Methods of determining cellular chemosensitivity

METHODS OF DETERMINING CELLULAR CHEMOSENSITIVITY

FIELD OF THE INVENTION

0003The invention relates to generally methods of determining cellular chemosensitivity by determining the pattern of sensitively of a cell to a panel of BH3 domain peptides.

BACKGROUND OF THE INVENTION

0005Programmed cell death, referred to as apoptosis, plays an indispensable role in the development and maintenance of tissue homeostasis within all multicellular organisms (Raff, Nature 356: 397-400, 1992). Genetic and molecular analysis from nematodes to humans has indicated that the apoptotic pathway of cellular suicide is highly conserved (Hengartner and Horvitz, Cell 76: 1107-1 1 14, 1994). In addition to being essential for normal development and maintenance, apoptosis is important in the defense against viral infection and in preventing the emergence of cancer.

0006Diverse intrinsic death signals emanating from multiple subcellular locales all induce the release of cytochrome c from mitochondria to activate Apaf-1 and result in effector caspase activation. Proteins in the BCL-2 family are major regulators of the commitment to programmed cell death as well as executioners of death signals at the mitochondrion. Members of this family include both pro- and anti-apoptotic proteins and share homology in up to four conserved regions termed BCL-2 homology (BH) 1-4 domains (Adams and Cory, 1998). The family can be divided into three main sub-classes. The anti-apoptotic proteins, which include BCL-2 and BCL-X_L, are all "multidomain," sharing homology throughout all four BH domains. However, the pro- apoptotic proteins can be further subdivided and include multidomain proteins, such as BAX and BAK, which possess sequence homology in BH1-3 domains. The more distantly related "BH3- only" proteins are to date all pro-apoptotic and share sequence homology within the amphipathic α-helical BH3 region, which is required for their apoptotic function (Chittenden et al., 1995; O'Connor et al., 1998; Wang et al., 1996; Zha et al., 1997).

0007Multidomain pro-apoptotic proteins such as BAX and BAK upon receipt of death signals participate in executing mitochondrial dysfunction. In viable cells, these proteins exist as monomers. In response to a variety of death stimuli, however, inactive BAX, which is located in the cytosol or loosely attached to membranes, inserts deeply into the outer mitochondrial membrane as a homo-oligomerized multimer (Eskes et al., 2000; Gross et al., 1998; Wolter et al., 1997 ). Inactive BAK resides at the mitochondrion where it also undergoes an allosteric conformational change in response to death signals, which includes homo-oligomerization (Griffiths et al., 1999; Wei et al., 2000). Cells deficient in both BAX and BAK are resistant to a wide variety of death stimuli that emanate from multiple locations within the cell (Wei et al., 2001).

0008The BH3-only molecules constitute the third subset of this family and include BID, NOXA, PUMA, BIK, BlM and BAD (Kelekar and Thompson, 1998). These proteins share sequence homology only in the amphipathic oc-helical BH3 region which mutation analysis indicated is required in pro-apoptotic members for their death activity. Moreover, the BH3-only proteins require this domain to demonstrate binding to "multidomain" BCL-2 family members. Multiple binding assays, including yeast two-hybrid, co-immunoprecipitation from detergent solubilized cell lysates and in-vitro pull down experiments indicate that individual BH3-only molecules display some selectivity for multidomain BCL-2 members (Boyd et al., 1995; O'Connor et al., 1998; Oda et al., 2000; Wang et al., 1996; Yang et ah, 1995). The BID protein binds pro-apoptotic BAX and BAK as well as anti-apoptotic BCL-2 and BCL-X_L (Wang et al., 1996; Wei et al., 2000). In contrast, BAD, and NOXA as intact molecules display preferential binding to anti-apoptotic members (Boyd et al., 1995; O'Connor et al., 1998; Oda et al., 2000; Yang et al., 1995)

0009SUMMARY OF THE INVENTION It various aspects, the invention provides methods of predicting sensitivity of a cell to a therapeutic agent by contacting the cell or a cellular component (e.g., mitochondria) thereof with a BH3 domain peptide and detecting apoptosis. The presence of apoptosis indicates that the cell is sensitive to the therapeutic agent. Alternatively, sensitivity of a cell to a therapeutic agent is determined by providing a BH3 profile of the cancer cell and comparing the BH3 profile to a control profile. A similarity of the BH3 profile in the cancer cell compared to the control profile indicates the cancer cell is sensitive to the therapeutic agent.

0010Also provided are methods of selecting an agent that is therapeutic for a subject by providing a cancer cell or cellular component thereof, contacting the cell or cellular component with a BH3 domain peptide or mimetic thereof and determining whether or not the BH3 domain peptide or mimetic induces apoptosis in the cancer cell or cellular component thereof to produce a test BH3 profile. The test BH3 profile is compared with a therapeutic agent BH3 profile. A similarity of the test BH3 profile compared to the therapeutic agent BH3 profile indicates that the agent is therapeutic for the subject.

0011Apoptosis is detected for example by detecting cytochrome C release from mitochondria. The therapeutic agent is a chemotherapeutic agent a BH3 domain mimetic, or antagonist of an anti-apoptotic protein. The BH3 domain peptide is derived from the BH3 domain of a BID, a BIM, a BAD, a BIK, a NOXA, a PUMA a BMF, or a HRK polypeptide. Exemplary BH3 domain peptides include SEQ ID NO: 1- 14 and 15. The BH3 domain peptide is an activator or a sensitizer of apoptosis. Preferably, the BH3 domain peptide is a sensitizer.

0012A profile containing a pattern of mitochondrial sensitivity to BH3 peptides taken from one or more subjects who have cancer is also provided by the invention.

0013Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting. Other features and advantages of the invention will be apparent from the following detailed description, and from the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

0015Figure IA is a bar chart showing the effects of sensitizer BH3 peptides on cytochrome c release from mouse liver mitochondria. Peptide concentrations were 10 uM unless otherwise indicated; tBID concentration was 13 nM. Average and standard deviation from at least three independent assays performed for each antiapoptotic protein are shown.

0016Figure IB is a bar chart showing the effects of tBID (first bar) and BCL-2 (1.2 uM, second bar) on cytochrome c release from mitochondria. The effect of sensitzer BH3 peptides on restoration of cytochrome c release is also shown. Note that in each case, restoration of cytochrome c release corresponds to a high affinity interaction in Table Ib.

0017Figure 1C is a bar chart showing the effects of tBID (first bar) and BCL-XL (0.25 uM, second bar) on cytochrome c release from mitochondria. The effect of sensitizer BH3 peptides on restoration of cytochrome c release is also shown.

0018Figure ID is a bar chart showing the effects of tBID (first bar; note tBID concentration was 43 nM in this experiment) and BCL-w (6.3 uM; second bar) on cytochrome c release from mitochondria. The effect of sensitizer BH3 peptides on restoration of cytochrome c release is also shown. Figure IE is a bar chart showing the effects of tBID (first bar) and MCL-I (1.1 uM; second bar) on cytochrome c release from mitochondria. The effect of sensitizer BH3 peptides on restoration of cytochrome c release is also shown.

0019Figure IF is a bar chart showing the effects of tBID (first bar) and BFL-I (2.4 uM; second bar) on cytochrome c release from mitochondria. The effect of sensitizer BH3 peptides on restoration of cytochrome c release is also shown.

0020Figure 2 is a bar chart showing the effect of BH3 peptides on cytochrome c release. MLM were treated with the indicated peptides at 10 uM in the presence or absence of 0.2 uM BCL-XL protein.

0021Figure 3A is a photograph of a Western blot showing the results of a GST-pulldown assay in which GST-BCL-w (or point mutant R96P) was combined with tBID protein and the indicated BH3 peptides (10 μ.M). The BHl domain R96P mutant of BCL-w lacks the ability to bind BH3 domains. For convenience, BCL-w binding pattern from Table Ib is excerpted below.

0022Figure 3B is a line graph showing the effects of BAD and NOXA BH3 peptides on displacement of a fluorescein tagged BIM BH3 peptide from BCL-2 and MCL-I proteins by fluorescence depolarization. Shown are representative plots from three independent experiments for each combination.

0023Figure 4A is a line graph showing the effects of IL- 3 withdrawal on the survival of wtFL5.12 and FL5.12-BCL-2 cells. Survival was imputed for cells not staining with Annexin V by FACS analysis. Shown is average and standard deviation of three independent experiments. Figure 4B is a line graph showing the effects of ABT-737, a BCL-2 antagonist, on the survival of IL-3 replete and IL-3 starved FL5.12-BCL-2 cells. Viability was assayed by absence of Annexin V staining. Shown is average and standard deviation of three independent experiments.

0024Figure 4C is a bar chart showing the effects of ABT-737 and ZVAD.fmk on the survival of FL5.12 cells (top). Also, a photograph of an immunoblot shows the effect of ABT-737 on PARP cleavage (bottom).

0025Figure 5 is a bar chart showing the effect of ABT-737 and ZVAD.fmk on FL5.12 viability. FL5.12 were grown either with IL-3 or in the absence of IL-3 for 24 hours, then treated as indicated for either 30 minutes or 1 , 2, or 3 hours. Cell death was measured by Annexin V staining via FACS analysis. Figure 6A is a bar chart showing the effects of BH3 peptides (10 uM) on cytochrome c release from mitochondria isolated from wtFL5.12 cells grown in the presence of IL-3 (blue bars); FL5.12-BCL-2 cells grown in the presence (red bars) or absence (tan bars) of IL-3 for 24 hours. Shown is average and standard deviation of three independent experiments. For convenience, BCL-2 binding pattern from Table Ib is excerpted below. Figure 6B is a bar chart showing the effects of NOXA and BAD BH3 on cytochrome c release from mitochondria isolated from wt and BCL-2 FL5.12 cells grown in the presence of IL-3. Shown is average and standard deviation of three independent experiments.

0026Figure 6C is a photograph of a Western blot depicting the effects of NOXA A, BAD, ABT-737, or control enantiomer on cytochrome c release from FL5.12 cells grown in the absence of IL-3 for 24 hours.

0027Figure 6D shows photographs of immunoblots depicting the effects of IL-3 withdrawal on BIM levels in FL5.12-BCL-2 whole cell lysates (left) and samples immunoprecipitated by an antibody directed against the human BCL-2 transgene product (right). Numbers at top refer to hours after IL-3 withdrawal. Control lane performed without antihuman BCL-2 antibody in pulldown at right.

0028Figure 6E is a photograph of a Western blot showing the results of an immunoprecipitation assay. BAX was immunoprecipitated using an antibody recognizing all BAX conformations (Δ21) or only the activated conformation with N-terminal exposure (NT). Death induced in the cells is indicated below. At right, mitochondria isolated from IL-3 starved cells were treated with ABT-737 or control enantiomer, and immunoprecipitation with NT performed as indicated. CD56 indicates control immunoprecipitation by an irrelevant antibody recognizing CD56. Figure 7A is a line graph showing the effects of MCL-I and BCL-2 on cell death induced by dexamethasone in 2B4 cells. Viability determined by absence of Annexin V staining by FACS. Shown is average and standard deviation of three independent experiments.

0029Figure 7B is a line graph showing the effects of BCL-2 antagonist ABT-737 on MCL-I dependent 2B4 cells and BCL-2 dependent 2B4 cells. Shown is average and standard deviation of three independent experiments.

0030Figure 7C is a bar chart showing the effects of BH3 peptides on cytochrome c release from mitochondria isolated from MCL-I -expressing 2B4 cells treated as indicated. Shown is average and standard deviation of three independent experiments. For convenience, MCL-I binding pattern from Table Ib is excerpted below.

0031Figure 7D is a bar chart showing the effects of BH3 peptides on cytochrome c release from mitochondria isolated from BCL-2-expressing 2B4 cells treated as indicated. For convenience, BCL-2 binding pattern form Table Ib is excerpted below.

0032Figure 7E is a photograph of an immunoblot showing the effects of dexamethasone on FLAG-MCL-I transfected 2B4 cells. FLAG antibody linked to agarose beads immunoprecipitated proteins complexing with FLAG-MCL-I. Increased BIM sequestration by MCL-I correlates with MCL-I dependence.

0033Figure 7F is a photograph of an immunoblot showing the effects of dexamethasone on FLAG-BCL-2 transfected 2B4 cells. Figure 7G is a bar chart showing the effects of BH3 peptides on dexamethasone-treated

0034FLAG-MCL-I 2B4 cells. Primed FLAG-MCL-I 2B4 cells transfected with BH3 peptides illustrate an MCL-I pattern.

0035Figure 8A is a bar chart showing the effects of BH3 peptides (10 μM, unless otherwise noted) on cytochrome c release from mitochondria isolated from liver. Figure 8B is a bar chart showing the effects of BH3 peptides on cytochrome c release from mitochondria isolated from BCL-2-dependent leukemia. For convenience, BCL-2 binding pattern from Table Ib is excerpted below. Shown is average and standard deviation of three independent experiments, except 30 and 100 μM treatments which were performed once.

0036Figure 8C is a picture of an immunoblot of samples from a BCL-2 dependent leukemia. First lane shows a whole cell lysate, 25 ug loaded; second lane - products of an immunoprecipitation using an antibody against the human BCL-2 transgene product; third lane - a control with Protein A beads alone; fourth lane - control immunoprecipitation using an irrelevant hamster monoclonal antibody recognizing murine CD-40.

0037Figure 8D is a bar chart showing the effects of BH3 peptides on cytochrome c release from mitochondria isolated from two SCLC cell lines, H146 and H1963. N=3 for each and the error bars represent the standard deviation.

0038Figure 9 is a model of selective cancer sensitivity to sensitizer BH3 mimetic treatment. Living unprimed cells (I) are primed for death following death stimuli (II). The leukemia cell is tonically primed for death without external intervention (II). Cells in the primed state undergo apoptosis in response to antiapoptotic antagonists (III) — those in the unprimed state do not. A) FL%.12-BCL-2. B) 2B4-MCL-1. C) myc/BCL-2 leukemia cells.

0039Figure 10 is a diagram depicting a model of BCL-2 family control of programmed cell death. Death signals cause induction or post-translational activation of BH3-only proteins. Activator BH3-only proteins, including BID and BIM, induce oligomerization of BAX and/or BAK, causing MOMP, cytochrome c release and caspase activation resulting in cell death. Antiapoptotic proteins prevent apoptosis by sequestering activator BH3-only proteins and

0040BAX/BAK, upstream of BAX/BAK oligomerization. Sensitzer BH3-only proteins promote cell death by binding the antiapoptotic proteins, displacing activator BH3-only proteins to trigger BAX./BAK oligomerization.

0041Figure 11 is a diagram depicting the intrinsic or mitochondrial programmed cell death pathway. In response to death signaling, activator BH3-only proteins are triggered to interact with BAX and BAK, inducing BAX and BAK oligomerization. This oligomerization is followed by permeabilization of the mitochondrial outer membrane, which releases proapoptotic factors like cytochrome c to the cytosol. Cytosolic cytochrome c forms a complex with APAF-I and Caspase-9 to make the holoenzyme known as the apoptosome, which in turn activates effector Caspase-3, leading to widespread proteolysis. This pathway can be interrupted by antiapoptotic members like BCL-2, which can bind activator BH3-only proteins, preventing their interaction with BAX and BAK. This inhibitory interaction can itself be antagonized by sensitizer BH3-only domains, which compete for the binding site in BCL-2, displacing activators bound by BCL-2. Figure 12A is a line graph showing the effect of ABT-737 on CLL cell viability. CLL cells from 24 patient samples were cultured for 48 hours with different concentrations of compounds. Death was quantitated by Annexin-V staining and normalized to solvent (DMSO) treated controls. Figure 12B is a line graph showing the effect of ABT-737 on CLL cell viability. CLL cells harvested from 5 patient samples were cultured for 4 hours with different concentrations of compounds. Death was quantitated as in (A).

0042Figure 12C is a line graph showing the effect of ABT-737 on normal PBMC viability. PBMCs were cultured for 24 hours in the presence of the indicated concentrations of compounds.

0043Figure 13A is a photograph of an immunoblot showing BCL-2 and BIM protein levels in whole cell lysates from CLL samples (number corresponds to patient number in Figure 12A). Three isoforms of BIM (BIM extra long-BIMEL, BIM long- BIML, and BIM short- BIMS) are shown.

0044Figure 13B is a photograph of an immunoblot showing BCL-2 and BIM protein levels in whole cell lysates from PBMC. Three normal PBMC lysates are at left, CLL lysates at right.

0045Figure 13C is a photograph of an immunoblot showing BCL-2 and BIM protein levels in whole cell lysates from two independent CLL primary samples made at time of cell harvest (pre) and 48 hours post-culture (post).

0046Figure 13D is a photograph of an immunoblot showing BCL-2 protein levels in primary CLL cells (A-F) and primary follicular lymphoma cells (FL). The immunoblots are indexed to lysates from the t(14;18)-containing H2 human lymphoma cell line.

0047Figure 14 is a bar chart showing the effects of BH3-only domain peptides (100 μM) or compounds (100 μM) on cytochrome c release (measured by ELISA) from mitochondria isolated from independent primary CLL patient samples. BADmu= a point mutant of the BAD BH3-only domain, used as a negative control. N=7, except for BADmu where N=5, ABT-737 and negative control enantiomer N=3. Error bars represent the standard deviation.

0048Figure 15A is a photograph of an immunoblot of BCL-2 and BIM proteins in whole cell lysates of seven independent CLL samples.

0049Figure 15B is a photograph of an immunoblot of BCL-2 and BIM proteins in whole cell lysates of primary CLL cells. Primary CLL cells were cultured for 24 hours with 100 nM ABT- 737, 100 nM negative control enantiomer, or vehicle (DMSO) + 200 μM ZVAD.fmk. Death was then quantitated by Annexin-V staining. Immunoprecipitation (i.p.) using lysates from each treatment group was performed using an anti-BCL-2 antibody. Results shown are representative of three independent experiments. Figure 15C is a photograph of an immunoblot. A BCL-2 antibody was used to immunoprecipitate BCL-2 from CLL lysates from four independent patient samples. After rinsing detergent away, the complex bound to the beads was then incubated with DMSO, 1 μM negative control enantiomer, or 1 μM ABT-737. Shown is the resulting immunoblot of the fraction displaced to the supernatant, probed for BIM.

0050Figure 15D is a photograph of an immunoblot. Freshly isolated CLL cells were incubated with DMSO, 1OnM, 10OnM, or iμM ABT-737 or negative control enantiomer for 4 hours. % dead was assessed by Annexin-V staining. Oligomerization of BAX was evaluated by anti-BAX immunoblot of chemically crosslinked whole cell lysates. Figure 15E is a bar chart showing the effect of 1% DMSO or 100 μM BAD BH3 peptide on mitochondria isolated from CLL samples. Samples were pre-incubated with antibodies directed against either the human BIM BH3 domain (Agent) or an irrelevant antigen (CD56) as indicated. N=5, bars show + standard deviation.

0051Figure 16A is a bar chart showing the effect of 100 μM BH3 peptides on cytochrome c release from mitochondria isolated from LPl cells.

0052Figure 16B is a photograph of an immunoblot of LPl and L363 cell lines comparing levels of MCL- I, BCL-2, and BIM.

0053Figure 16C is a line graph showing the effect of 48-hour treatment with ABT-737 on the viability of L363 cells and LPl cells. N=3, bars show + standard deviation. Figure 17 is a diagram depicting a model of ABT-737 induced death at the mitochondria.

0054Mitochondrial BCL-2 sequesters BIM in CLL cells. Upon addition of ABT-737, BIM is displaced and BCL-2 becomes occupied by ABT-737. Freed BIM then interacts with BAX or BAK₁ inducing oligomerization leading to cytochrome c release and irreversible commitment to programmed cell death. BCL-2 primed with activator BH3-only proteins renders the cancer cell sensitive to treatment with ABT-737 and possibly other chemotherapeutic agents.

0055Figure 18A is a chart showing the interaction pattern between BH3 peptides and anti- apoptotic proteins.

0056Figures 18B-E are a series of bar charts showing BH3 profiles for various lymphoma cell lines. Figure 19 A-B are a series of line graph showing cell sensitivity to various agents Figure 20 are a series of bar charts showing a comparison of mitochondrial and cell- based BH3 profiling.

DETAILED DESCRIPTION OF THE INVENTION

0058The present invention is based in part by the discovery of a third cellular state with respect to programmed cell death. This state has been named "primed for death". Until the present discovery, only two states had been identified with respect to programmed cell death, alive and dead. Cells that are primed for death require tonic antiapoptotic function for survival.

0059Using a panel of peptides derived from BH3 domains of BH3-only proteins that selectively antagonize individual BCL-2 family members BCL-2, BCL-XL, BCL-w, MCL-I and BFL-I, it was shown that cellular "addiction" to individual antiapoptotic proteins can be diagnosed based on mitochondrial response to these peptides. This panel of peptides is shown in Table 1 and are referred to herein BH3 domain peptides. Antiapoptotic proteins BCL-2, BCL- XL, MCL-I, BFL-I and BCL-w each bear a unique pattern of interaction with this panel of proteins. Cellular dependence on an antiapoptotic protein for survival is decoded based on the pattern of mitochondrial sensitivity to this peptide panel. This strategy is called BH3 profiling.

0060<img file="WO2007123791A2_D0001.tif" />

0061Mitochondria were probed to determine a cell's state using our panel of sensitizer BH3- peptides, selective antagonists of antiapoptotic BCL-2 family members. Mitochondria that are primed for death are dependent on antiapoptotic protein function to prevent MOMP, so that they release cytochrome c when exposed to sensitizer BH3 peptides (See, Figure 1, Figure 4A, Figure 5C, and Figure 6B). In contrast, unprimed cells do not release cytochrome c when exposed to sensitizer BH3 peptides. Any cell from which mitochondria can be isolated can therefore be so tested and categorized as being primed or unprimed. Testing of mitochondria directly has the advantage of eliminating any contribution of transcription, translation, or post- translational modification events that might be triggered by transfection of peptide, protein, or expression vector into a living cell. A "snapshot" of the apoptotic state at a given time may be taken with minimal perturbation of the extant apoptotic machinery. In summary, the methods of the invention allow capture if information about a fundamental aspect of cellular physiology. Importantly, mitochondrial behavior was correlated to whole cell behavior in several models. Mitochondria were primed when cells were enduring a physiologic challenge, and BH3 profiling revealed a dependence on antiapoptotic proteins only when a cellular dependence was also demonstrated. As shown below in the EXAMPLES, FL5.12 cells and mitochondria became primed for death only after IL-3 withdrawal. For 2B4 cells, cells and mitochondria were primed for death only after dexamethasone treatment. For the primary BCL-2 dependent leukemia cells, the genomic instability, myc oncogene activation and checkpoint violation inherent to the cancer phenotype were sufficient to induce mitochondrial priming without further external intervention. The SCLC cell lines H164 and H1693 revealed a BCL-2 pattern of sensitivity to BH3 profiling likewise are sensitive to the BCL-2 antagonist ABT-737. In each case, mitochondrial studies correctly diagnosed the cellular dependence on an antiapoptotic BCL-2 family member. Furthermore, the identity of the individual family member could be - decoded based on the pattern of mitochondrial sensitivity to our peptide panel These results indicate that in some cells, like IL-3 replete FL5.12-BCL-2 cells, BCL-2 overexpression provides extra antiapoptotic reserve. In others, like the murine leukemias, high levels of BCL-2 are present, but the BCL-2 is so highly occupied by activator BH3 proteins that the cell has very poor antiapoptotic reserve, and is actually primed for death.

0062Not all cells are sensitive to antagonism of antiapoptotic proteins. Sensitive cells are "primed for death" with death signals carried by a select subset of proapoptotic proteins of the BCL-2 family. Some cancer cells may be tonically primed for death, and thus are selectively susceptible to agents that provoke or mimic sensitizer BH3-only domains. It has been postulated that inhibition of apoptosis is a requirement of oncogenesis(Green and Evan, 2002; Hanahan and Weinberg, 2000). In what may be an attempt to meet this requirement, many types of cancer cells overexpress antiapoptotic BCL-2 family members. Understanding how these proteins function is therefore critical to understanding how cancer cells maintain survival. The methods of the present invention allows the systematic investigation how antiapoptotic BCL-2 family members interact with BH3-only family members to control mitochondrial outer membrane permeabilization (MOMP) and commitment to apoptosis. Antiapoptotic proteins show selective affinity for binding BH3 peptides derived from BH3-only proteins. Furthermore, antagonism of antiapoptotic family members results in MOMP only when the antiapoptotic proteins are "primed" with activator BH3 proteins, validating the critical role of activator BH3 domains in activating BAX/BAK. In cell culture models, activator "priming" can be observed following experimentally-induced death signaling, and that such priming confers dependence on antiapoptotic family members. Remarkably, dependence on antiapoptotic BCL-2 family members can be captured functionally by the pattern of mitochondrial sensitivity to sensitizer BH3 domains. Accordingly, the invention features methods of determining the sensitivity of a cell to a therapeutic agent by identifying whether or not a cell is primed for death by determining the pattern of mitochondrial sensitivity to BH3 domain peptides.

0063BH3 Profiling

0064In various methods, sensitivity of a cell to an agent is determined. Cell sensitivity is determined by contacting a cell or cellular component (e.g., mitochondria) with a BH3 domain peptide. A cell is sensitive to an agent if apoptosis is detected. Alternatively, cell sensitivity is determined by providing a test BH3 profile of the cell and comparing the profile to a cancer cell BH3 profile. A similarity of the test profile and the control profile indicates that the cell is sensitive to an agent. A BH3 profile is a pattern of sensitivity to BH3 peptides of the cell. Sensitivity is indicated by apoptosis. A cancer cell BH3 profile is a pattern of sensitivity to BH3 peptides in a cancer cell whose responsiveness or lack there of to a particular agent is known. Optionally, the test BH3 profile is compared to more that one cancer cell BH3 profile. Thus, by comparing the test BH3 profile to the control BH3 profile sensitivity to an agent is determined.

0065The cell or cellular component is a cancer cell or a cell that is suspected of being cancerous. The cell is permeabilized to permit the BH3 peptides access to the mitochondria.

0066Cells are permeabilized by methods known in the art. For example, the cell are permeabilized by contacting the cell with digitonin. After the cell is permeabilized, the cells are treated with a potentiometric dye. Examples of potentiometric dyes incude he green-fluorescent JC-I probe (5,5',6,6'-tetrachloro- 1 , 1 ',3,3'-tetraethylbenzimidazolylcarbocyanine iodide) or dihydrorhodamine 123.

0067JC-I is a lipophilic, cationic dye that enters mitochondria in proportion to the membrane potential JC- I exists as a monomer in water at low membrane potential (M). However, at higher potentials, JC-I forms red- fluorescent "J-aggregates". As a monomer the dye has an absorption/emission maxima of 527 nm while at high membrane potential the emission maximum is 590 nm. Thus, ratio measurements of the emission of this cyanine dye can be used as a sensitive measure of mitochondrial membrane potential. The dye allows for a dual measurement of dye concentration that does not require the measurement of a nuclear or cytoplasmic reference values. Studies using isolated mitochondria have shown that the 527 nm emission from monomeric JC-I increases almost linearly with membrane M potentials ranging from 46 to 182 mV, whereas the 590 nm J-aggregate emission is less sensitive to M values less negative than 140 mv and is strongly sensitive to potential values in the range of 140 to 182 mV (Di Lisa et al., 1995) Optical filters designed for fluorescein and tetramethylrhodamine can be used to separately visualize the monomer and J-aggregate forms, respectively. Alternatively, both forms can be observed simultaneously using a standard fluorescein longpass optical filter set.

0068Dihydrorhodamine 123 an uncharged, nonfluorescent agent that can be converted by oxidation to the fluorescent laser dye rhodamine 123 (R123). The cell is from a subject known to or suspected of having cancer. The subject is preferably a mammal. The mammal is, e.g., a human, non-human primate, mouse, rat, dog, cat, horse, or cow. The subject has been previously diagnosed as having cancer, and possibly has already undergone treatment for cancer. Alternatively, the subject has not been previously diagnosed as having cancer. The agent is a therapeutic agent such as a chemotherapeutic agent. For example the agent is a mimetic of sensitizer BH3 domains or an antagonist of an anti-apoptotic protein. Apoptosis, i.e., cell death is identified by know methods. For example, characteristics of apoptosis include the cell shrinks, develop bubble-like blebs on their surface, have the chromatin (DNA and protein) in their nucleus degraded, and have their mitochondria break down with the release of cytochrome c, loss of mitochondrial membrane potential, break into small, membrane- wrapped, fragments, or phosphatidylserine, which is normally hidden within the plasma membrane, is exposed on the surface of the cell.

0069The difference in the level apoptosis of a cell that has been contacted with a BH3 peptide compared to a cell that has not been contacted with a BH3 peptide is statistically significant. By statistically significant is meant that the alteration is greater than what might be expected to happen by chance alone. Statistical significance is determined by method known in the art. For example statistical significance is determined by p-value. The p-value is a measure of probability that a difference between groups during an experiment happened by chance. (P(z>z_Observed))- For example, a p-value of 0.01 means that there is a 1 in 100 chance the result occurred by chance. The lower the p-value, the more likely it is that the difference between groups was caused by treatment. An alteration is statistically significant if the p-value is or less than 0.05. Preferably, the p-value is 0.04, 0.03, 0.02, 0.01, 0.005, 0.001 or less.

0070The invention also includes a profile of a pattern of mitochondrial sensitivity to BH3 sensitizer peptides taken from one or more subjects who have cancer.

0071BH3 Domain Peptides

0072A BH3 domain peptide is less than 195 amino acids in length, e.g., less than or equal to 150, 100, 75, 50, 35, 25 or 15 amino acid in length. For example a BH3 peptide includes the sequence of SEQ ID NO: 1-13 shown in Table 1.

0073A BH3 domain peptide include a peptide which includes (in whole or in part) the sequence NH₂- XXXXXXIAXXLXXXGDXXXX -COOH (SEQ ID NO: 14) or NH₂- XXXXXXXXXXLXXXXDXXXX -COOH (SEQ ID NO: 15). As used herein X may be any amino acid. Alternatively, the BH3 domain peptides include at least 5, 6, 7, 8, 9, 15 or more amino acids of SEQ ID NO: 14 or SEQ ID NO: 15).

0074Optionally, the BH3 domain peptide is attached to transduction domain. A transduction domain compound that directs a peptide in which it is present to a desired cellular destination , Thus, the transduction domain can direct the peptide across the plasma membrane, e.g., from outside the cell, through the plasma membrane, and into the cytoplasm. Alternatively, or in addition, the transduction domain can direct the peptide to a desired location within the cell, e.g., the nucleus, the ribosome, the ER, mitochondria, a lysosome, or peroxisome.

0075In some embodiments, the transduction domain is derived from a known membrane- translocating sequence. Alternatively, transduction domain is a compound that is known to facilitate membrane uptake such as polyethylene glycol, cholesterol moieties, octanoic acid and decanoic acid.

0076For example, the trafficking peptide may include sequences from the human immunodeficiency virus (HIV) 1 TAT protein. This protein is described in, e.g., U.S. Patent Nos. 5,804,604 and 5,674,980, each incorporated herein by reference. The BH3 domain peptide is linked to some or all of the entire 86 amino acids that make up the TAT protein. For example, a functionally effective fragment or portion of a TAT protein that has fewer than 86 amino acids, which exhibits uptake into cells can be used. See e.g., Vives et al, J. Biol. Chem., 272(25): 16010-17 (1997), incorporated herein by reference in its entirety. A TAT peptide that includes the region that mediates entry and uptake into cells can be further defined using known techniques. See, e.g., Franked et al, Proc. Natl. Acad. Sci, USA 86: 7397-7401 (1989). Other sources for translocating sequences include, e.g., VP22 (described in, e.g., WO 97/05265; Elliott and O'Hare, Cell 88: 223-233 ( 1997)), Drosophila Antennapedia (Antp) homeotic transcription factor, HSV, poly-arginine, poly lysine, or non-viral proteins (Jackson et al, Proc. Natl. Acad. Sci. USA 89: 10691 - 10695 ( 1992)).

0077The transduction domain may be linked either to the N-terminal or the C-terminal end of BH3 domain peptide. A hinge of two proline residues may be added between the transduction domain and BH3 domain peptide to create the full fusion peptide. Optionally, the transduction domain is linked to the BH3 domain peptide in such a way that the transduction domain is released from the BH3 domain peptide upon entry into the cell or cellular component.

0078The transduction domain can be a single (Le., continuous) amino acid sequence present in the translocating protein. Alternatively it can be two or more amino acid sequences, which are present in protein, but in the naturally-occurring protein are separated by other amino acid sequences. The amino acid sequence of naturally-occurring translocation protein can be modified, for example, by addition, deletion and/or substitution of at least one amino acid present in the naturally-occurring protein, to produce modified protein. Modified translocation proteins with increased or decreased stability can be produced using known techniques. In some embodiments translocation proteins or peptides include amino acid sequences that are substantially similar, although not identical, to that of naturally-occurring protein or portions thereof. In addition, cholesterol or other lipid derivatives can be added to translocation protein to produce a modified protein having increased membrane solubility.

0079The BH3 domain peptide and the transduction domain can be linked by chemical coupling in any suitable manner known in the art. Many known chemical cross-linking methods are non-specific, i.e.; they do not direct the point of coupling to any particular site on the transport polypeptide or cargo macromolecule. As a result, use of non-specific cross-linking agents may attack functional sites or sterically block active sites, rendering the conjugated proteins biologically inactive.

0080One way to increasing coupling specificity is to directly chemical coupling to a functional group found only once or a few times in one or both of the polypeptides to be cross- linked. For example, in many proteins, cysteine, which is the only protein amino acid containing a thiol group, occurs only a few times. Also, for example, if a polypeptide contains no lysine residues, a cross-linking reagent specific for primary amines will be selective for the amino terminus of that polypeptide. Successful utilization of this approach to increase coupling specificity requires that the polypeptide have the suitably rare and reactive residues in areas of the molecule that may be altered without loss of the molecule's biological activity.

0081Cysteine residues may be replaced when they occur in parts of a polypeptide sequence where their participation in a cross-linking reaction would otherwise likely interfere with biological activity. When a cysteine residue is replaced, it is typically desirable to minimize resulting changes in polypeptide folding. Changes in polypeptide folding are minimized when the replacement is chemically and sterically similar to cysteine. For these reasons, serine is preferred as a replacement for cysteine. As demonstrated in the examples below, a cysteine residue may be introduced into a polypeptide's amino acid sequence for cross-linking purposes. When a cysteine residue is introduced, introduction at or near the amino or carboxy terminus is preferred. Conventional methods are available for such amino acid sequence modifications, whether the polypeptide of interest is produced by chemical synthesis or expression of recombinant DNA. Coupling of the two constituents can be accomplished via a coupling or conjugating agent. There are several intermolecular cross-linking reagents which can be utilized, See for example, Means and Feeney, CHEMICAL MODIFICATION OF PROTEINS, Holden-Day, 1974, pp. 39- 43. Among these reagents are, for example, J-succinimidyl 3-(2-pyridyldithio) propionate (SPDP) or N, N'- (1,3-phenylene) bismaleimide (both of which are highly specific for sulfhydryl groups and form irreversible linkages); N, N'-ethylene-bis- (iodoacetamide) or other such reagent having 6 to 1 1 carbon methylene bridges (which relatively specific for sulfhydryl groups); and l,5-difluoro-2, 4-dinitrobenzene (which forms irreversible linkages with amino and tyrosine groups). Other cross-linking reagents useful for this purpose include: p,p'-difluoro-m,m'- dinitrodiphenylsulfone (which forms irreversible cross-linkages with amino and phenolic groups); dimethyl adipimidate (which is specific for amino groups); phenol-1,4- disulfonylchloride (which reacts principally with amino groups); hexamethylenediisocyanate or diisothiocyanate, or azophenyl-p-diisocyanate (which reacts principally with amino groups); glutaraldehyde (which reacts with several different side chains) and disdiazobenzidine (which reacts primarily with tyrosine and histidine). Cross-linking reagents may be homobifunctional, i.e., having two functional groups that undergo the same reaction. A preferred homobifunctional cross-linking reagent is bismaleimidohexane ("BMH"). BMH contains two maleimide functional groups, which react specifically with sulfhydryl-containing compounds under mild conditions (pH 6.5-7.7). The two maleimide groups are connected by a hydrocarbon chain. Therefore, BMH is useful for irreversible cross-linking of polypeptides that contain cysteine residues.

0082Cross-linking reagents may also be heterobifunctional. Heterobifunctional cross-linking agents have two different functional groups, for example an amine-reactive group and a thiol- reactive group, that will cross-link two proteins having free amines and thiols, respectively. Examples of heterobifunctional cross-linking agents are succinimidyl 4-(N-maleimidomethyl) cyclohexane-1-carboxylate ("SMCC"), m-maleimidobenzoyl-N-hydroxysuccinimide ester ("MBS"), and succinimide 4-(p-maleimidophenyl) butyrate ("SMPB"), an extended chain analog of MBS. The succinimidyl group of these cross-linkers reacts with a primary amine, and the thiol-reactive maleimide forms a covalent bond with the thiol of a cysteine residue.

0083Cross-linking reagents often have low solubility in water. A hydrophilic moiety, such as a sulfonate group, may be added to the cross-linking reagent to improve its water solubility. Sulfo-MBS and sulfo-SMCC are examples of cross-linking reagents modified for water solubility. Many cross-linking reagents yield a conjugate that is essentially non-cleavable under cellular conditions. However, some cross-linking reagents contain a covalent bond, such as a disulfide, that is cleavable under cellular conditions. For example, Traut's reagent, dithiobis (succinimidylpropionate) ("DSP"), and N-succinimidyl 3-(2-pyridyldithio) propionate ("SPDP") are well-known cleavable cross-linkers. The use of a cleavable cross-linking reagent permits the cargo moiety to separate from the transport polypeptide after delivery into the target cell. Direct disulfide linkage may also be useful.

0084Numerous cross-linking reagents, including the ones discussed above, are commercially available. Detailed instructions for their use are readily available from the commercial suppliers. A general reference on protein cross-linking and conjugate preparation is: Wong, CHEMISTRY OF PROTEIN CONJUGATION AND CROSS-LINKING, CRC Press (1991).

0085Chemical cross-linking may include the use of spacer arms. Spacer arms provide intramolecular flexibility or adjust intramolecular distances between conjugated moieties and thereby may help preserve biological activity. A spacer arm may be in the form of a polypeptide moiety that includes spacer amino acids, e.g. proline. Alternatively, a spacer arm may be part of the cross-linking reagent, such as in "long-chain SPDP" (Pierce Chem. Co., Rockford, IL., cat. No. 21651 H).

0086The BH3 domain peptides and/or the transduction domain peptides can be polymers of L- amino acids, D-amino acids, or a combination of both. For example, in various embodiments, the peptides are D retro-inverso peptides. The term "retro-inverso isomer" refers to an isomer of a linear peptide in which the direction of the sequence is reversed and the chirality of each amino acid residue is inverted. See, e.g., Jameson et al., Nature, 368, 744-746 (1994); Brady et al., Nature, 368, 692-693 (1994). The net result of combining D-enantiomers and reverse synthesis is that the positions of carbonyl and amino groups in each amide bond are exchanged, while the position of the side-chain groups at each alpha carbon is preserved. Unless specifically stated otherwise, it is presumed that any given L-amino acid sequence of the invention may be made into a D retro-inverso peptide by synthesizing a reverse of the sequence for the corresponding native L-amino acid sequence. Alternatively, the BH3 domain peptides and/or the transduction domain peptides are cyclic peptides. Cyclic peptides are prepared by methods known in the art. For example, macrocyclization is often accomplished by forming an amide bond between the peptide N- and C-termini, between a side chain and the N- or C-terminus [e.g., with K₃Fe(CN)ή at pH 8.5] (Samson et al., Endocrinology, 137: 5182-5185 (1996)), or between two amino acid side chains. See, e.g., DeGrado, Adv Protein Chem, 39: 51-124 (1988).

0087BH3 domain peptides and/or the transduction domain peptides are easily prepared using modern cloning techniques, or may be synthesized by solid state methods or by site-directed mutagenesis. A domain BH3 peptide and/or the transduction domain peptides may include dominant negative forms of a polypeptide. In one embodiment, native BH3 domain peptides and/or transduction domain peptides can be isolated from cells or tissue sources by an appropriate purification scheme using standard protein purification techniques. In another embodiment, BH3 domain polypeptides and/or transduction domain peptides are produced by recombinant DNA techniques. Alternative to recombinant expression, BH3 domain peptides and/or transduction domain peptides can be synthesized chemically using standard peptide synthesis techniques.

0088An "isolated" or "purified" protein or biologically active portion thereof is substantially free of cellular material or other contaminating proteins from the cell or tissue source from which the BH3 domain peptide is derived, or substantially free from chemical precursors or other chemicals when chemically synthesized. The language "substantially free of cellular material" includes preparations of BH3 peptides and/or transduction domain peptides in which the protein is separated from cellular components of the cells from which it is isolated or recombinantly produced. In one embodiment, the language "substantially free of cellular material" includes preparations of BH3 domain peptides and/or the transduction domain peptides having less than about 30% (by dry weight) of non- BH3 domain peptide and/or non- transduction domain peptides (also referred to herein as a "contaminating protein"), more preferably less than about 20% of non- BH3 peptide and/or πon- transduction domain peptides, still more preferably less than about 10% of non- BH3 peptide and/or non- transduction domain peptides, and most preferably less than about 5% non-BH3 domain peptide and/or non- transduction domain peptides . When the BH3 domain peptide and/or the transduction domain peptides or biologically active portion thereof is recombinantly produced, it is also preferably substantially free of culture medium, i.e., culture medium represents less than about 20%, more preferably less than about 10%, and most preferably less than about 5% of the volume of the protein preparation. The language "substantially free of chemical precursors or other chemicals" includes preparations of BH3 domain peptides and/or the transduction domain peptides in which the protein is separated from chemical precursors or other chemicals that are involved in the synthesis of the protein. In one embodiment, the language "substantially free of chemical precursors or other chemicals" includes preparations of BH3 domain peptides and/or transduction domain peptides having less than about 30% (by dry weight) of chemical precursors or non-BH3 domain peptide and/or non-transduction domain peptides chemicals, more preferably less than about 20% chemical precursors or non-BH3 domain peptide and/or non- transduction domain peptides chemicals, still more preferably less than about 10% chemical precursors or non-BH3 domain peptide chemicals, and most preferably less than about 5% chemical precursors or non-BH3 domain peptide and/or non- transduction domain peptides chemicals.

0089The term "biologically equivalent" is intended to mean that the compositions of the present invention are capable of demonstrating some or all of the same apoptosis modulating effects, i.e., release of cytochrome C or BAK oligomerization although not necessarily to the same degree as the BH3 domain polypeptide deduced from sequences identified from cDNA libraries of human, rat or mouse origin or produced from recombinant expression symptoms.

0090Percent conservation is calculated from the above alignment by adding the percentage of identical residues to the percentage of positions at which the two residues represent a conservative substitution (defined as having a log odds value of greater than or equal to 0.3 in the PAM250 residue weight table). Conservation is referenced to sequences as indicated above for identity comparisons. Conservative amino acid changes satisfying this requirement are: R-K; E-D, Y-F, L-M; V-I, Q-H.

0091BH3 domain peptides can also include derivatives of BH3 domain peptides which are intended to include hybrid and modified forms of BH3 domain peptides including fusion proteins and BH3 domain peptide fragments and hybrid and modified forms in which certain amino acids have been deleted or replaced and modifications such as where one or more amino acids have been changed to a modified amino acid or unusual amino acid and modifications such as glycosylation so long as the hybrid or modified form retains the biological activity of BH3 domain peptides . By retaining the biological activity, it is meant that cell death is induced by the BH3 polypeptide, although not necessarily at the same level of potency as that of the naturally- occurring BH3 domain polypeptide identified for human or mouse and that can be produced, for example, recombinantly. The terms induced and stimulated are used interchangeably throughout the specification. Preferred variants are those that have conservative amino acid substitutions made at one or more predicted non-essential amino acid residues. A "conservative amino acid substitution" is one in which the amino acid residue is replaced with an amino acid residue having a similar side chain. Families of amino acid residues having similar side chains have been defined in the art. These families include amino acids with basic side chains (e.g., lysine, arginine, histidine), acidic side chains (e.g., aspartic acid, glutamic acid), uncharged polar side chains (e.g., glycine, asparagine, glutamine, serine, threonine, tyrosine, cysteine), nonpolar side chains (e.g., alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, tryptophan), beta-branched side chains (e.g., threonine, valine, isoleucine) and aromatic side chains (e.g., tyrosine, phenylalanine, tryptophan, histidine). Thus, a predicted nonessential amino acid residue in a BH3 domain polypeptide is replaced with another amino acid residue from the same side chain family.

0092Alternatively, in another embodiment, mutations can be introduced randomly along all or part of a BH3 coding sequence, such as by saturation mutagenesis, and the resultant mutants can be screened to identify mutants that retain activity.

0093Also included within the meaning of substantially homologous is any BH3 domain peptide which may be isolated by virtue of cross-reactivity with antibodies to the BH3 domain peptide described herein or whose encoding nucleotide sequences including genomic DNA, mRNA or cDNA may be isolated through hybridization with the complementary sequence of genomic or subgenomic nucleotide sequences or cDNA of the BH3 domain peptides herein or fragments thereof. The invention will be further illustrated in the following non-limiting examples.

EXAMPLE 1: GENERAL METHODS

0095Reagents ABT-737 and its negative control enantiomer which has lower affinity for BCL-2 family members were obtained from Abbott Laboratories(OItersdorf T, 2005).

0096GST-pulldown

009710 ug GST-BCL-w (or BH3 binding -defective R96P point mutant) were incubated with glutathione-agarose beads for one hour at 4° C in binding buffer (140 mM NaCl, 10 mM Tris pH7.4). Beads were rinsed and incubated with approximately 0.2 ug tBID for 1 hr at 4° C. Beads were washed again and incubated with peptides for 1 hour at 4°C. tBID protein was eluted from beads with 50 mM glutathione and loaded on a denaturing NuPAGE gel.

0098Cytochrome c release Mitochondria were purified from liver and FL5.12 cells as previously described(Letai et al., 2002). Mitochondria were purified from leukemia cells and 2B4 cells as previously described for FL5.12 cells. Mitochondria were incubated with treatments for 45 (mouse liver mitochondria) or 35 minutes (FL5.12, 2B4, and leukemic mitochondria). Release of cytochrome c was determined by a comparison of cytochrome c in the pellet and supernatant following treatment, quantitated by ELISA (R&amp;D systems). When results of multiple experiments were averaged, results from solvent-only (DMSO) treatments values were subtracted from each, so that 0 release reflects that observed in solvent-only treatments.

0099Alternatively, mitochondria were purified from freshly isolated CLL cells and cell lines by mechanical disruption followed by differential centrifugation, as previously described (Letai et al., 2002). Mitochondrial suspensions were made at 0.5 mg protein/ml, except for the case of ABT-737 and negative control enantiomer treatments, where 0.1 mg/ml was used. Release of cytochrome c was determined by a comparison of cytochrome c in the pellet and supernatant quantitated by ELISA (R&amp;D systems).

0100Peptides Peptides were synthesized by Tufts University Core Facility and purified by HPLC.

0101Identity was confirmed by mass spectrometry. Stock solutions were made in DMSO. Peptides used for fluorescence polarization were synthesized with an N-terminus fluorescein tag and a 6- aminohexanoic acid linker. Sequences were taken from published sequences of murine BAD (LWAAQRYGRELRRMSDEFEGSFKGL) (Kelekar et al., 1997), BADmu (LWAAQRYGREARRMSDEFEGSFKGL, note L-> A, a point mutation abrogating binding to BCL-2) (Zha et al., 1997), NOXA A (AELPPEFAAQLRKIGDKVYC) (Oda et al., 2000), BMF (HQA EVQl A RKLQL [A DQ FH R) (Puthalakath et al., 2001) and human BID (EDIIRNIARHLAQVGDSMDR) (Wang et al., 1996), BIM (MRPEIWIAQELRRIGDEFNA) (O'Connor et al., 1998), BIK (MEGSDALALRLACIGDEMDV) (Boyd et al., 1995), BNIP3-CC (VVEGEKEVEALKKSADWVSD) (Yasuda et al., 1999), hari-kiri (HRK) (SSAAQLTAARLKALGDELHQ) (Inohara et al., 1997) and PUMA (EQWAREIGAQLRRMADDLNA) (Nakano and Vousden, 2001). Recombinant proteins

0102Antiapoptotic proteins were expressed in bacteria and affinity purified using glutathione- agarose (for GST-linked proteins) as previously described (Letai et al., 2002) or by nickel-NTA agarose beads (for His-tagged MCL-I) according to manufacturer protocol (Qiagen). Selected samples were additionally purified by anion exchange FPLC to obtain sufficient purity as judged by Coomassie staining of a denaturing gel. In each case, the C-terminus transmembrane domain was truncated to maintain solubility in aqueous solution. For binding assays, GST-linked proteins were used for BCL-2, BCL-XL, BCL-w and BFL-I; His-tagged MCL-I was used. For the mitochondrial assays, the same proteins were used except for MCL-I, where a GST-linked protein was used. Constructs expressing GST-MCL-I, GST-BFL-I, and GST-BCL-w were kind gifts of Tilman Oltersdorf; His-tagged MCL-I construct was a kind gift of Ruth Craig. The human sequence was used for each. Recombinant tBID was made as previously described; it contained double cysteine to serine substitutions which maintains wild-type ability to induce cytochrome c release(Oh et al., 2005). Fluorescence polarization binding Assays

0103Binding assays were performed using fluorescence polarization as previously described(Letai et al., 2002). A minimum of three independent experiments were used to determine each dissociation constant. For BIM BH3 displacement assays, 25 nM fluorescein- 1 inked BIM BH3 peptide was bound to 0.5 uM GST-BCL-2 or 0.1 uM GST-MCL-I in binding buffer. NOXA or BAD BH3 peptides were then titrated and displacement of BIM BH3 monitored by loss of fluorescence polarization. fmmunoprecipitation

0104Cell lysates (250 ug) were incubated with 6C8 hamster anti-human BCL-2 antibody (3 ug) for at least 1 hour at room temperature in 1% CHAPS buffer ( 5 mM sodium phosphate pH 7.4; 2.5 mM EDTA; 100 mM sodium chloride; 1% w/v CHAPS, in the presence of protease inhibitors (Complete tablets; Roche)). Protein A-sepharose beads (Sigma) were added to precipitate complexes containing BCL-2. The beads were mixed with loading buffer prior to loading supernatant onto gel. For FLAG-MCL-I immunoprecipitation, 2B4 cells were treated with 100 nM dexamethasone for 24 hours and lysed in 1% CHAPS buffer. 250 ug protein lysate was incubated with anti-FLAG antibody-conjugated agarose beads (Sigma) for 1 hour at 4°C. Protein was eluted from washed beads with 2.5 ug FLAG peptide. Eluant was loaded onto denaturing gel for electrophoresis. Alternatively, Cell lysates (250 μg) were incubated with 6C8 hamster anti-human BCL-2 antibody (3 μg) for at least 1 hour at 4°C in 0.1% Triton-XlOO buffer. Protein A-sepharose beads (Sigma) were added to precipitate complexes containing BCL-2. The beads were mixed with loading buffer prior to loading supernatant onto a 10% Bis-Tris polyacrylamide gel for analysis. For displacement reactions, 50μg of lysate were incubated with 3μg 6C8 BCL-2 antibody for at least 1 hour at 4°C in 0.1% Triton-X100 buffer or CHAPS buffer. Protein A-sepharose beads were added and incubated for 1 hour. Then the beads were pelleted and washed 3 times and resuspended in HE buffer (1 mM EDTA and 10 mM HEPES, pH 7.4, as in (Chipuk et al. 2004)). lμM ABT-737, lμM negative control enantiomer, or DMSO was added to the tube, incubated overnight; the supernatant was loaded onto a 10% Bis-Tris polyacrylamide gel (Invitrogen) for analysis. lmmunoblots

0105Protein lysates were obtained by cell lysis in 1% CHAPS buffer. Protein samples were size fractionated on NuPAGE 10% Bis-Tris polyacrylamide gels (Invitrogen). Antibodies were used to detect the following proteins on membrane: BIM (Calbiochem, 22-40); BCL-2 (Pharmingen, /100); PUMA (Prosci, NT); rabbit polyclonal anti-murine BID(Wang et al., 1996) ;BAK (Upstate, NT); BAX (Santa Cruz, N-20); Acdn (Chemicon, MAB1501); CD-40 (Pharmingen, HM40-3); MCL-I (Rockland). BAX oligomerization performed as previously described(Letai et al., 2002).

0106Alternatively, protein lysates were obtained by cell lysis in Triton-X100 (142.5 mM NaCl, 5mM MgC12, 10 mM HEPES, 1 mM EGTA, 0.1% Triton-X100 (Sigma)), RIPA (150 mM NaCl, 2 mM EDTA, 0.1 M Na2HPO4 pH 7.2, 0.2 mM NaVO4, 50 mM NaF, 1% sodium deoxycholate, 0.1 % SDS, and % NP-40(Sigma)) or CHAPS ( 100 mM NaCl, 5 mM NaPO4, 2.5 mM EDTA, 1% CHAPS (Sigma)) buffer supplemented with a Complete protease inhibitor cocktail tablet (Roche). Protein samples were electrophoretically separated on NuPAGE 10% Bis-Tris polyacrylamide gels (Invitrogen). Antibodies were used to detect the following proteins on membrane: BIM (Calbiochem 22-40 or Abgent BH3 domain); BCL-2 (Pharmingen, /100); MCL- I (Chemicon, RC- 13).

0107Annexin- V assay Cells were stained with fluorescent conjugates of Annexin-V (BioVision) and propidium iodide (PI) and analyzed on a FACSCalibur machine (Becton-Dickinson).

0108Isolation and short-term culture of human cells

010915 ml of blood in heparin treated tubes was obtained from each anonymous CLL patient and processed without freezing. Equal volume of media (RPMI medium, 10% human bovine serum, supplemented with 10 μg/ml insulin and 10 mg/ml transferrin) was mixed with each sample and CLL cells isolated by centrifugation through Ficoll-PAQUE Plus (Amersham). Cells were washed twice in media and cultured at a density of 2.0x106 cells/ml for up to 48 hours. Samples for Figure 1 IA were obtained from 24 consecutive patients identified with WBC >50,000/μl. All guidelines and regulations were followed in accordance with IRB protocols #99-224 (Dana-Farber Cancer Institute). Normal PBMCs were obtained from tubing discarded following platelet donation by anonymous normal donors and processed as above, except cultured without insulin and transferrin.

0110CLL clinical criteria FISH analysis was performed by the Brigham and Women's Hospital cytogenetics laboratory using a CLL panel of multicolor probe sets (Vysis, Inc.) (Dohner et al., 2000). CLL cells were processed and IgVH and ZAP70 status were determined by the CLL Research Consortium Tissue core using previously established methods (Rassenti et al. 2004). Somatic hypermutation in the IgVH locus was classified as absent when >98% homology to germline was measured. Patient samples were classified as ZAP70 positive when >20% cells were positive; CD38 positive when >30% cells were positive. Multiple Myeloma Cell culture LPl and L363 cells (kind gift from Ruben Carrasco) were cultured in Iscove's modified Dulbecco's medium with 10% fetal bovine serum.

0111Cell culture FL5.12 cells were cultured as described previously in Iscove's modified Dulbecco's medium, 10% fetal bovine serum, 1000 ug/ml G418 with or without IL-3 provdied by 10% WEHI-3B supplement (supernatant of IL-3 secreting WEHI-3B cells). FL5.12 cells were stably transfected with a vector containing a neomycin resistance construct and either human BCL-2 cDNA (FL5.12-BCL-2) or no insert (wt). 2B4 cells were cultured in RPMI 1640 supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100 ug/ml streptomycin, 10 uM non-essential amino acids and 8 ul/L betamercaptoethanol. Stably transfected 2B4 cells were isolated after trasnfection with pLZR-GFP retroviral vector or vector with Flag-Mcl-1 (kind gift from Joe Opferman). Caspase Inhibition Experiment

011210⁶ cells/mL Neo or BCL-2 FL5.12 cells were plated in the media containing 1L-3, or 2.OxIO⁶ cells/mL cells were washed twice with IXPBS and plated in media without IL-3 for 24 hours. Cells receiving caspase inhibitor were incubated with 200μM ZVAD.fmk (Calbiochem) for 1 hour prior to any additional treatment. Cells were treated with lμM ABT-737 or NCE for 30 minutes, 1 , 2, 3 or 4 hours as indicated and then stained with Annexin-V/PI to assess apoptotic status. For protein analysis, cells were harvested, washed with IX PBS twice and lysates made in 1% CHAPS buffer. lOug was loaded onto a 10% Bis-Tris protein gel. Resulting immunoblots were probed with anti-PARP antibody (BioVision), which recognizes both cleaved and uncleaved PARP protein.

0113BAX oligomer ization

011410⁷ freshly isolated CLL cells were incubated with DMSO, 10 nM, 100 nM, or 1 μM ABT-737 or negative control enantiomer for 4 hours, % dead assessed by Annexin-V staining and FACS analysis, then treated with 0.3% saponin and 10 μM BMH for 30 minutes on ice. Cells were then lysed, loaded onto a 10% BISTris polyacrlyamide gel, transferred, and immunoblotted for BAX.

0115Mice

0116Leukemia prone mice were generated as previously described (Letai, 2004). Mouse experimental protocols conform to the relevant regulatory standards and were approved by the Dana-Farber Cancer Institute Animal Care and Use Committee.

0117Statistical analyses

0118Experimental replicates were performed using lysates or mitochondria from different CLL samples. In main body, where a P value is given, it was obtained by using a two-tailed Students Mest, and P &lt; 0.05 was considered statistically significant. GraphPad Prism software was used to determine EC50 values by non-linear dose-response curve fitting and to perform Mann-Whitney nonpaiametric testing in Table 5.

0119EXAMPLE 2: ANTIAPOPTOTIC PROTEINS DEMONSTRATE DISTINCT PROFILES OF BINDING SENSITIZER BH3 PEPTIDES

0120To determine selectivity in interactions among antiapoptotic BCL-2 family members and BH3 domains of BH3-only proteins, fluorescence polarization binding assays

0121(FPA) were used. Antiapoplotic proteins BCL-2, BCL-XL, MCL-I, BCL-w, and BFL-I were purified from transfected bacteria as GST fusion proteins. BH3-domains were synthesized as 20-25-mers as shown in Table 2a. Oligopeptides used for FPA were tagged with an N-terminal FITC moiety. Table 2b quantitates binding by dissociation constants.

0122Table 2

0123<img file="WO2007123791A2_D0002.tif" />

0124b

0125<img file="WO2007123791A2_D0003.tif" />

0126It is immediately notable that the antiapoptotic family members may be distinguished from each other based on affinity for individual BH3 domains. For instance, BCL- XL may be distinguished from BCL-2 and BCL-w by its much greater affinity for HRK BH3. Otherwise, though there are quantitative distinctions among binding patterns of BCL-2, BCL-XL and BCL-w, the quantitative binding patterns are quite similar, suggesting similarity in the hydrophobic binding pockets of these three molecules.

0127In contrast with this group, MCL-I does not bind BAD BH3, in agreement with data generated by pull-down (Opferman et al., 2003), yeast two-hybrid (Leo et al., 1999); and surface plasmon resonance (Chen et al., 2005) assays. Murine NOXA is unique among the known BH3-only proteins in that it possesses two 5 putative BH3 domains (Oda et al., 2000). It is notable that while the other four proteins interact with neither of the NOXA BH3 domains tested, MCL-I interacts with both. This suggests that the interaction between NOXA and MCL- 1 is indeed biologically significant. The ability to bind both BH3 domains suggests the possibility of novel multimeric interactions between MCL-I and murine NOXA, or alternatively differential control over exposure of the two BH3 domains in NOXA. Also distinct is BFL-I. While it binds BID and BIM, it binds only PUMA among the sensitizers tested. It is also notable that the activators BID and BIM BH3 are bound by all of the antiapoptotics tested, distinguishing them from the sensitizers which, except PUMA, show a more selective pattern of binding. Of additional note is that the BH3 domain obtained from BNIP-3α binds to none of the proteins tested, and does not activate BAX or BAK. While the possibility that BNIP BH3 interacts with an untested multi-domain pro- or antiapoptotic BCL-2 family member cannot be excluded, it is also possible that BNIP-3α does not function as a BH3- family member at all (Ray et al., 2000). EXAMPLE 3: DEPENDENCE ON INDIVIDUAL ANTIAPOPTOTIC PROTEINS MAY BE DEDUCED BY PATTERN OF SENSITIVITY TO SENSITIZER BH3 PEPTIDES; INHIBITION OF ANTIAPOPTOTIC PROTEIN IS INSUFFICIENT FOR MOMP UNLESS ACTIVATOR TBID IS PRESENT

0128Previous results have shown that the BH3 domains of BID and BIM possess the ability to induce BAX and BAK oligomerization and cytochrome c release in a purified mitochondrial system (Letai et al., 2002). This class is referred to as the BH3 domain "activators." BH3 domains from BAD and BIK (termed "sensitizers") were unable to induce cytochrome c release on their own. However, when an activator was bound and sequestered by BCL-2, preventing interaction of the activator with BAX or BAK, sensitizers could provoke mitochondrial apoptosis by competitively inhibiting BCL-2's binding of the activator, freeing the activator to oligomerize BAX or BAK and induce cytochrome c release. Thus, the two sensitizer BH3 domains were shown to be antagonists of BCL-2 antiapoptotic function. The ability to antagonize BCL-2 function correlated with high-affinity binding to BCL-2.

0129In Table Ib above, the expanded range of BH3 domains tested in the present study demonstrate distinct patterns of binding to antiapoptotic proteins. To test if selective binding corresponded to ability of individual BH3 domains to selectively antagonize antiapoptotic function, a purified mitochondrial system was constructed in which the critical apoptosis decision making molecular machinery was reconstituted. For the activator function, caspase-8 cleaved BID protein, tBID, was used. tBID is an archetypical activator protein, capable of inducing BAX/BAK oligomerization and cytochrome c release in purified mitochondria (Wei et al., 2000) and synthetic liposomes (Kuwana et al., 2005; Kuwana et al., 2002). tBID's induction of cytochrome c release and apoptosis requires BAX or BAK (Cheng et al., 2001; Wei et al., 2001). The multidomain proapoptotic function was provided by the BAK which resides in mouse liver Mitochondria; mouse liver mitochondria contain no detectable BAX protein (Letai et al., 2002). The dominant antiapoptotic function was provided by one of the 6 five different recombinant antiapoptotic proteins used in the binding assays. BH3 peptides provided the sensitizer function.

0130As cytochrome c release is the readout for the system, it was important to test whether the peptides by themselves release cytochrome c in mouse liver mitochondria like activators BID or BIM BH3 (Letai et ah, 2002). Figure IA is a confirmatory assay that shows none of the sensitizer BH3 peptides by themselves can induce cytochrome c release significantly above background, even at concentrations 10- fold higher than those used in Figure IB-F. While this has previously been shown for BAD, BIK, NOXA A, and NOXA B BH3's, this is a novel finding for the HRK, BNIP, PUMA and BMF BH3 domains. In each of the subsequent panels, cytochrome c release by tBID is demonstrated, followed by inhibition of cytochrome c release by addition of either BCL-2 (b), BCL-XL (c), BCL-w (d), MCL-I (e), or BFL-I (0- The ability of the panel of BH3 domains to antagonize antiapoptotic protection as measured by cytochrome c release was determined. Remarkably, in each case, the ability to antagonize antiapoptotic function maps to the binding specificities in Table Ib. This is important confirmation that the binding pattern elucidated in Table Ib corresponds to biological function.

0131It is important to emphasize that treatment with sensitizer peptides alone, even those that bind and antagonize all the antiapoptotics tested, such as PUMA BH3, or the combination of NOXA and BAD BH3 is insufficient to cause cytochrome c release (Figure IA). Furthermore, when the panel of sensitizer BH3 peptides was tested in the presence of the antiapoptotic protein BCL-XL, there was still no cytochrome c release, formally ruling out the possibility that the BH3 peptides were somehow directly converting antiapoptotic proteins to a proapoptotic function (Figure 2). To induce MOMP and cytochrome c release, there appears to be an absolute requirement for an activator function, here provided by the tBID protein. These data critically demonstrate that the panel of peptides can determine whether a mitochondrion depends on an antiapoptotic protein to maintain integrity. Furthermore, the identity of the critical antiapoptotic protein can be deduced based on the pattern of sensitivity to the panel of sensitizer BH3 peptides. This strategy is termed BH3 profiling.

0132EXAMPLE 4: SENSITIZERS DISPLACE ACTIVATORS FROM ANTIAPOPTOTIC PROTEINS Since sensitizer BH3 peptides cannot induce cytochrome c release on their own, but can induce cytochrome c release when activator and antiapoptotic proteins are present, in a pattern that mirrors their binding to antiapoptotic proteins, it was hypothesized that the sensitizers are displacing activators from the antiapoptotic proteins. As one test of this hypothesis, the ability of sensitizer peptides to displace tBID from antiapoptotic protein Bcl-w was examined utilizing a GST-pulldown assay. Proteins bound to glutathioneagarose beads were eluted with glutathione and analyzed by Western blot. In Figure 3A, tBID is displaced from BCl-w by sensitizer BH3 peptides in a pattern that replicates the pattern in Figure ID. As an additional test, the displacement of the activator BlM BH3 peptide from BCL-2 and MCL-I by BAD and NOXA BH3 peptides was determined. In Figure 3B, consistent with Table Ib, BAD BH3 efficiently displaces BIM BH3 from BCL-2, but not MCL-I, whereas NOXA A BH3 7 efficiently displaces BIM from MCL-I , but not BCL-2. These experiments support the ability of sensitizer BH3 peptides to displace activators from the antiapoptotic binding cleft. EXAMPLE 5: A CELLULAR REQUIREMENT FOR BCL-2 CORRESPONDS TO A "BCL-2 PATTERN" OF MITOCHONDRIAL SENSITIVITY TO THE SENSITIZER BH3 PANEL

0133In order to test whether mitochondrial dependence on individual antiapoptotic protein function can be correlated with cellular behavior, cellular models of defined antiapoptotic dependence were investigated. First, it was determined if cellular requirement for BCL-2 for cellular survival correlates with the BCL-2 signature of mitochondrial sensitivity to sensitizer

0134BH3 domains found in Figure I B. The pro-lymphocytic murine FL5.12 cell line requires IL-3 to maintain survival. Apoptosis induced by IL-3 withdrawal is inhibited by overexpression of BCL-2 (Figure 4A). Therefore, BCL-2-overexpressing FL5.12 (FL5.12-BCL-2) cells deprived of IL-3 are a model of BCL-2 dependent survival. FL5.12-BCL-2 cells grown in the presence of IL-3 are examples of BCL-2 independent cells.

0135While the dependence on BCL-2 of IL-3 deprived FL5.12 cells is demonstrated genetically in Figure 4A, the dependence was confirmed using a cell-permeable BCL-2 antagonist. ABT-737 has been shown to antagonize BCL-2 (and BCL-XL and BCL-w) (Oltersdorf et al., 2005). In agreement with the prior report, ABT-737 induced cell death in the IL-3 starved, but not the IL-3 replete BCL-2 protected cells (Figure 4B). Moreover, ABT-737 was non-toxic to the unstressed IL-3 replete wt FL5.12 cells. This cell death was caspase dependent, demonstrating that death occurred using the apoptotic pathway (Figure 4C). IL-3 - cells were grown in the absence of IL-3 for 24 hours prior to initiation of treatment with compound. All cells were treated with compounds for 24 hours prior to harvest. Having credentialed a BCL-2 dependent cellular system, it was next determined if BCL-2 dependence could be isolated at the level of mitochondria. It was hypothesized that removal of IL-3 would "load" the BCL-2 on the mitochondria with activator BH3 proteins. It was further hypothesized that mitochondria bearing "loaded" BCL-2 would release cytochrome c when treated with sensitizer BH3 peptides which compete for the BCL-2 binding cleft. ABT -737 inhibition of BCL-2 in FL5.12-BCL-2 cells primed by IL-3 withdrawal induced an apoptosis that is caspase dependent and very rapid (Figure 5). The interpretation that the IL-3 starved FL5.12- BCL-2 cells were "primed" for death is supported by the rapidity of their death following ABT- 737 treatment (Figure 5).

0136Mitochondria were isolated from wt FL5.12 cells and FL5.12-BCL-2 cells in the presence of IL-3, and from FL5.12-BCL-2 cells following 24 hours of IL-3 deprivation. Due to advanced apoptosis, mitochondria could not be isolated in sufficient quantities from wt FL5.12 cells after IL-3 deprivation. Figure 6A shows that while activators BID and BIM potently induce cytochrome c release from mitochondria isolated from wt FL5.12 cells, the remaining sensitizer peptides do not (blue bars). Thus, inhibition of antiapoptotic family members is by itself not sufficient to induce MOMP. Next, BCL-2 overexpression inhibits release induced by 10 μM BID BH3, but not 10 μM BIM BH3, in accordance with dose-response curves previously demonstrated (red bars) (Letai et al., 2002). When mitochondria from FL5.12-BCL-2 cells deprived of IL-3 were tested, however, certain senstizer peptides demonstrated the ability to induce cytochrome c release (tan bars), and sensitivity to 10 μM BID BH3 is restored. It is most notable that only those sensitizer peptides with high affinity for BCL-2 cause MOMP. BIK BH3 does not induce cytochrome c release in this setting, but it should be noted that it has approximately 10-fold lower affinity than BAD, PUMA or BMF BH3 for BCL-2. It can be seen, therefore, that cellular BCL-2 dependence can be "diagnosed" from the pattern of mitochondrial sensitivity to the panel of sensitizer BH3 peptides. This dependence can be "diagnosed" whether the activator involved is a recombinant protein, as in Figure 1, or a more complex mix involving more that one molecule, as is likely the case following IL-3 withdrawal. Nofe that inhibition of BCL-2 alone is not sufficient to induce cytochrome c release, as seen by the failure of all of the sensitizer peptides to induce release in the IL-3 replete FL5.12-BCL-2 mitochondria (Figure 6A). In fact, even the combination of peptides, BAD and NOXA BH3, which provide a broad spectrum of antiapoptotic protein binding, cannot induce cytochrome c release in the absence of an activator molecule (Figure 6B). To induce MOMP, the BCL-2 must first be "primed" by molecules communicating a death signal, generated by IL-3 withdrawal. Mitochondria isolated from FL5.12-BCL-2 cells grown in the absence of IL-3 for 24 hours were treated with NOXA A or BAD peptides (30 uM) or ABT-737 or control enantiomer at 10 uM for 35 minutes. Mitochondrial pellets were subjected to chemical crosslinking as previously described (Letai et al., 2002). BCL-2 blocks apoptosis upstream of BAX oligomerization, and BAD BH3 and ABT-737 inhibition of BCL-2 on IL-3 starved mitochondria results in BAX oligomerization (Figure 6C). Therefore, it was hypothesized that this death signal might be an activator BH3 protein.

0137BIM has previously been shown to play a role in death following IL-3 withdrawal in FL5.12 cells (Harada et al., 2004). Figure 6D shows that total cellular BIM levels, as well as levels of BIM complexed to BCL-2, dramatically increase following IL-3 withdrawal. It is notable that levels of BCL-2, BAX, and BAK stay roughly constant during the same time period. These results suggest that the activator BIM (and perhaps PUMA) is a dynamic mediator of the death response following IL-3 withdrawal in FL5.12 cells, and that it is sequestered to prevent apoptosis. Cells and mitochondria bearing "loaded" BCL-2 are then "addicted" to BCL-2, and die when BCL-2 function is antagonized. Furthermore, cellular BCL-2 addiction can be diagnosed by the pattern of mitochondrial sensitivity to sensitizer BH3 domains. The negative control of immunoprecipitation using anti-human BCL-2 antibody on lysates from IL-3 starved FL5.12-BCL-XL cells yielded no bands, not shown.

0138This model predicts that BCL-2 acts upstream of BAX activation by intercepting activator BH3 molecules. To test this prediction, in Figure 6E, immunoprecipitation was performed with an antibody that recognizes only the activated form of BAX which exposes an N-terminus epitope (Desagher et al., 1999; Hsu and Youle, 1997). wt or BCL-2 expressing FL5.12 cells were exposed to IL-3 withdrawal as indicated. IL-3 withdrawal induced BAX activation in wt FL5..12 cells, while total BAX levels remained constant. However, when BCL- 2 protected against death from IL-3 withdrawal, it also prevented BAX conformational change, consistent with BCL-2's sequestering activators like BIM prior to their interaction with BAX (compare fourth and eight lanes). Furthermore, treatment with ABT-737 restored cytochrome c release and BAX activation, consistent with ABT-737 functioning by displacing activators from BCL-2. Taken together, these results indicate that BCL-2 blocks apoptosis upstream of BAX activation.

0139EXAMPLE 6: BH3 PROFILING CAN DISCRIMINATE MCL-1 CELLULAR DEPENDENCE FROM BCL- 2 CELLULAR DEPENDENCE

0140To test if the model of antiapoptotic "priming" could be extended beyond BCL-2 to other antiapoptotic proteins, the behavior of cells protected by BCL-2 was compared to those protected by MCL-I. 2B4 cells transfected with Flag-MCL-1, BCL-2, or empty vector constructs were cultured for 24 hours in the presence of the indicated concentration of dexamethasone. The murine hybridoma 2B4 cell line is sensitive to dexamethasone treatment. Overexpression of FLAGtagged MCL-I or BCL-2 confers resistance to dexamethasone- induced apoptosis (Figure 7A). Therefore, dexamethasone-treated, FLAG-MCL-I expressing cells are a model of cellular MCL-I dependence, while dexamethasone-treated, BCL-2 expressing cells are a model of cellular BCL-2 dependence. 2B4 cells were incubated with dexamethasone and either ABT-737 or enaπtiomer for 24 hours. Treatment of the MCL-I -protected dexamethasone-treated cells with ABT-737 has no effect, showing the cells are not dependent on BCL-2 for survival. In stark contrast, 2B4 cells protected from dexamethasone-induced apoptosis by BCL-2 are very sensitive to ABT-737 (Figure 7B).

0141The cellular data provoke the prediction that mitochondria isolated from 2B4-MCL-1 cells treated with dexamethasone would be sensitive to NOXA and insensitive to BAD BH3, the opposite of the pattern observed with IL-3-starved FL5.12-BCL-2 cells. Mitochondria were isolated from dexamethasone-treaied and untreated vector-transfected and FLAG-MCL-I- transfected 2B4 cells. Apoptosis was too advanced to permit isolation of mitochondria from dexamethasone treated vector-transfected 2B4 cells. As can be seen in Figure 7C, only mitochondria isolated from the MCL-I dependent cells recapitulate an "MCL-I pattern" of sensitivity to sensitizer BH3 peptides. As with the FL5.12 cells, since sensitizer BH3 peptides cause little cytochrome c release in untreated cells, it is clear that sensitizer BH3 peptide inhibition of MCL-I (and other antiapoptotic proteins that might be present) is not by itself sufficient to induce apoptosis. An additional death signal (initiated by dexamethasone treatment in this case) is needed to "prime" MCL-I so that MCL-I antagonism by sensitizers results in mitochondrial permeabilization. To demonstrate the robustness of this strategy, BH3 profiling was performed on 2B4 cells treated with dexamethasone, but this time protected with BCL-2. Consistent with the priming model, a BCL-2 pattern is revealed (Figure 7D). Thus, MCL-I dependence, like BCL-2 dependence, also can be "diagnosed" by mitochondrial sensitivity to the sensitizer BH3 panel. As with the FL5.12 cells, it was determined if dexamethasone treatment resulted in increased sequestration of an activator BH3 protein by MCL-I and BCL-2. Vector or FLAG- MCL-I transfected 2B4 cells were treated with 0 or 100 nM dexamethasone and lysed. Figure 7E shows that FLAG-MCL- 1 sequesters increased amounts of BIM following the death signaling induced by dexamethasone treatment, as does BCL-2 (Figure 7F). Note that levels of BAX and BAK stay constant during the treatment. Also note that it appears that the small amount of BAX bound to cells before treatment with dexamethasone decreases after treatment. One interpretation is that the BAX is displaced by increased levels of BIM binding to BCL-2. This is significant because it suggests that displacement of BAX from MCL-I is insufficient to induce MOMP and death. To further demonstrate that the mitochondrial assays reflect true cellular dependence, peptides were transfected via electroporation into I⁷LAG-MCL- 1-transfected 2B4 cells that had been treated with dexamethasone, putatively priming MCL-I with death signals, carried at least in part by BIM. Percent killing was ascertained by Annexin-V staining. Note that killing by transfection with NOXA A is significantly greater than that with BAD, recapitulating the same MCL-I pattern observed in isolated mitochondria from this same cell line (Figure 7G, compare with Table Ib, Figure Ie, Figure 7C). N=3 and the error bars represent the standard deviation. These results support the cellular relevance of the mitochondrial BH3 profiling assays.

0142EXAMPLE 7: DEPENDENCE ON BCL-2 IN A LEUKEMIA CORRESPONDS TO MITOCHONDRIAL SENSITIVITY TO SENSITIZERS IN A "BCL-2 PATTERN" AND SEQUESTRATION OF BIM

0143Dependence on antiapoptotic proteins is perhaps of greatest importance in the context of cancer, in which antiapoptotic BCL-2 family proteins are subjects of intense investigation as therapeutic targets. While the concept of oncogene addiction has received attention recently(Jonkers and Berns, 2004; Weinstein, 2002), the molecular details of the addiction to specific oncogenes is poorly understood. A validated model of oncogene addiction, a BCL-2 dependent murine leukemia, was used to examine the molecular basis for BCL-2 "addiction."

0144Previous results have described a mouse acute lymphocytic leukemia model in which c- myc is constituitively expressed and human BCL-2 is repressibly expressed. In this model, when BCL-2 transgene expression is eliminated by administration of doxycycline, the leukemic cells undergo apoptosis, resulting in rapid resolution of the leukemia (Letai, 2004). This provides an ideal in vivo model of a BCL-2 dependent cancer. It was hypothesized that dependence on BCL- 2 was due to a similar mechanism to that of the IL-3 deprived FL5.12-BCL-2 cells - that is, a death signal was being initiated and carried by an activator BH3 molecule, but BCL-2 was binding it and preventing its interaction with multidomain proapoptotic proteins. Mitochondria were isolated from leukemia cells and exposed to sensitizer BH3 peptides.

0145Subsequently, cytochrome c release was measured. As an internal control, mitochondria were isolated from liver from the leukemic mice in parallel (Figure 8A). The sensitizer BH3 peptides were unable to induce cytochrome c release from non-malignant hepatocyte mitochondrias from the leukemic mice, just as they were unable to induce cytochrome c release from non-malignant liver mitochiondria from normal mice (Figure Ia) or from non-malignant FL5.12 (Figure 6A) or 2B4 mitochondria (Figure 7C). Intriguingly, certain sensitizer BH3 peptides were capable of inducing near total cytochrome c release from the leukemic mitochondria (Figure 8B). Significantly, the pattern of peptides which induced release corresponded exactly to those peptides which bind with high affinity to BCL-2 (Table Ib), namely BAD, BIK, PUMA, and BMF. Note that, consistent with its approximately ten-fold lower affinity than BAD BH3 for BCL-2, BIK BH3 requires a ten-fold higher concentration to demonstrate cytochrome c release. A ten-fold increase in NOXA A peptide concentration has no effect, consistent with the extremely low affinity NOXA A has for BCL-2.

0146These results suggest that in this leukemia model, death signals are being continually initiated, and BCL-2 is required to sequester the activator BH3 molecule to prevent apoptosis. In contrast to the non-malignant systems tested above, leukemic cell BCL-2 behaves as if already "primed" with activator protein(s) without any further intervention, such as growth factor withdrawal or dexamethasone treatment. Figure 8C shows that BIM is expressed in the leukemia cells, and it is bound by BCL-2. Supporting the signal importance of BIM in transmitting death signals in this model, BID is also present in the lysate, but is not bound by BCL-2. Note that PUMA is also found to be bound by BCL-2, consistent with a report showing PUMA deficiency could accelerate myc-induced lymphomagenesis (Hemann et al., 2004). Since the PUMA BH3 lacks the ability to directly activate BAX or BAK, it was hypothesized that PUMA is acting as a sensitizer in this context, in effect decreasing the amount of BCL-2 available to bind BIM and possibly BAX or BAK.

0147If BCL-2 maintains survival of this leukemia cell primarily by sequestering BIM, then one would predict that BIM loss of function could substitute for BCL-2 overexpression to cooperate with c-myc in leukemogenesis. In fact, this experiment has already been performed. It was found that BIM deficiency can indeed cooperate with c-myc to produce a pre-B lymphocytic leukemia like the one produced here by the cooperation of BCL-2 overexpression with c-myc (EgIe et al., 2004). These results support a model in which BCL-2 is necessary for survival of leukemia largely because it is required to sequester BIM, preventing activation of BAX/BAK and subsequent MOMP. The leukemia cells are therefore neither normal and healthy, nor dead, but rather primed for death. EXAMPLE 8: BH3 PROFILING PREDICTS SENSITIVITY TO ABT-737

0148As another test of the ability of BH3 profiling to detect in vivo BCL-2 dependence, two small cell lung cancer (SCLC) cell lines were examined. Mitochondria were isolated from two SCLC cell lines, H146 and H1963 and exposed to panel of BH3 peptides. Cytochrome c release quantitatcd by ELISA. Both were sensitive to treatment with ABT-737 in vitro and in an in vivo murine xenograft model (Oltersdorf T, 2005). Both H146 and H1963 demonstrate a pattern of sensitivity diagnostic of BCL-2 sensitivity (Figure 8D). This provides further support, in addition to the results of Figure 4B and 7E, that mitochondrial BH3 profiling is a powerful predictor of what cells are sensitive to BH3 mimetic drugs in vitro and in vivo. EXAMPLE 9: ABT-737 KILLS CLL CELLS IN THE LOW NANOMOLAR RANGE

0149ABT-737 is a cell-permeant small molecule with affinity to BCL-2, BCL-XL, and BCL-w in the sub-nanomolar range. The negative control enantiomer (enant) is a stereoisomer of ABT-737 that binds to BCL-2 with lower affinity, and has been used as a loss of function control (Oltersdorf et al., 2005). CLL cells have shown sensitivity to ABT- 737, so they were selected as a possible initial cancer model of BCL-2 dependence (Oltersdorf et al., 2005). Freshly isolated primary CLL cells were incubated with ABT-737 or negative control enantiomer. After 48 hours, death was assessed using Annexin-V staining. In all twenty-four CLL samples tested, apoptosis was induced within 48 hours by ABT-737 with an EC50 of 4.5 + 2.2 nM (Figure 12A) (range 1.9-9.4 nM, see Table 3). The negative control enantiomer was less potent (mean EC50 574 nM) (Figure 12A). Directly targeting the apoptotic machinery might be expected to induce apoptosis rapidly. Five samples treated for four hours responded similarly compared to 48-hour treatment (Figure 12B). Non-malignant peripheral blood mononuclear cells (PBMC) from normal donors were resistant to ABT-737 with an EC50 >1000 nM. (Figure 12C) Table 3

0150<img file="WO2007123791A2_D0004.tif" />

0151i- ₁ αtr « ion v

EXAMPLE 10: BCL-2 AND BIM LEVELS ARE CONSISTENT AMONG CLL SAMPLES

0153Given the consistency of response to ABT-737, it was hypothesized that BCL-2 protein expression was also uniform in CLL cells. In addition, proapoptotic BIM has been shown to be an important determinant of commitment to apoptosis in lymphocytic cells (Bouillet et al., 1999; Opferman et al., 2003). Antiapoptotic BCL-2 and proapoptotic BIM levels among 15 CLL samples were remarkably uniform (Figure 13A); levels of BCL-2 and BIM in PBMCs were consistently lower (Figure 13B). To ensure that short-term culture did not affect BCL-2 or BlM levels and perhaps alter response to ABT-737, protein lysates were made from CLL samples at time of isolation and after 48 hours in culture. Neither BCL-2 nor BIM levels changed during culture (Figure 13C). BCL-2 levels in CLL cells were compared to levels in follicular lymphoma, a cancer in which BCL-2 is overexpressed due to the t(14;18) translocation (Figure 13D). BCL-2 levels were notably similar in the two diseases.

0154Table 4 summarizes the clinical characteristics of the source patients in Figure 12 A. EC50 values were compared between groups dichotomized by factors previously identified as prognostically useful in CLL. This analysis revealed that in no case did a difference in mean EC50 between groups exceed 2 nM. A nonparametric statistical comparison of the groups showed that none differed with statistical significance except the groups dichotomized by leukocyte count (Table 5). Thus, biological response to ABT-737 appears to be largely independent of traditional prognostic factors.

0155Table 4

0156Table 5

0157<img file="WO2007123791A2_D0005.tif" />

0158EXAMPLE 11: CLL MITOCHONDRIA REVEAL A TONIC DEPENDENCE ON BCL-2 FUNCTION TO MAINTAIN OUTER MEMBRANE INTEGRITY Since BCL-2 opposes the intrinsic, or mitochondrial, pathway of apoptosis, it was hypothesized that the toxicity of ABT-737 was based on a mitochondrial requirement for BCL-2 function in CLL. A panel of peptides has been characterized, which derived from the BH3 domains of BH3-only proteins that behave as selective antagonists of BCL-2 function in several genetically defined model systems (Letai et al., 2002) (and MC, VDM,

0159Nishino M , Wei G, Korsmeyer S, Armstrong S, AL, in preparation). For instance, BH3 peptides from BAD, PUMA, BMF, and, with lower potency, BIK bind and inhibit BCL-2 function, whereas BH3 domains from NOXA, HRK and BNIP-3A do not interact with BCL-2. This panel has been validated on other antiapoptotic family members, and this pattern of interaction is distinct for each antiapoptotic protein, so that the function of each may be specifically detected by this "BH3 profiling." Therefore, mitochondria that depend on BCL-2 function for maintenance of their outer membrane integrity show induction of outer membrane permeability when treated with BAD, PUMA and BMF, but not NOXA, HRK, and BNIP-3A peptides.

0160CLL mitochondria were incubated with BH3 peptides as well as ABT-737 and negative control enantiomer (Figure 14). Note that activators BID and BIM BH3 peptides interact with all antiapoptotic proteins tested and furthermore can directly activate BAX and BAK (Letai et al., 2002), so that they act as positive controls for cytochrome c release assays. BAD, PUMA and BMF induce cytochrome c release, whereas the NOXA, HRK, and BN1P-3A peptides, and a point null mutant of BAD BH3, do not. This pattern is diagnostic of mitochondrial BCL-2 dependence. ABT-737 also induced cytochrome c release, validating that its target is located at the mitochondria of CLL cells and required to maintain mitochondrial outer membrane integrity. Therefore, these "BH3 profiling" experiments demonstrate that the CLL mitochondria depend on BCL-2 function to maintain outer mitochondrial membrane integrity, elucidating a mechanism for the exquisite sensitivity of CLL cells to ABT-737 treatment. Since the BH3 peptides in the sensitizer panel lack the ability to directly activate BAX and BAK, these experiments also implicated the presence of an activator molecule constitutively sequestered by BCL-2 in CLL. EXAMPLE 12: BIM BOUND TO BCL-2 PRIMES CLL CELLS FOR KILLING BY ABT-737

0161Previously studies have shown that ABT-737 and the sensitizer BH3 peptides act as antagonists of antiapoptotic BCL-2, but lack the ability to directly activate BAX and BAK. In order to induce MOMP, sensitizers require the presence of an activator like BIM or BID (Letai et al., 2002; Oltersdorf et al., 2005). The results above, therefore, suggest an activator is bound to BCL-2, then displaced by ABT-737 or the BCL-2 binding BH3 peptides. Following displacement, it was hypothesized that the freed activator could induce MOMP via interaction with BAX and BAK. Figure 13 demonstrated the presence of BIM in CLL samples. Levels of BID, the other established activator BH3-only protein, were very faint by immunoblot, and cleaved, activated BID was almost to completely undetectable (not shown). It was hypothesized that BIM was indeed bound by BCL-2. Figure 15A shows that BIM is sequestered by BCL-2 in primary CLL cells. Furthermore, treatment with ABT-737, and to a lesser degree, the less potent negative control enantiomer, causes a dramatic reduction in the amount of BIM bound to BCL-2 (Figure 15B). Interestingly, displaced BIM seems to be rendered less stable, as total cellular levels of BIM, but not BCL-2, are reduced following ABT-737 treatment. Co-treatment with a pan-caspase inhibitor ZVAD-fmk protects against ABT-737-induced death, further implicating an apoptotic death. ZVAD-fmk also preserves cellular BIM levels, suggesting that displaced BIM may be cleaved by caspases as previously reported (Chen and Zhou, 2004). Detergents in lysates can interfere with the binding of sensitizers to the BCL-2 hydrophobic cleft (not shown), so it was hypothesized that the observed persistence of BIM bound to BCL-2 after ABT-737 and ZVAD-fmk treatment might be an artifact due to detergent in the lysate inhibiting ABT-737's binding to BCL-2. Two competing hypotheses of ABT-737 mechanism of action were generated. In the first, ABT-737 displaces BIM from BCL-2, inducing BAX oligomerization and caspase activation, and finally caspase cleavage of the displaced BIM. In the second, BIM degradation is merely a consequence of MOMP and caspase activation that is initiated by a mechanism independent of BIM displacement.

0162To test these competing hypotheses, whole cell lysates were examined for BCL-2:BIM complex levels in the presence or absence of ABT-737. In detergent-free conditions, ABT-737, but not the negative control enantiomer, displaces BIM from BCL- 2 into the supernatant (Figure 15C). BIM has been shown to activate BAX and induce its oligomerization (Letai et al., 2002; Marani, et al., 2002; Yamaguchi and Wang, 2002).

0163Consistent with this mechanism, CLL cells display oligomerization of BAX within four hours of treatment with ABT-737 (Figurel5D).

0164If the first hypothesis is correct, and BIM is required for inducing the MOMP following BCL-2 antagonism, selective sequestration of BIM should cause a reduction in cytochrome c release following antagonism of BCL-2 on CLL mitochondria. In Figure 15E, BCL-2 function is antagonized with the sensitizer BAD BH3 peptide. As shown in Figure 12, BAD BH3 by itself induced cytochrome c release. However, addition of an antibody that binds the BH3 domain of BIM caused a dramatic reduction in cytochrome c release. An irrelevant antibody had no effect. Prevention of cytochrome c release by masking BIM supports the first hypothesis, that antagonism of ABT-737 is toxic to CLL cells due to the displacement of BIM (Figure 15C) from a BCL-2:BIM complex. Displaced BIM then induces BAX oligomerization (Figure 15D), MOMP and commitment to programmed cell death. An important implication of these results is that BCL-2 expression is necessary but not sufficient to dictate response to antagonism of BCL-2 by ABT-737 or sensitizer BH3 peptides. Activator BH3-only proteins like BIM must be sequestered by BCL-2 for the cell to be sensitive to BCL-2 antagonism. EXAMPLE 14: PREDICATION OF DRUG RESPONSE BY BH3 PROFILING

0165BH3 profiling allows for the the prediction of response of cancer cells to anti-cancer therapeutics. For the purposes of this application, therapeutics can be divided into those that target anti-apoptotic BCL-2 proteins, and conventional agents. As a model test system, four lymphoma cell lines SU-DHL4, SU-DHL6, SU-DHL8, and SU-DHLlO were employed. To perform BH3 profiling, the ability of a panel of sensitizer peptides to induce mitochondrial outer membrane permeabilzation (MOMP) in mitochondria isolated from the lymphoma cells were tested For easy reference, Figure 18A shows the interaction pattern between the BH3 peptides and anti-apoptotic proteins. MOMP was measured by quantifying cytochrome c release by ELISA.

0166As shown in Figure 18, BH3 profiling proved able to distinguish these three classes of blocks in our sample of four lymphoma lines. SU-DHL4 and SU-DHL6 a "primed" phenotype, based on the sensitivity to sensitizer BH3 peptides. Note that a strong response to the PUMA BH3 peptide, which interacts with all of the antiapoptotic proteins, provides a useful gauge of whether the mitochondria are primed. The pattern of sensitivity (PUMA, BMF, BAD, +/- BIK) indicated a dependence on BCL-2 for SU-DHL-4. SU-DHL 6 also was primed, as shown by a strong PUMA BH3 and BMF BH3 signal. The weaker, but definite, response to both of the more selective BH3 peptides BAD BH3 and NOXA A BH3 implicate combined dependence on BCL-2 and MCL-I. SU-DHL-8 appeared to be poorly primed, given the limited response to PUMA BH3 and other sensitizers, but nonetheless demonstrated an intact effector arm by responding strongly to activators BIM BH3 and BID BH3. SU-DHL-10 responded poorly to both sensitizer and activator peptides, indicating the loss of the effector arm. Only cells that are dependent on BCL-2 for survival are predicted to respond to BCL-2 antagonists like ABT-737. Therefore, BH3 profiling predicts that SU-DHL4 and SU-DHL6 should respond to ABT-737. This hypothesis was tested and confirmed that BH3 profiling accurately predicted response to ABT-737 (Figure 19A). Furthermore, BH3 profiling would predict that SU-DHL4 and SU-DHL6 might be more sensitive to other chemotherapy agents, as they are "primed" for death and SUDHL-8 and SUDHL-10 were not. To test this, cells were treated with vincristine. As predicted by BH3 profiling SU-DHL4 and SU-DHL6 were the most sensitive cell lines (Figure 19B). EXAMPLE 15: CELL BASED BH3 PROFILING A method that converts the mitochondrial-based BH3 profiling to a cell-based assay was developed. In this assay, cells are permeabilized by digitonin to permit peptide access to mitochondria. Cells are treated with the fluorescent dye JC-I to evaluate loss of mitochondrial transmembrane electrochemical potential due to treatment with the peptides. Loss of mitochondrial integrity due to apoptosis can be observed by a shift in the fluorescence peak from 590 nm to 520 nm. Multiple assays may be read in parallel on a 96- or 384-well plate in the TECAN Safire2 fluorimeter. To test this system, it has been applied it to several cell lines of known BCL-2 dependence, and have obtained excellent correlation between results of mitochondrial BH3 profiling and cellular BH3 profiling. Two examples may be seen in Figure 20.

OTHER EMBODIMENTS

0168While the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.