples were analyzed in triplicates and separated on an analytical column (Poroshell 120 EC-C18, 2.7 µm, 50 cm × 75 µm, Agilent Technologies) with a 115-min gradient. Peptides were first eluted at a constant flow rate of 300 nl / min using 9 to 36% solvent B (0.1% v / v formic acid in 80% acetonitrile) over 97 min, raised to 99% in 3 min, then held for 3 min and equilibrated in 9% B for 1 min. The mass spectrometer was operated in data-dependent mode. Electrospray ionization was performed at a 2.1 kV static spray voltage; the temperature of the ion transfer tube was set to 275 °C, and the RF lens voltage was set to 55%. Full scan MS spectra from the m / z range of 375-1600 were acquired at a resolution of 60,000 after accumulating to the^Qo\i_\m_^ km`-set automated gain control (AGC) target. Higher energy collisiondissociation (HCD) was performed with 35% normalized collision energy (NCE), at an orbitrap resolution of 30,000. Dynamic exclusion time was set to 90 s and a 0.7 m / z isolation window was used for fragmentation. Data search was performed using MaxQuant (version 2.1.3.0) with an integrated Andromeda search engine, against the human Swissprot protein database (Downloaded on October 10th, 2022, containing 20,398 reviewed sequences). Digestion was defined as Trypsin / P and a maximum of 2 missed cleavages were allowed. Cysteine carbamidomethylation was set as a fixed modification. Protein N-terminal acetylation, methionine oxidation, and phosphorylation on serine, threonine, and tyrosine were set as variable modifications. Label-free quantification (LFQ) and the match-between-runs feature were enabled for protein quantification. A false discovery rate (FDR) of 1% was applied to both peptide spectrum matches (PSMs) and protein identification using a target-decoy approach. For total proteome measurements, intensity-based absolute quantification (iBAQ) was enabled. Quantitative data filtering was conducted using the Perseus software (version 1.6.14.0). Proteins that cross-matching to bovine contaminants were removed along with potential contaminants, reverse peptides, and proteins only identified by sites. LFQ intensities were log2-transformed. Proteins that were quantifiable in at least two out of three replicates were retained. Imputation was performed based on the normal distribution. Resazurin assays for venetoclax sensitization 3 × 104Z-138, Jeko-1 and Maver-1 cells with a viability of > 90% were seeded in serum-free IMDM in a black 96-well plate (Greiner Bio-One). For primary cultures, 3 × 104PBMCs of CLL patients were thawed and seeded in IMDM supplemented with 10% FBS in a black 96-well plate. After 1 h, cells were treated with increasing concentrations of venetoclax in the absence or presence of 10 ^M IT1t, AMD070, AMD3100, TG-0054 or VUN401. After 48 h incubation, 44 ^M resazurin was added into the culture medium. After 1 h incubation, fluorescence cytotoxicity read-out was performed using a CLARIOstar plate reader at 540 / 30 nm excitation and 590 / 30 nm emission. Spheroid assays of CLL patients were thawed, plated in in ultra-low absorption plates and centrifuged for 10 min at 1000 rpm and subsequently incubated for 24 h to allow spheroid formation. Three-dimensional (3D) cultures were cultured in IMDM supplemented with 10% FBS and 1% P / S and were stimulated and treated as indicated. Culture plates were then placed in an IncuCyte live-cell imager (Essen Biosciences) in an incubator at 37cC and 5% CO2. Scans were taken every 5 h using the single spheroid assay for live-cell analysis application and 4x magnification. Spheroid area was quantified using IncuCyte software as a proxy for spheroid growth. Corresponding step-by-step protocols were previously described (Haselager et al., 2023. Hemasphere 7: e938). After culture, spheroids were resuspended and disintegrated to ensure proper antibody staining. Cells were incubated with monoclonal antibodies for surface staining for 30 min at 4cC. Cells were stained with the following antibodies: CD5, CD19, CD4, CD8, CD25, CD95 and a viability dye. Samples were measured on a Canto II flow cytometer (BD Biosciences). Samples were analyzed using FlowJo software. Cytotoxicity assays for venetoclax sensitization 3 × 104Z-138, Jeko-1 and Maver-1 cells with a viability of > 90% were seeded in serum-free IMDM in a transparent flat 96-well plate (Greiner Bio-One). After 1 h, cells were treated with increasing concentrations of venetoclax in the absence or presence of 10 ^M IT1t, AMD070, AMD3100, TG-0054 or VUN401. After 48 h incubation, 100 nM MitoTracker Orange and 20 nM Topro-3 was added into the culture medium. Following staining, fluorescence cytotoxicity read-out was performed using a Attune NxT flow cytometer. Data analysis All and bar plots were visualized and statistical analyses were performed using Prism version 10.0 (GraphPad), unless indicated otherwise. Curves were fitted using least squares nonlinear regressions, assuming a sigmoidal fit (for concentration-response curves). Significance of differences was determined as indicated in figure legends. Antibody generation Two llamas were immunized with lipoparticles expressing human CXCR4 (Integral Molecular), followed by two rounds of panning on these lipoparticles using scFv / Fab phagemid libraries. Animal studies were conducted in accordance with European directive 2010 / 63 / EU and with national legislative regulations after local ethical approval by the Ethical Committee for Animal Testing. Total RNA was isolated from PBMCs, transcribed into cDNA (Superscript IV, Thermo Fischer Scientific) and VH and VL genes were amplified by PCR and cloned as scFv constructs into the pCB3 phagemid vector (van der Woning et al., 2016. mAbs 8: 1126-1135). Phages expressing scFv constructs were subjected to three rounds of panning on CXCR4 expressing CCRF-CEM, Jurkat and CXCR4 expressing CHO-K1 cells. Cells were washed and binding phages were eluted with 1 mg / ml trypsin in PBS for 30 minutes while shaking, which was then neutralized using 4 mg / ml 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride (AEBSF, Sigma- Aldrich). Eluted phages were rescued by infection into TG1 at OD600 of 0.5 for 30 minutes at 37 degrees, followed by addition of 2TY medium containing 100 µg / ml ampicillin and 2% glucose (w / v) and overnight incubation at 37°C while shaking. A total of 4 masterplates were picked with single clones from the outputs of the panning. These clones were expressed as scFv fragments and screened for CXCR4 binding as crude periplasmic extracts in flow cytometry on CHO-K1-CXCR4 cells and ELISA on CXCR4 expressing lipoparticles (Integral Molecular). Transwell migration studies Z-138 cells were serum-starved using IMDM supplemented with 0.5% BSA and subsequently put at a concentration of 5 x 106cells / mL. When applicable, 10 ^M of CXCR4 antagonists was added at this stage. Next, cells were placed in a head-over-head rotator and incubated for 2 h at RT. Following incubation, 5 x 105cells were gently transferred to a 5 µm pore sized polyester membrane tissue culture insert (Sarstedt). The insert was then positioned at the bottom of a 24-wells plate (Greiner Bio-One), containing 3.16 nM CXCL12 and incubated for 4 h. Following incubation, the tissue culture inserts were carefully removed from the 24-well plate. Subsequently, solutions were extracted from the wells and stained with 1 µg / mL propidium iodide (Thermo Fisher Scientific), while kept on ice. The amount of viable migrated cells in each well was measured using an Attune Nxt Flow Cytometer. Example 1: Disruption of non-endogenous CXCR4 oligomerization Basal CXCR4 oligomerization was previously demonstrated to be disrupted by the CXCR4 N-terminus-binding single-domain antibody VUN401 and transmembrane minor pocket-binding small molecule inverse agonists, including IT1t (Van Hout et al., 2018. Biochem Pharmacol 158: 402-412; Isbilir et al., 2020. Proc Natl Acad Sci U S A 117: 29144- 29154; Ward et al., 2021. J Biol Chem 296: 100139).To dissect the role of CXCR4 oligomerization in downstream signaling, we set out to expand the toolbox of CXCR4 ligands with distinct oligomerization-modulating properties. To this end, an oligomerization BRET assay using C-terminally YFP- and Rluc-tagged CXCR4 was employed to test a panel of CXCR4-targeting small molecules and previously developed single-domain antibodies (Figure 1A, B). As expected, close proximity between the two CXCR4 protomers caused a hyperbolic increase in BRET upon increasing CXCR4-YFP expression (Figure 1A). CXCL12 stimulation led to a further rise in the BRET ratio (Fig.1A, B), which has previously been explained as a conformational rearrangement within a pre-existing CXCR4 dimer (Percherancier et al., 2005. J Biol Chem 280: 9895- 9903; Paradis et al., 2022. Nat Commun 13: 6826; Perpina-Viciano et al., 2020. Mol Pharmacol 98: 72-87). In accordance with previous studies (Isbilir et al., 2020. Proc Natl Acad Sci U S A 117: 29144-29154; Ward et al., 2021. J Biol Chem 296: 100139), IT1t and VUN401 decreased CXCR4 oligomerization in a concentration-dependent manner (Fig. 1B). In contrast, major pocket-binding bicyclam AMD3100 and VUN400, a single-domain antibody that primarily interacts with ECL2 of CXCR4 (Van Hout et al., 2018. Biochem Pharmacol 158: 402-412) did not affect CXCR4 oligomerization BRET (Figure 1B). A bivalent construct of the cluster-disrupting single-domain antibody VUN401 prevented this disruption and did not affect the BRET signal (Fig.1B). In contrast to reducing the CXCR4 oligomerization state, CXCR4 oligomerization was induced by a bivalent construct of an otherwise oligomerization-neutral single-domain antibody (VUN400), called VUN400-Fc. VUN400-Fc increased the BRET ratio similarly to that of the endogenous agonist CXCL12. In a CXCR4 antagonist panel consisting of several small molecules and single- domain antibody VUN401, only the latter could disrupt receptor oligomerization without inverse agonistic effects on constitutive G^i / o activation (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154). We wondered whether this is a specific feature of VUN401 or whether this activity is correlated to CXCR4 oligomer disruption. To this end, we performed new immunizations and selected additional CXCR4-binding single-domain antibodies. Novel CXCR4-targeting single-domain antibodies (i.e., VUN410-VUN421) were then screened for their effects on CXCR4 oligomerization BRET (data not shown). Ofthese, the majority could ]` ^g\nndad`_ \n ^TSL400-gdf`^: ijo \aa`^odib ]\n\g AVAP4oligomerization BRET. In contrast, VUN411, VUN418, VUN419 and VUN421 were^TSL401-gdf`^: reducing the CXCR4 oligomerization BRET. VUN421 behaved similarly toVUN401, whereas the other three monomerizing single-domain antibodies were less efficacious in decreasing the CXCR4 oligomerization BRET signal, and VUN418 was also less potent (Fig. 1C). Next, the oligomer-disruptive effects of AMD070 (mavorixafor) and TG-0054 (burixafor) were assessed. These small molecules, targeting the minor pocket and an unknown region of CXCR4 (Mona et al., 2016. Org BiomolChem 14: 10298-10311; Rosenberg et al., 2019. Call Chem Biol 26: 662-6673), respectively, were not taken along in previous oligomerization studies. AMD070 showed similar oligomerization-disruptive effects as IT1t, albeit partially compared to VUN401 (Fig. 1D). For TG-0054, an increase rather than a decrease of oligomerization BRET signal was observed. Additionally, the effects of a conventional camelid antibody and its derived Fab fragment, 2D01, were determined. Like VUN401, the single-domain but not the bivalent format was capable of disrupting CXCR4 oligomers (Fig. 1E). To validate the decreased oligomerization BRET signals were truly due to decreased oligomeric complexity rather than conformational changes, Spatial-intensity Distribution Analysis (SpiDA) was performed with VUN411, VUN421, AMD070 and TG- 0054, using IT1t and VUN401 as controls. In accordance with the BRET data, AMD070 and IT1t reduced the amount of CXCR4 oligomers, to a smaller extent than VUN401, whereas TG-0054 did not have any effect (Fig.1F). Moreover, both VUN411 and VUN421 led to a significant reduction of the CXCR4 oligomeric state, albeit partially compared to VUN401. Hence, CXCR4-targeting ligands differentially modulate receptor oligomerization, via distinct underlying mechanisms. Example 2: CXCR4 oligomerization modulates JAK2 recruitment in a heterologous system Inhibition of basal signaling by minor pocket-binding small molecules caused disruption of CXCR4 oligomers, and therefore basal CXCR4 signaling was proposed to drive receptor oligomerization (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144- 29154). To investigate this hypothesis, we compared the ability of the various small molecules and VUN401 to induce receptor monomerization and inhibit basal G^i / o activity of a CXCR4 constitutively active mutant (CAM, N119S) in HEK293T cells. In accordance with previous results (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154; Ward et al., 2021. J Biol Chem 296: 100139; Zhang et al., 2002. J Biol Chem 277: 24515-14521; Mona et al., 2016. Org BiomolChem 14: 10298-10311; Rosenberg et al., 2019. Call Chem Biol 26: 662-6673), minor pocket binder and oligomer-disruptor IT1t inhibited basal G^i / o activation (data not shown). In contrast, and in line with the hypothesis, major pocket binder and non-oligomer-disrupting AMD3100 promoted rather than inhibited basal G^i / o signaling (data not shown). However, the other minor pocket binder AMD070 (Mona et al., 2016. Org BiomolChem 14: 10298-10311; Rosenberg et al., 2019. Call Chem Biol 26: 662- 6673) and VUN401 did disrupt CXCR4 oligomers but without affecting or promoting basal G^i / o activation, respectively (data not shown). TG-0054 showed an opposite profile by not disrupting oligomers and still inhibiting basal G^i / o activity. CXCR4 is known to activate the JAK / STAT pathway in various malignancies (Montresor et al., 2018. Oncotarget 9: 35123-35140; Liu et al., 2014. Oncol Rep 32: 2760- 2768; Pfeiffer et al., 2009. Br J Cancer 100: 1949-1956). Multiple studies reported constitutive interactions of CXCR4 with non-receptor tyrosine kinase JAK2, which CXCL12 could further enhance in some cellular backgrounds (Liu et al., 2014. Oncol Rep 32: 2760-2768; Pfeiffer et al., 2009. Br J Cancer 100: 1949-1956; Soriano et al., 2003. Eur J Immunol 33: 1328-1333; Ahr et al., 2005. J Biol Chem 280: 6692-6700). In turn, JAK2 can phosphorylate STAT3, causing dimerization and nuclear translocation of this transcription factor. Interestingly, CXCL12-induced activation of the JAK2 / STAT3 signaling cascade has been suggested to be G protein-independent but dependent on CXCR4 dimerization (Montresor et al., 2018. Oncotarget 9: 35123-35140; Vila-Coro et al., 1999. FASEB J 13: 1699-1710). To validate the hypothesized relationship between CXCR4 oligomerization and JAK2 / STAT3 signaling, we developed a BRET assay to study the proximity of CXCR4 and JAK2 (Fig.2A). CXCR4-NanoLuc and increasing amounts of mV- JAK2 led to a hyperbolic increase of BRET ratio, in accordance with previously reported basal interactions of these two proteins (Liu et al., 2014. Oncol Rep 32: 2760-2768; Pfeiffer et al., 2009. Br J Cancer 100: 1949-1956; Soriano et al., 2003. Eur J Immunol 33: 1328- 1333; Ahr et al., 2005. J Biol Chem 280: 6692-6700). Stimulation with endogenous agonist CXCL12 reduced the CXCR4-JAK2 BRET signal in a concentration-dependent manner (Fig. 2A, 2B). As expected, inhibition of G^i / o proteins with PTx did not prevent the CXCL12-induced effect (Fig. 2B). Next, we utilized the different CXCR4 oligomer modulators to probe the influence of receptor oligomerization on the interaction with JAK2. No effects were observed for AMD3100 and VUN400, ligands that did not influence CXCR4 oligomerization (Figure 4C). In contrast, oligomer-inducing ligands VUN400-Fc and VUN410-Fc decreased the CXCR4-JAK2 BRET signal, similar to CXCL12, whereas all monomerizing ligands enhanced the BRET ratio. An inverse linear correlation is apparent when plotting the BRETmax values of the CXCR4-JAK2 association for each ligand against the CXCR4 oligomerization BRETmax values (Fig.2D). Not only ligands but also receptor mutations affect the basal oligomeric state of CXCR4. For instance, SpiDA and BRET-based measurements demonstrated mutating a TMV residue in the dimer interface, W1955.34, disrupted CXCR4 dimerization (Ward et al., 2021. J Biol Chem 296: 100139, Paradis et al.¸2022. Nat Commun 13: 6826; Wu et al., 2010. Science 330: 1066- 1071). We examined whether this mutation had similar effects as monomerizing ligands on the association of CXCR4 with JAK2. Indeed, lower basal oligomerization BRET for the W195L mutant was accompanied by increased JAK2 association BRET compared to CXCR4 WT (data not shown). Together, these findings suggest that altering the oligomeric state of CXCR4 may influence non-canonical signaling events like the constitutive interaction of the receptor with JAK2. Example 3: Cluster sizes of endogenous CXCR4 receptors in hematological cancer cell lines. CXCR4 oligomerizes constitutively and expression-dependently in heterologous expression systems (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154; Ward et al., 2021. J Biol Chem 296: 100139). To investigate whether this also occurs in an endogenous setting, an antibody-based method for the detection of untagged GPCR oligomers in living cells was developed. For this purpose, it was required to have an antibody that 1) binds to CXCR4 with high affinity, 2) does not change the basal oligomeric state of the receptor and 3) has no effects on basal receptor signaling. Out of twelve single- domain antibodies tested, VUN415 displayed such properties and was selected as the lead candidate (data not shown). Next, an experimental bioluminescence resonance energy transfer (BRET) setup was developed with VUN415 genetically fused to NanoLuciferase (NanoLuc, Nluc) as donor and VUN415 conjugated with a fluorescent dye as acceptor (data not shown). In an adapted NanoB2approach (van den Bor et al., 2023. Cell Rep Methods 3: 100422), close- proximity of two or more clustering receptors should lead to BRET between antibody donor and acceptor constructs bound to different CXCR4 subunits, also termed protomers, thereby providing information about the relative receptor oligomeric state. Increasing equimolar (1:1) concentrations of the two antibody fusion constructs led to a concentration- dependent, saturable increase in antibody binding on CXCR4-overexpressing HEK293T cells (data not shown), as well as an increase in BRET ratio (data not shown). No increase in BRET was apparent for CXCR4-negative CHO-K1 or CRISPR Cas9 knockout HEK293T cells, in line with the lack of receptor expression and thus absence of antibody binding (data not shown). Similar results were obtained using a fixed antibody concentration with varying ratios of the two fusion constructs, although in this case a hyperbolic saturable increase in oligomerization BRET signal was detected for CXCR4-overexpressing HEK293T cells (data not shown). The increased BRET ratio was not due to lower concentrations of the NanoLuc-tagged antibody, since exchanging VUN415-ATTO565 for unlabeled VUN415 yielded no increase in BRET signal (data not shown). Dimerization and higher-order oligomerization of proteins can be artificially induced by fusing them to a FK506-binding protein (FKBP) domain, which can be chemically crosslinked by AP20187 and other ligands (Hofman et al., 2010. J Biol Chem 285: 39481-39489). To validate whether the observed increased BRET signals are indeed a consequence of receptor clustering, CXCR4 was fused C-terminally to a FKBP dimerization domain to force clustering. As expected (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154; Ward et al., 2021. J Biol Chem 296: 100139; Ge et al.¸2017. Sci Rep 7: 16873), low receptor expression levels were associated with the predominant monomeric CXCR4, while a robust increase in antibody oligomerization BRET was observed upon stimulation with AP20187 for FKBP-tagged CXCR4, but not for the untagged receptor (data not shown). To test previously reported expression-dependency of CXCR4 oligomerization (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154; Ward et al., 2021. J Biol Chem 296: 100139), an ecdysone-inducible CXCR4 construct was employed (Lee et al., 2016. Mol Ther Nucleic Acids 5: e367). As anticipated, stimulation with ecdysone agonist tebufenozide led to concentration-dependent increases in CXCR4 expression as well as antibody oligomerization BRET signal (data not shown). After the development and initial validation in HEK293T cells, the ability of the adapted NanoB2approach to detect endogenous CXCR4 oligomers in hematological cancer cells was assessed. The focus on these malignancies was because they display increased levels of CXCR4 which are considered to play a prominent pathological role (Mohle et al., 1999. Leukemia 13: 1954-1959; Moreno et al., 2015. J Pathol 235: 445-455; Wester et al., 2015. Theranostics 5: 618-630). Using the Namalwa Burkitt lymphoma cell line as a proof- of-concept, it was set out to detect endogenous CXCR4 oligomers. A robust increase in BRET signal was apparent using VUN415-NanoLuc with VUN415-ATTO565. In contrast, no BRET occurred in a control setup using an antibody against the highly expressed ITGB1 receptor as a BRET partner (data not shown). The lack of BRET for the ITGB1 control was not due to a lack of antibody binding, as both fluorescent single-domain antibodies demonstrated clear concentration-dependent binding to Namalwa cells, resulting in comparable fluorescent intensities (data not shown). Hence, the NanoB2approach can monitor specific oligomers of heterologously and endogenously expressed receptors. Subsequently, a large and diverse panel of hematological cancer cell lines was employed to assess whether enhanced endogenous expression of CXCR4 is associated with an increased oligomeric state. First, full concentration-response curves for antibody-based oligomerization detection were generated using a small selection of cell lines with varying CXCR4 expression levels and disease subtypes (Fig.3A). For the CLL cell line MEC-1 with non-detectable surface expression of CXCR4, no oligomerization BRET signal could be detected. For the other two cell lines, clear concentration-dependent increases in oligomerization BRET signal were apparent, with similar BRET50 but much higher BRETmax values for the CXCR4highcell line Z-138 (mantle-cell lymphoma, MCL) than for RPC1-WM1 cells (Waldenstrom macroglobulinemia) with low CXCR4 expression. The normalized oligomerization BRETmax values, which were not affected by plate reader gain settings (data not shown), were used as a measure of relative oligomerization level. In a large panel with hematological cancer cell lines, Z-138 cells stood out as the cell line with the highest level of CXCR4 present in oligomers (Fig. 2B). Although other factors such as lipid composition are likely involved as well (Pluhackova et al., 2016. PLoS Comput Biol 12: e1005169; Gardeta et al., 2022. Front Immunol 13: 925559), endogenous oligomeric complexity of CXCR4 generally followed the trend of receptor expression levels, with CII and Z-138 as outliers (Fig. 2B). To further investigate the link between receptor expression and oligomerization, the effects of genetic manipulation of CXCR4 expression levels were tested. Doxycycline- inducible expression of CXCR4 enhanced the oligomeric state in RPC1-WM1 (data not shown) and MEC-1 cells, whereas siRNA-mediated silencing of CXCR4 caused a marked reduction of endogenous receptor oligomerization in Z-138 (Fig. 3C) and Namalwa cells (data not shown). To validate the BRET-based findings of endogenous CXCR4 oligomerization in hematological cancer cell lines under basal conditions, dSTORM single-molecule imaging^jh]di`_ rdoc Pdkg`t^n I-function analysis (Kiskowski et al., 2009. Biophys J 97: 1095-1103) was employed. Using Alexa647-conjugated VUN415, CXCR4 was labeled in RPCI- WM1 (CXCR4low) and Z-138 (CXCR4high) cells. In contrast to RPCI-WM1 and Z-138 cells, negative control CHO-K1 (CXCR4negative) cells displayed negligible amounts of localized events per cell (Fig.3D). To further assess the specificity of VUN415-AF647, samples were incubated with an excess of CXCR4 antagonist AMD3100. Both RPCI-WM1 and Z-138 contained specific localized events as demonstrated by the elevation of their number compared to the corresponding AMD3100-treated sample (data not shown). For RPCI- WM1 cells, a higher proportion of non-specific events was observed (128 ± 31 non-specific events per µm²) compared to Z-138 (61 ± 17 non-specific events per µm²). Next, a statistical cluster-\i\gtndn ]\n`_ ji Pdkg`t^n I-function and localizationoutput was performed to analyze differences between RPCI-WM1 and Z-138 cells. Thenk\od\g _dnomd]podji k\oo`mi ja oc` gj^\gdu`_ `q`ion r\n kgjoo`_ pndib Pdkg`t^n F api^odji(linear-om\inajmh`_ Pdkg`t^n I), rcd^c ^jhk\m`n nk\od\g _dnomd]podji k\oo`min oj m\i_jhdistributions generated by Monte-Carlo simulations (data not shown). Both RPCI-WM1 and Z-138 showed significant clustering of CXCR4 receptors (positive L(r)-r value) (Fig. 2D). In contrast, no clustering was observed for CHO-K1 cells (data not shown). The average CXCR4 cluster diameter in RPCI-WM1 CXCR4lowcells (48 ± 34 nm) was found to be significantly larger than in Z-138 CXCR4highcells (36 ± 23 nm, Fig.3E). When assessing cluster stoichiometry, Z-138 appeared to contain fewer small clusters (up to 5 receptors / cluster) and more large-size clusters (>5 receptors / clusters) than RPCI-WM1 (Fig.3F). Taken together, these findings suggest that native CXCR4 constitutively organizes into multimeric structures in non-Hodgkin lymphoma and multiple myeloma cell lines, which is linked to receptor expression levels. Example 4: Pharmacological disruption of endogenous CXCR4 clusters by small molecules After detecting endogenous CXCR4 oligomers in hematological cancer cells, it was examined whether these native oligomers can be pharmacologically disrupted. It was first attempted to identify ligands that can bind CXCR4 alongside the oligomer detection single-domain antibodies (i.e. labeled VUN415). For this, binding of ATTO565-labeled single-domain antibodies to NanoLuc-tagged CXCR4 (NanoB2-based approach, van den Bor et al., 2023. Cell Rep Methods 3: 100422) was assessed in the absence and presence of different small molecules and VUN401. AMD070 (AMD11070, mavorixafor) did not affect the binding of VUN415 to CXCR4 (Fig. 4A). In contrast, AMD3100 (plerixafor), TG-0054 (burixafor) and VUN401 completely prevented binding of VUN415 and IT1t appeared to stabilize the binding of the antibody to the receptor. As IT1t was previously shown to be amongst the most efficacious disruptors of recombinantly overexpressed CXCR4 oligomers (Isbilir et al., 2020. Proc Natl Acad Sci USA 117: 29144-29154), a screening was performed to identify an antibody that is non-competitive with this ligand. While IT1t prevented CXCR4 binding for most single-domain antibodies (data not shown), including VUN400, binding of VUN401 was partially inhibited and VUN417 binding was stabilized (Fig. 4B). Importantly, VUN416 binding was unaffected by IT1t (Fig. 4B). Further screening of a large panel of antibodies and different small molecules indicated that VUN415 / AMD070 and VUN416 / IT1t pairs were the only non-competitive antibody-small molecule pairs (data not shown). Then, the pharmacological disruption of endogenous CXCR4 oligomers in Z-138 cells, the hematological cancer cell line with the highest CXCR4 oligomeric complexity (Fig. 3B), was assessed. As detected by VUN415-NanoLuc / -ATTO565-based BRET, AMD070 indeed caused a partial decrease in CXCR4 oligomerization in Z-138 cells (Fig. 4C). VUN416-NanoLuc / -ATTO565-based detection revealed IT1t fully monomerized endogenous CXCR4 in Z-138 cells. Importantly, also in these cells, the binding of both VUN415-NanoLuc and VUN416-NanoLuc was unaffected by their respective small molecules (data not shown). Taken together, the identification of non-competitive antibody-small molecule pairs allowed for the detection of differential endogenous CXCR4 oligomer disruption by small molecules. Example 5: CXCR4 clusters drive basal mobility of hematological cancer cell lines CXCL12-induced CXCR4 nanoclustering has been linked to CXCR4 functionality, including adhesion and migration responses (Martínez-Muñoz et al., 2018. Mol Cell 70: 106-119). The potential role of CXCR4 oligomerization in basal migration of Z-138 cells with high expression levels and oligomeric state of CXCR4 was investigated using a transwell migration assay. To this end, cells were treated with previously characterized CXCR4 cluster disrupters IT1t and VUN401 or 2D01 Fab. The bivalent 2D01-Fc, the small molecule AMD3100, and the single-domain antibody VUN415, which all do not influence CXCR4 cluster status, were included as negative controls. These data demonstrated that molecules that reduced CXCR4 cluster size specifically impaired the basal migration of Z- 138 cells (Fig.5). Example 6: Disruption of CXCR4 clusters inhibits primary CLL spheroid growth The functional role of CXCR4 clusters in patient-derived primary CLL and MCL cells was investigated after oligomerization of native CXCR4 could be detected in a panel of primary cells from five CLL patients and two MCL patients, which could almost be completely disrupted by IT1t (Fig. 4). In a 3D lymph node-mimicking CLL model derived of patient peripheral blood cells (Haselager et al., 2023. Hemasphere 7: e938), CXCR4 cluster disrupters IT1t and AMD070 inhibited spheroid growth (Fig. 6A) without having cytotoxic effects (i.e. no significant difference in cell viability in primary CLL spheroid model was observed for CXCR4 cluster disrupters IT1t and AMD070 compared to vehicle, Fig.6B). In contrast, non-monomerizing antagonists AMD3100 and TG-0054 did not affect spheroid growth (Fig. 6A). Compared to AMD070, IT1t inhibited spheroid growth more potently and additionally inhibited the expression of CLL and CD4 activation markers CD95 and CD25, respectively (data not shown). These data indicate that CXCR4 clustering extends to CLL patient-derived cultures and selective pharmacological disruption thereof has therapeutic potential. Example 7: CXCR4 oligomers drive anti-apoptotic signaling To investigate whether CXCR4 oligomers serve as signaling platforms, a phospho- proteome analysis in absence and presence of cluster-disrupting agents was conducted (Fig. 7A). Z-138 cells, which have the highest CXCR4 cluster size, were treated with monomerizing small molecule IT1t and monomerizing antibody VUN401. Quality control results gave confidence to proceed with phosphor-proteome analysis (data not shown). To explore changes in phosphorylation pattern upon sustained exposure to CXCR4- monomerizering ligands, we performed Gene Ontology (GO) functional enrichment analysis using the downregulated phospho-sites after treatment with IT1t or VUN401 for 60 min. (Fig. 7B). For IT1t and VUN401, 12 and 6 significantly differentially regulated GO biological processes were identified, respectively. Of these, 5 were overlapping, which might indicate that these processes are related to CXCR4 oligomerization. Among the overlapping processes, we found the positive regulation of apoptosis (Fig. 7B). STRING networking analysis revealed medium confidence interaction (interaction score > 0.400) for 60 phospho-proteins in the IT1t-VUN401 overlapping apoptosis sensitization network (Fig. 7C). In total, 110 phospho-sites were present in the network, of which 18 have been functionally annotated for regulating apoptosis. For instance, BAD phosphosites S75 and S118, whose phosphorylation in response to survival stimuli blocks the activity of this pro- apopotic BCL-2 family protein (Zhou et al., 2000. J Biol Chem 275: 25046-25051; Tan et al., 2000. J Biol Chem 275: 25865-15969), are both downregulated in response to IT1t and VUN401 treatment of Z-138 cells. Similarly, IT1t and VUN401 both reduced the phosphorylation of the inhibitory S157 residue of pro-apoptotic caspase-2 (Pediani et al., 2016. J Biol Chem 291: 13132-13146). Example 8: Disruption of CXCR4 oligomerization sensitizes to chemotherapy- induced cell death The BH3 mimetic venetoclax (ABT-199) is a potent, selective BCL-2 inhibitor that is approved for CLL and specific AML patients (Vereertbrugghen et al., 2021. Front Oncol 11: 598319; Souers et al., 2013. Nat Med 19: 202-208). However, chronic venetoclax exposure can lead to acquired resistance in B-cell lymphoma cell lines (Choudhary et al., 2015. Cell Death Dis 6: e1593), highlighting the need of combinatorial therapies. Given the central role of the intrinsic apoptosis pathway in the CXCR4 oligomerization- associated signaling profile, it was tested whether CXCR4 cluster disrupting ligands sensitized Z-138, Jeko-1 and Maver-1 MCL cells to venetoclax-induced cell death in a resazurin assay. Co-treatment with cluster disrupting ligands IT1t and to a smaller extent VUN401 enhanced the sensitivity of Z-138 cells for venetoclax-induced drop in metabolic activity. In contrast, this effect was not apparent for non-monomerizing ligand AMD3100 (Fig. 8A, Table 3 and Table 4). The effect of IT1t appeared to be CXCR4-specific, as the addition of a high concentration AMD3100 blocked the sensitization effect (data not shown). To further explore the role of CXCR4 oligomerization in this sensitization phenotype, we tested more monomerizing and non-monomerizing small molecules. To validate that the drop in metabolic activity is cell death, a FACS readout was performed making use of viability dyes that allowed distinction between early- and late apoptosis. Not only IT1t but also the other monomerizing ligand, AMD070, significantly enhanced the cytotoxic effect of venetoclax in Z-138, Jeko-1 and Maver-1 cells, whereas no effect for non-cluster-disrupting small molecules AMD3100 and TG-0054 was observed (Fig.8B). A double titration with IT1t and venetoclax in Z-138 cells in the resazurin assay format revealed IT1t to dose-dependently sensitize venetoclax-mediated cytotoxicity with saturating effects at 10 µM (Fig. 8C). Synergy assessment using the Bliss independence model with earlier obtained cytotoxicity data (Fig. 8B) revealed strong synergy for IT1t and venetoclax, while VUN401 displayed synergy to a lower extent (Fig.8D). Coincubation with the pan-caspase inhibitor qVD revealed the synergy to be at least in part caspase dependent (Fig. 8E). Importantly, similar to what was observed with the MCL cell lines, cluster disrupter IT1t, but not the non-disrupter AMD3100, potentiated the cytotoxic effects of venetoclax in primary CLL cultures (Fig. 8F, 8G). These data indicate that CXCR4 clustering promotes anti-apoptotic signaling and associated phenotypes in hematological cancer cell lines and primary CLL cultures, which can be targeted using CXCR4-monomerizing ligands. Example 9: Cell viability assay for venetoclax sensitization in DMS 79 cells. DMS 79 cells (ATCC; CRL-2049) were diluted in the recommended medium by thesupplier and dispensed in a 384-well plate at a density of 100 - 6400 ^`ggn k`m r`gg di 45 ^gmedium. The margins of the plate were filled with phosphate-buffered saline. Plated cellsr`m` di^p]\o`_ di \ cphd_dad`_ \ohjnkc`m` ja 5 % AM2 \o 37 ^A. ?ao`m 24 cjpmn, 5 ^g ja^jhkjpi_ _dgpodji r\n \__`_ \i_ kg\o`n r`m` apmoc`m di^p]\o`_. ?o o=`i_, 24 ^g jaATPlite 1Step^ (PerkinElmer) solution was added to each well, and subsequently shaken for 2 minutes. After 10 minutes of incubation in the dark, the luminescence was recorded on an Envision multimode reader (PerkinElmer). Sensitization to venetoclax-induced cell death by CXCR4-oligomer disruptor IT1t was not only observed in hematological cancer cell lines and primary CLL cultures, but also in the lung small cell cancer cell line DMS 79 (Fig. 9). Example 10. Constitutive cell migration assessment. A live-cell imaging experiment was performed, where a consistent proportion of the control-treated cells (5-10%) showed significant basal constitutive cell migration during a 4-hour period (Fig. 10). Monomerizing ligands IT1t and VUN401 impaired the basal cell migration of this population significantly, whereas non-monomerizing ligands AMD3100 and VUN415 did not (Fig. 10A). When analyzing the trajectories, both IT1t and VUN401 impaired the average migration speed of the highly migratory Z-138 cell population significantly (Fig. 10B). IT1t showed a significant inhibition of the average traveled distance, whereas VUN401 showed a similar trend (Fig. 10C). Collectively, these results highlight a role for CXCR4 oligomers in constitutive cell migration, which can be modulated by oligomer disruptors.

[0002] Table 3: Effects CXCR4 ligands on venetoclax potency in different MCL cell lines in a resazurin assay format. Data are pooled mean ± SEM of at least three independent experiments, each performed in triplicate. Significant differences (P < 0.05) compared to vehicle condition are indicated in bold, according to a one-way ANOVA followed byBpii`o^n kjno cj^ o`no. **** P < 0.0001.Z-138 Maver-1Rm`\oh`io pGA50 N pGA50 NT`cd^g` 9.3 ^ 0.1 L.?. 9.6 ^ 0.1 L.?.+?KB3100 9.3 ^ 0.1 0.99 9.6 ^ 0.1 0.93+GR1o 10.1 ± 0.1 **** 10.0 ± 0.1 0.01+TSL401 9.7 ± 0.2 0.01 L.B. L.B.Jeko-1 Rm`\oh`io pGA50 NT`cd^g` 8.2 ^ 0.1 L.?.+?KB3100 8.1 ^ 0.3 0.95+GR1o 8.9 ± 0.1 0.05+TSL401 L.B. L.B.Table 4: Effects CXCR4 ligands on venetoclax potency in different MCL cell lines in a FACS viability assay format. Data are pooled mean ± SEM of at least three independent experiments. Significant differences (P < 0.05) compared to vehicle condition are indicated in bold, according to a paired T-test. * P < 0.05, ** P < 0.001. Z-138 Maver-1 Jeko-1 Treatment pEC50 P pEC50 P pEC50 PVehicle 8.5 ± 0.6 N.A. 9.3 ± 0.3 N.A. 8.8 ± 0.6 N.A.+IT1t 9.7 ± 0.4 * 9.6 ± 0.3 0.1 9.8 ± 0.5 *+AMD070 8.8 ± 0.7 ** 9.5 ± 0.5 0.2 9.0 ± 0.7 0.3+TG-0054 8.3 ± 1 0.3 9.2 ± 0.3 0.1 8.9 ± 0.4 0.8+AMD3100 8.3 ± 0.9 0.3 9.2 ± 0.4 0.3 8.9 ± 0.5 0.4+VUN401 8.8 ± 0.3 0.1 N.D. N.D. N.D. N.D.

Claims

CLAIMS 1. A CXCR4 cluster inhibitor, for use in a method of treating a tumor comprising CXCR4 clusters in a subject, the method further comprising simultaneous or sequential administering a cell death-inducing compound to the subject, whereby the CXCR4 cluster inhibitor is IT1t, VUN401, 2D01, VUN411, VUN418, VUN419, VUN421, or any combination thereof.

2. The CXCR4 cluster inhibitor for use of claim 1, wherein the CXCR4 cluster inhibitor is IT1t or VUN401.

3. The CXCR4 cluster inhibitor for use of claim 1 or 2, wherein the CXCR4 cluster inhibitor is IT1t.

4. The CXCR4 cluster inhibitor for use of any one of claims 1-3, wherein the CXCR4 cluster inhibitor is an antibody conjugated to IT1t.

5. The CXCR4 cluster inhibitor for use of any one of claims 1-4, wherein the cell death-inducing compound is a TNF-related ligand and / or a BCL-2 inhibitor.

6. The CXCR4 cluster inhibitor for use of any one of claims 1-5, wherein the cell death-inducing compound is venetoclax.

7. The CXCR4 cluster inhibitor of any one of claims 1-6, wherein the CXCR4 cluster inhibitor is administered before the cell death-inducing compound.

8. A composition comprising a CXCR4 cluster inhibitor and a cell death-inducing compound, preferably a pharmaceutical composition, wherein the CXCR4 cluster inhibitor is selected from IT1t, 2D01, VUN401, VUN411, VUN418, VUN419, VUN421, or a combination thereof.

9. The composition of claim 8, wherein the CXCR4 cluster inhibitor is IT1t or VUN401.

10. The composition of claim 8, wherein the CXCR4 cluster inhibitor is an antibody fragment conjugated to IT1t.

11. The composition of any one of claims 8-10, wherein the cell death-inducing compound is a TNF-related ligand and / or a BCL-2 inhibitor, preferably venetoclax.

12. A kit of parts, comprising a CXCR4 cluster inhibitor, a cell death-inducing compound, and instructions to provide an individual having a tumor comprising CXCR4 clustering with the CXCR4 cluster inhibitor and the cell death-inducing compound, wherein the CXCR4 cluster inhibitor is IT1t, VUN401, 2D01, VUN411, VUN418, VUN419, VUN421, or any combination thereof.

13. A method of treating a tumor comprising CXCR4 clusters in a subject, comprising simultaneous or sequential administering a CXCR4 cluster inhibitor and a cell death- inducing compound to the subject, whereby the CXCR4 cluster inhibitor is IT1t, VUN401, 2D01, VUN411, VUN418, VUN419, VUN421, or any combination thereof.

14. The method of claim 13, wherein the tumor is a lymphoma, myeloma, leukemia, glioma, rhabdomyosarcoma or prostate cancer.

15. The method of claim 13 or claim 14, wherein the tumor is a non-Hodgkin lymphoma.

16. A method of reducing basal motility of a cell by administering a CXCR4 cluster inhibitor, whereby the CXCR4 cluster inhibitor is IT1t, VUN401, 2D01, VUN411, VUN418, VUN419, VUN421, or a combination thereof.

17. The method of claim 16, whereby the CXCR4 cluster inhibitor is IT1t and / or VUN401.

Images