Synthetic antibodies to BAX and uses thereof

Abstract

Synthetic fragment antigen-binding (Fab) antibodies are disclosed that bind to an N- terminal activation site of BCL-2-associated X-protein (BAX) and inhibit BAX activation. Also disclosed are methods of using the Fabs for measuring inactive monomeric BAX levels, screening for small molecules that bind to an N-terminal activation site of BAX, inhibiting apoptotic cell death, and predicting the ability of a cancer therapy to promote apoptotic cell death.

Description

<DP=1>²- 1 -²SYNTHETIC ANTIBODIES TO BAX AND USES THEREOF²CROSS-REFERENCE TO RELATED APPLICATION²

[0001] This application claims the benefit of U.S. Provisional patent ²application No. ²62 / 216,400, filed on September 10, 2015.²STATEMENT OF GOVERNMENT INTEREST²

[0002] This invention was made with government support under grant numbers ²HL095929, ²CA178394 and CA155472 awarded by the National Institutes of Health.²BACKGROUND OF THE INVENTION²

[0003] Throughout this application various publications are referred to by ²number in ²parentheses. Full citations for these references may be found at the end of ²the specification. ²The disclosures of these publications, and all patents, patent application ²publications and books ²referred to herein to more fully describe the art to which the subject ²invention pertains.²

[0004] Apoptosis plays a critical role in maintaining normal tissue ²homoeostasis in ²multicellular organisms and its deregulation results in an imbalance of ²homeostasis ²contributing to several diseases (1,2) The BCL-2 protein family plays a ²central role in ²regulating the mitochondrial pathway of apoptosis (3,4) The mitochondrial ²outer membrane ²permeabilization (MOMP) is considered the key event that is regulated by the ²complex network ²of protein-protein interactions between pro-apoptotic and anti-apoptotic ²members of the BCL-²2 family. Activation of pro-apoptotic members BAX and / or BAK is required for ²induction of ²MOMP, whereas the anti-apoptotic or survival proteins such as BCL-2, BCL-XL ²and MCL-1, ²inhibit the pro-apoptotic proteins and prevent MOMP. Activation of BAX and BAK ²or ²inhibition of anti-apoptotic BCL-2 proteins is regulated by direct interaction ²with the BH3-²only proteins.²

[0005] The activation pathway of BAX represents the gateway to apoptosis ²and ²understanding the function of BAX and its regulation mechanisms is an area of ²intensive ²investigation. BAX is predominantly found in the cytosolic compartment in an ²inactive ²conformation (5,6). Upon cellular stress, BAX is triggered and undergoes a ²series of ²conformational changes that enable its translocation to the mitochondrial ²membrane and²Date Recue / Date Received 2023-12-15²<DP=2>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-2-²oligomerization leading to MOMP induction (7,8). The structure of the BAX ²monomer in the ²inactive conformation was previously determined by nuclear magnetic resonance ²(NMR) ²spectroscopy (9). The inactive BAX structure adopts a typical BCL-2 fold, ²consisting of nine ²a-helices linked with variable loops. In contrast to BAK and anti-apoptotic ²BCL-2 proteins ²that reside at the mitochondrial outer membrane, the structure of BAX was ²determined with ²its hydrophobic C-terminal helix a9 bound to the canonical hydrophobic groove. ²When the ²C-terminal a9 helix dissociates from the canonical hydrophobic groove, it ²binds to the ²mitochondrial outer membrane facilitating the mitochondrial translocation of ²BAX (9). ²Structural analysis of a hydrocarbon stapled BIM BH3 helix bound to monomeric ²BAX ²uncovered an activation site at the N-terminal surface of BAX (10). This ²activation site ²regulates the trigger mechanism for conformational activation of cytosolic BAX ²leading to ²the release of the hydrophobic a9 helix and exposure of the hydrophobic a2 ²helix (BH3 ²domain) (10-12). Mitochondrial translocated BAX undergoes further ²conformational changes ²on the membrane that induce BAX oligomerization and MOMP (13-16), or is ²inhibited by ²anti-apoptotic Bc1-2 proteins (17-20).²100061 A ²number of diseases and disorder are associated with premature or unwanted cell²death and characterized by abnormal activation of BAX. The present invention ²addresses the ²need for inhibitors of BAX activation for therapeutic treatments.²SUMMARY OF THE INVENTION²

[0007] The ²present invention discloses synthetic fragment antigen-binding (Fab)²antibodies that bind to an N-terminal activation site of BCL-2-associated X-²protein (BAX) ²and inhibit BAX activation. Also disclosed are methods of using the Fabs for ²measuring ²inactive monomeric BAX levels, screening for small molecules that bind to an N-²terminal ²activation site of BAX, inhibiting apoptotic cell death, and predicting the ²ability of a cancer ²therapy to promote apoptotic cell death.²BRIEF DESCRIPTION OF THE DRAWINGS²100081 Fig. ²1A-1D. Discovery of synthetic antibody fragments that bind with high²affinity to BAX at overlapping binding sites. (A) Synthetic antibody fragments ²(Fabs) that ²bind to BAX with the corresponding variable sequences of complementarity ²determining ²regions (CDRs) and half-maximal binding titers (EC50) as determined by ELISA. ²(B) Top ²panel: Binding affinity (KD), association rate constant (ka) and dissociation ²rate constant (kd) ²for select synthetic Fabs that bind to BAX as determined by biolayer ²interferometry. ELISA ²binding EC5os are shown for comparison. Bottom panel: A representative ²biolayer²<DP=3>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-3-²interferometry experiment using immobilized 3G11 and three different BAX ²concentrations ²(C) Synthetic Fabs, except 2A5, compete biotinylated 3G11 (b3G11) from binding ²to BAX ²using a competitive ELISA. (D) Binding of selected phage-bound expressed Fabs ²(4)) to ²BAX as competitively inhibited by the corresponding free synthetic Fab ²proteins (100 nM). ²All synthetic Fabs, except 2A5, show at least 40% competition efficacy to ²different phage-²bound expressed Fabs. Data shown in (B-D) represent mean SD from at least ²three ²independent experiments. In (A), SEQ ID NOs for CDRL3, CDRH1, CDRH2, and CDRH3 ²²are respectively, SEQ ID NO:1-4 for Fab 2B1, SEQ ID NO:5-8 for 3E8, SEQ ID ²NO:9-12 for ²3G11, SEQ ID NO:13-16 for 2D9, SEQ ID NO:17-20 for 3G9, SEQ ID NO:21-24 for ²2C11, ²SEQ ID NO:25-28 for 3H4, SEQ ID NO:29-32 for 2A6, SEQ ID NO:33-36 for 3H1, SEQ ²ID ²NO:37-40 for 2A2, SEQ ID NO:41-44 for 3G3, SEQ ID NO:45-48 for 2D5, SEQ ID ²NO:49-²52 for 2D2, and SEQ ID NO:53-56 for 2A5.²100091 Fig. ²2A-2F. Synthetic Fabs inhibit BAX activation induced by pro-apoptotic tBID²in liposomal assay. (A) Synthetic BAX-binding Fobs or a VEGF specific Fab ²(YADS1) at 2 ²1.1M have no capacity to induce BAX-mediated liposomal ANT / DPX release while ²tBID ²induced potent BAX-mediated liposomal release. (B) BAX-binding Fabs at 2 tiM, ²except ²2A5 and YADS1, inhibit tBID-induced BAX-mediated liposomal ANT / DPX release. ²Bars ²represent mean SD value at 90 min that is normalized to tBID-induced maximum ²²ANTS / DPX release at 90 min (C-F). Representative liposomal ANTS / DPX release ²experiments in kinetic representation showing the inhibitory activity of Fabs ²3G11, 2B1 and ²3E8 at 0.5 p.M, 1 M and 2 EIM Fab whereas 2A5 had no inhibitory effect. ²Experiments ²performed with 400 nM BAX, 30 nM tBID and up to 2 0 / 1 Fabs (A-F). Data shown ²in (A-F) ²represent mean SD from triplicates normalized using either tBid-induced BAX ²mediated ²ANTS / DPX release as 100% (A-B) or 1% triton release as 100% (C-F). Data shown ²are ²representative of at least three independent experiments.²100101 Fig. ²3A-3B. Synthetic Fab 3G11 inhibit BAX-mediated cytochrome c release and²BAX mitochondria] translocation induced by pro-apoptotic tBID in isolated ²mitochondria. ²(A) Fab 3G11 inhibits dose-responsively tBID-induced BAX-mediated cytochrome c ²release ²from isolated BAK.4" mitochondria. (B) 3G11 Fab inhibits dose-responsively ²tBID-induced ²BAX mitochondria] translocation in isolated BAK4- mitochondria. Data shown are ²²representative of at least three independent experiments.²100111 Fig. ²4. 3G11 binds to the N-terminal surface of BAX. % deuterium incorporation²of unbound BAX conformation was determined in solution. HXMS analysis suggests ²²increased rates of deuterium exchange in the N-terminal region of BAX, the al -²a2 loop and²<DP=4>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-4-²helices a2, part of a5, a8, and a9. The relative difference of % deuterium ²incorporation of ²BAX conformation bound to 3G11 minus the % deuterium incorporation of BAX ²conformation alone was also determined. HXMS analysis suggests increased ²protection from ²deuterium incorporation in residues of helices al, a6, part of a5 and the al-²a2 loop compared ²to the unbound BAX. The regions of significant protection from deuterium ²incorporation (> - ²20%) are highlighted in dark on the sequence of BAX (SEQ ID NO:57). The ribbon ²²representation of the full length BAX structure (PDB ID; 16F6) is shown in the ²bottom ²portion of the figure.²100121 Fig. 5A-5F. Select BAX mutants inhibit functional inhibition and ²binding to BAX ²by 3G11. (A-D) Representative liposomal ANTS / DPX release experiments in ²kinetic ²representation showing the inhibitory activity of 3G11 (2 p,M) with BAX WT ²that is ²weakened by the IC21E mutation and completely abolished by the R134E or the ²double ²mutation R134E / K21E. Experiments performed with 400 nM BAX WT or BAX mutants, ²30 ²nM tBID and 2 KM 3G11. (E) % Inhibition based on the maximum tBID-induced BAX ²activation for BAX WT and mutants at 90 min in presence or absence of 3G11 ²Fab. (F) ²ELISA binding profiles for 3G11 binding to BAX WT and mutants. Half-maximal ²binding ²titers (EC50) were determined as follows: 8 2 nM for BAX WT, 29 4 for ²IC21E, 89 2 ²nM for R134E mutant and 91 2 nM for the R134E / IC21E double mutant. Data ²shown in ²(A-D) represent mean SD from triplicates and at least two independent ²experiments.²DETAILED DESCRIPTION OF THE INVENTION²100131 The invention provides a synthetic fragment antigen-binding (Fab) ²antibody that ²binds to an N-terminal activation site of BCL-2-associated X-protein (BAX) and ²inhibits ²BAX activation. Preferably, the antibody maintains BAX in its inactive, ²monomeric form. ²The antibody can bind to residues of helices al and a6, and to residues of al-²a.2 loop of ²BAX. Preferably, the antibody binds to BAX with a half-maximal binding (EC50) ²affinity of ²2nM-70nM. Preferably, the antibody blocks the interaction of BAX with a BAX ²activating ²partner, such as, for example, tBID, PUMA, BIM and / or NOXA.²

[0014] In different embodiments, the Fab is selected from the group ²consisting of²2B1 Fab, having a CDRL3 region comprising the amino acid sequence ²QYSGSGHYLI (SEQ ID NO:1), a CDRH1 region comprising the amino sequence IYSSSM ²(SEQ ID NO:2), a CDRH2 region comprising the amino acid sequence SISSSSSYTS ²(SEQ ²ID NO:3) and a CDRH3 region comprising the amino acid sequence ²RGYWYYWAWWASAMD (SEQ ID NO:4);²<DP=5>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-5-²3E8 Fab, having a CDRL3 region comprising the amino acid sequence QSSYSLI ²(SEQ ID NO:5), a CDRH1 region comprising the amino sequence LSYYSM (SEQ ID ²NO:6), a CDRH2 region comprising the amino acid sequence SISPYYGYTY (SEQ ID ²NO:7) and a CDRH3 region comprising the amino acid sequence ²RGGAYYFGYYGSGSYAMD (SEQ ID NO:8);²3G11 Fab, having a CDRL3 region comprising the amino acid sequence QWSFGPI ²(SEQ ID NO:9), a CDRH1 region comprising the amino sequence ISYYSM (SEQ ID ²NO:10), a CDRH2 region comprising the amino acid sequence SIYPYSSSTY (SEQ ID ²NO:11) and a CDRH3 region comprising the amino acid sequence RSSAMD (SEQ ID ²NO:12);²2D9 Fab, having a CDRL3 region comprising the amino acid sequence QWSHYLI ²(SEQ ID NO: 13), a CDRH1 region comprising the amino sequence LYYYSM (SEQ ID ²NO: 14), a CDRH2 region comprising the amino acid sequence SISPSYGYTS (SEQ ID ²NO:15) and a CDRH3 region comprising the amino acid sequence RSSFYYYALD (SEQ ²ID ²NO:16);²3G9 Fab, having a CDRL3 region comprising the amino acid sequence ²QHYYYSPWPI (SEQ ID NO:17), a CDRH1 region comprising the amino sequence ²LYSYYI (SEQ ID NO:18), a CDRH2 region comprising the amino acid sequence ²SISPYYSSTY (SEQ ID NO:19) and a CDRH3 region comprising the amino acid ²sequence ²RSSYSYAGMD (SEQ ID NO:20);²2C11 Fab, having a CDRL3 region comprising the amino acid sequence QSYVSPI ²(SEQ ID NO:21), a CDRH1 region comprising the amino sequence ISSYYI (SEQ ID ²NO:22), a CDRH2 region comprising the amino acid sequence SISSYYSSTY (SEQ ID ²NO:23) and a CDRH3 region comprising the amino acid sequence ²RVSYGHAYVGYSSGMD (SEQ ID NO:24);²3H4 Fab, having a CDRL3 region comprising the amino acid sequence QSWYYSYPI ²(SEQ ID NO:25), a CDRH1 region comprising the amino sequence LSYSSM (SEQ ID ²NO:26), a CDRH2 region comprising the amino acid sequence SISSYYSYTS (SEQ ID ²NO:27) and a CDRH3 region comprising the amino acid sequence RYYGYGGGID (SEQ ²ID ²NO:28);²2A6 Fab, having a CDRL3 region comprising the amino acid sequence QSAGGYPLI ²(SEQ ID NO:29), a CDRH1 region comprising the amino sequence IYYSSM (SEQ ID ²NO:30), a CDRH2 region comprising the amino acid sequence SISPYSSYTS (SEQ ID²<DP=6>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-6-²NO:31) and a CDRH3 region comprising the amino acid sequence RSFGYGWAFD (SEQ ²ID ²NO:32);²3H1 Fab, having a CDRL3 region comprising the amino acid sequence QHSYPI ²(SEQ ID NO:33), a CDRHI region comprising the amino sequence ISYSSI (SEQ ID ²NO:34), a CDRH2 region comprising the amino acid sequence SIYSYSGSTY (SEQ ID ²NO:35) and a CDRH3 region comprising the amino acid sequence RYGAMD (SEQ ID ²NO:36);²2A2 Fab, having a CDRL3 region comprising the amino acid sequence QYYYPI ²(SEQ ID NO:37), a CDRH1 region comprising the amino sequence ISSSSI (SEQ ID ²NO:38), ²a CDRH2 region comprising the amino acid sequence SIYSYYGYTY (SEQ ID NO:39) ²and ²a CDRH3 region comprising the amino acid sequence RYSAMD (SEQ ID NO:40);²3G3 Fab, having a CDRL3 region comprising the amino acid sequence ²QGAWSGGHLI (SEQ ID NO:41), a CDRH1 region comprising the amino sequence ²LSYSSM (SEQ ID NO:42), a CDRH2 region comprising the amino acid sequence ²YISPYYGYTY (SEQ ID NO:43) and a CDRH3 region comprising the amino acid ²sequence ²RGWAYYYGYWGPSGLD (SEQ ID NO:44);²2D5 Fab, having a CDRL3 region comprising the amino acid sequence ²QWGYSHSHLI (SEQ ID NO:45), a CDRH1 region comprising the amino sequence ISYSS ²(SEQ ID NO:46), a CDRH2 region comprising the amino acid sequence SISPYYGSTY ²(SEQ ID NO:47) and a CDRH3 region comprising the amino acid sequence RSHFGALD ²(SEQ ID NO:48);²2D2 Fab, having a CDRL3 region comprising the amino acid sequence ²QSYYWVSPF (SEQ ID NO:49), a CDRH1 region comprising the amino sequence LYYSSI ²(SEQ ID NO:50), a CDRH2 region comprising the amino acid sequence SIYPYSGSTY ²(SEQ ID NO:51) and a CDRH3 region comprising the amino acid sequence RSYGYAMD ²(SEQ ID NO:52); and²2A5 Fab, having a CDRL3 region comprising the amino acid sequence ²QYHYWYYPI (SEQ ID NO:53), a CDRH1 region comprising the amino sequence ISYYSM ²(SEQ ID NO:54), a CDRH2 region comprising the amino acid sequence SISPYYGSTY ²(SEQ ID NO:55) and a CDRH3 region comprising the amino acid sequence RAGAMD ²(SEQ ²ID NO:56).²100151 The ²complete heavy and light chain variable domain amino acid sequences for the²14 Fabs are shown below.²1) 3E8²<DP=7>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-7-²heavy chain, 242 residues²EVQLVESGGGLVQPGGSLRLSCAASGFNLSYYSMHWVRQAPGKGLEWVASISPYYG ²YTYYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARGGAYYFGYYGSGS ²YAMDYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVS ²WNS GALTSGVHTF PAVLQS SGLYS LS SVVTVPS SSLGTQTYICNVNHKPSNTKVDKK ²VEPKSCDKTHTCHHHHHH (SEQ ID NO:58)²light chain, 225 residues²DIQMTQSPSSLSASVGDRVTITCRASQSVSSAVAWYQQKF'GKAPKLLIYSASSLYSGV ²PSRFSGSRSGTDFTLTISSLQPEDFATYYCQQSSYS¨LITFGQGTKVEIKRTVAAPSV ²FIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNS QES V TEQDSKD ST ²YSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:59)²2) 3G3²heavy chain, 240 residues²EV Q LV ES GGGLV Q P GGS LRL S C AAS GFNL S YS S MHWVRQ AP GKGLEWV AYIS P YYG ²YTYYADSVKGRFTISADTSICNTAYLQMNSLRAEDTAVYYCARGWAYYYGYWGPS-²-GLDYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSW ²NSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVE ²PKSCDKTHTC HHHHHH (SEQ ID NO:60)²light chain, 228 residues²DIQMTQ SPS SL S AS VGDRVTITCRAS QSV S SAVAWYQQKPGKAPKLLIYSAS SLY S GV ²PSRFSGSRSGTDF1LTISSLQPEDFATYYCQQGAWSGGHLITFGQGTKVEIKRTVAAPS ²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS ²TYSLS STLTLSKADYEKHKVYACEVTHXGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:61)²3) 3G9²heavy chain, 234 residues²EVQLVESGGGLVQPGGSLRLSCAASGFNLYSYYIHMTVRQAPGKGLEWVASISPYYSS ²TYYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARSSYSYA------GM ²DYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSG ²ALTS GVHTFPAVLQS S GLY SL S SV V TVP S S S LGT QTYICNVNHKP SNTKVDKKVEPKS ²CDKTHTC HHHHHH (SEQ ID NO:62)²light chain, 228 residues²DIQMTQSPSSLSASVGDRVTITCRASQSVSSAVAWYQQKPGKAPKWYSASSLYSGV²PSRF SGS RS GTDF TLTIS SLQPEDFATYYCQQHYYYSPWPI _______________________ IF ²GQGTKVEIKRTV AAP S²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS ²TYSL S STLTL SKADYEKHKVYAC EV THQGLS SPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:63)²4) 3G1I²heavy chain, 230 residues²EVQLVESGGGLVQPGGSLRLSCAASGFNISYYSMHWVRQAPGKGLEWVASIYPYSSS ²TYYAD SVKGRFTI S ADTS KNTAYL QMN SLRAEDTAVYYC ARS S-------AMD ²YWGQGTLVTV SSASTKGPSVFPLAPS SKSTSGGTAALGCLVKDYFPEPVTV SWNS GA²<DP=8>²CA 02998115 2018-03-08²WO 2017 / 044308 _______________________________________________________ ²PCT / US2016 / 048508²-8-²LTS GVH ______________________________________________________________ 11: ²PAVL QS SGLYSLS SVVTVPSSSLGTQTYICNVNHKP SNTKVDKKVEPKSC²DKTHTCHHHHHH (SEQ ID NO:64)²light chain, 225 residues²DIQMTQ SP S SLS AS VGDRVTITCRASQ SVS SAVAWYQQKPGICAPICLUYSAS SLY S GV ²PSRFSGS RS GTDFTL TIS SL QPEDF ATYYCQQW SFG¨PITF GQ GTKV EIKRTV AAP SV ²FIF PP S D S QLKS GTAS VV C LLNNFYP REAKV QW KVDNAL Q S GN S QES V TE Q D SKD ²ST ²YSLSSTLTLSICADYEICHKVYACEVTHQGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:65)²5) 3H1²heavy chain, 230 residues²EVQLVESGGGLVQPGGSLRLSCAASGFNISYSSIHWVRQAPGKGLEWVASIYSYSGS ²TYYADSVKGRF'TISADTSKNTAYLQMNSLRAEDTAVYYCARYG--------AM ²DYWGQGTLVTVS SA STKGP S VFPLAP S SKSTS GGTAALGCLVICDYFP EPVTV SWN SG ²AL TS GVHTFPAVLQS SGLYSLSSVVTVPSS SLGTQTYICNVNHICP SNTKVDICKV EP KS ²CDKTHTCHHHHHH (SEQ ID NO:66)²light chain, 224 residues²DIQ MTQ SP S S LS AS VGDRVTITCRA S Q SVS SAVAWYQQ1CPGICAPICLLIYSAS SLYSGV²PSRFSGSRSGTDF'TLTISSLQPEDFATYYCQQHSY---PITFGQGTKVEIKRTVAAPSV²FIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESV _____________________ ²FEQDSICDST²YSLSSTLTLSICADYEICHICVYACEV'THQGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:67)²6) 3H4²heavy chain, 234 residues²EVQLV ES GGGLVQPGGSLRL SC AAS GFNL SYS S MHWVRQAPGKGLEWVASISSYYS ²YTSYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARYYGYGG---------G ²IDYWGQ GTLV TVS SAS TKGPSVFPLAPS SKS TS GGTAAL GCLV KDYF PEPVTV SWNS ²GALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPK ²SCDKTHTC HHHHHH (SEQ ID NO:68)²light chain, 227 residues²DIQMTQSPSSLSASVGDRVTITCRASQSVSSAVAWYQQICPGICAPKWYSASSLYSGV ²PSRFSGSRSGTDFTLTISSLQPEDFATYYCQQSWYYSY¨PITFGQGTKVEIKRTVAAPS ²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSICDS ²TYSLS S TLTL SICADYEKHKVYAC EV THQ GLS SPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:69)²7) 2A2²heavy chain, 230 residues²EVQLVESGGGLVQPGGSLRLSCAASGFNISSSSIHWVRQAPGKGLEWVASIYSYYGY ²TYYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARYS---------AMD ²YWGQ GT LVTV SSASTKGPSVFPLAPS SKS TSGGTAALGC LVKDYFPEPVTV SWNS GA ²LTS GVHTFPAVLQS SGLYSLS SVV'TVPSSSLGTQTYICNVNHKP SNTKVDICKVEPKSC ²DKTHTC HI-11-1HHH (SEQ ID NO:70)²light chain, 224 residues²<DP=9>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-9-²DIQMTQSPS SLS ASVGDRVTITCRAS QSV S SAVAWYQQKPGICAPKWYS AS SLY S GV ²PSRFSGSRSGTDFTLTISSLQPEDFATYYCQQYYY--PITFGQGTKVEIKRTVAAPSV ²FIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNS QESVTEQDSKD ST ²YSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:71)²8) 2A5²heavy chain, 230 residues²EVQLVESGGGLVQPGGSLRLSCAASGFNIYSSSMHWVRQAPGKGLEWVASISSSSSY ²TSYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCAR¨AG--AMDYWGQG ²TLV TV S S ASTKGP S VFP LAP S S KSTS GGTAALGC LVICDYFP EPV TV S'WN S GALT S ²GVH ²TFPAVLQSSGLYSLSSVVTVP SSSLGTQTYICNVNHICPSNTKVDIC.KVEPKSCDKTHTC ²HHHHHH (SEQ ID NO:72)²light chain, 227 residues²DIQMTQSPS S LS ASVGDRVTITCRASQSVS SAVAWYQQKPGKAPKLLIYS AS SLY SGV ²PSRFSGSRSGTDFTLTISSLQPEDFATYYCQQYHYWYY¨PITFGQGTKVEIKRTVAAPS ²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS ²TYSLSSTLTLSICADYEKHKVYACEVTHXGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:73)²9) 2A6²heavy chain, 234 residues²EV QLVES GGGLVQPGGSLRL SCAAS GFNIYYS S MHWVRQ AP GKGLEWVASISPYS SY ²TSYADSVKGRFTISADTSICNTAYLQMNSLRAEDTAVYYCARSFGYGW--------AF ²DYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVICDYFPEPVTVSWNSG ²ALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKP SNTKVDICKVEPKS ²CDKTHTC HHHHHH (SEQ ID NO:74)²light chain, 227 residues²DIQMTQSPSSLSASVGDRVTITCRASQSVSSAVAWYQQKF'GKAPICLLIYSASSLYSGV ²PSRFSGSRSGTDFTLTISSLQPEDFATYYCQQSAGGYP¨LITFGQGTKVEIKRTVAAPS ²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS ²TYSLSSTLTLSKADYEICHKVYACEVTHXGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:75)²10) 2B1²heavy chain, 239 residues²EV QLVESGGGLVQPGGS LRLSCAAS GFNIYS S SMHVVVRQ AP GKGLEWVAS IS SS S SY²TSYADSVKGRFTISADTSICNTAYLQMNSLRAEDTAVYYCARGYWYYWAWWAS--²AMDYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVICDYFPEPVTVSW²NS GALTS GVH __________________________________________________________ IF ²PAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVE²PKSCDKTHTC HHHHHH (SEQ ID NO:76)²light chain, 228 residues²DIQMTQ SPS SLS ASVGDRVTITCRAS QSV S SAVAWYQQ1CPGKAPKWYSAS SLY S GV²PSRFSGSRSGTDF __ ILTISSLQPEDFATYYCQQYSGSGHYLITFGQGTKVEIKRTVAAPS ²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS ²TYSLSSTLTLSICADYEICHKVYACEVTHQGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:77)²<DP=10>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-10-²11) 2C11²heavy chain, 240 residues²EVQLVES GGGLVQPGGSLRL S CAAS GFNI S S YYIHWVRQAPGKGLEWVAS I S SYYSS ²TYYAD SVKGRFTI S ADTS KNTAYLQMNS LRAEDTAVYYC ARV SYGHAYVGY S S---G ²MDYWGQGTLV TV S S AS TKGP S VFP LAP S S KS TS GGTAALG CLVKDYFP EPVTV S WN ²SGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSN'TKVDKKVEP ²KSCDKTHTC HHHHHH (SEQ ID NO:78)²light chain, 225 residues²DIQ MTQ S PS S LS AS VGDRVTITCRASQSVS SAV AWYQQKP GKAPKLLIYS AS SLY S GV ²PSRFSGSRSGTDFTLTISSLQPEDFATYYCQQSYVS---PITFGQGTKVEIKRTVAAPSV ²FIFPP SD S QLKS GTASVVC LLNNFYPREAKVQWKVDNALQ S GNS QESV TEXD SKI) ST ²YSL S STLTLSKADYEKHKVYACEVTHXGL SSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:79)²12) 2D2²heavy chain, 232 residues²EV QLV ES GGGLVQPGGS LRL SCAASGFNLYYS S IHWVRQAPGKGLEWVAS IYPY S GS ²TYYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARSYGY---------AM ²DYWGQGTLVTVS SASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSG ²ALTS GVHTFP AV LQ S SGLYSL SSVVTVPSS SLGT QTYIC NVNHKP SNTKV DKKV EP KS ²CDKTHTC HHHHHH (SEQ ID NO:80)²light chain, 227 residues²DIQMTQSPSSLSASVGDRVTITCRASQSVSSAVAWYQQKPGKAPKWYSASSLYSGV ²PSRFSGSRSGTDFTLTISSLQPEDFATYYCQQSYYWVS¨PFTFGQGTKVEIKRTVAAPS ²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS ²TYSLS STLTL SKADYEKHKVYACEVTHQGLS SPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:81)²13) 2D5²heavy chain, 232 residues²EVQLV ES GGGLVQPGGS LRL S CAAS GFNI SY S SIHWVRQAPGKGLEWVASISPYYGS ²TYYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARSHFG-----------ALD ²YWGQ GTLV TV S S ASTKGP SVFPLAP S SKS TSGGTAALGC LVKDYFP EPVTV SWNS GA ²LTS GVHTFPAVLQS SGLYSLS SVVTVPSSSLGTQTYICNVNHKP SNTKVDKKVEPKSC ²DKTHTC HHHHHH (SEQ ID NO:82)²light chain, 228 residues²DIQ MTQ SP S SL S AS VGDRVTITCRAS Q SV S SAV AWYQQKP GKAPKLLIYS AS SLY S GV²PSRFSGSRSGTDF __ LLTISSLQPEDFATYYCQQWGYSHSHLITFGQGTKVEIKRTVAAPS ²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS ²TYSLSSTLTLSKADYEKHKVYACEVTHXGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:83)²14) 2D9²heavy chain, 234 residues²EV QLV ES GGGLVQPGGS LRL S CAAS GFNLYYYS MHWVRQAPGKGLEWVA S IS P SYG²YTSYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARSSFYYY----AL²<DP=11>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-11-²DYW GQ GTLVTV S S ASTKGP S VFPL AP S SKSTS GGTAAL GCLVKDYFP EPV TV SWN S G²AL TS GVHTFP AVLQS SGLYSL S SVVTVPSS SLGTQTYICNVNHKPSNTKVD²KKVEPKSCDKTHTC HHHHHH (SEQ ID NO:84)²light chain, 225 residues²DIQMTQ SPS S L S AS VGDRVTITCRAS QSV S SAVAVVYQQKP GKAPKLLIYS AS SLY S GV ²P SRF S GS RS GTDF TL TIS S L QPEDFATYYCQQWS HY---LITF GQ GTKVEIKRTVAAP S ²VFIFPPSDSQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS ²TYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGECGGSDYKDDDDK ²(SEQ ID NO:85).²

[0016] The Fabs bind to and inhibit BAX.²

[0017] The ²Fabs can be produced reproducibly in large amounts using bacterial²expression or mammalian expression systems.²100181 The ²invention further provides a method of inhibiting apoptotic cell death by²contacting BCL-2-associated X-protein (BAX) with any of the synthetic Fab ²antibodies ²disclosed herein in an amount effective to inhibit activation of BAX and ²apoptotic cell death. ²The synthetic Fab antibody can be conjugated to a therapeutic agent and / or to ²an agent that ²facilitates transport across a cell membrane. The synthetic Fab antibody can ²administered to ²a subject using any known method effective to administer an amount of the ²antibody that ²inhibits activation of BAX, including, for example, lentiviral or adenoviral ²expression, a ²plasmid, nanoparticles or liposomes.²

[0019] The ²synthetic Fab antibody can be administered to a subject having, for example,²a disease or disorder selected from the group consisting of a cardiovascular ²disease or ²disorder (e.g., arteriosclerosis, heart failure, heart transplantation, ²aneurism, chronic ²pulmonary disease, ischemic heart disease, hypertension, thrombosis, and / or²cardiomyopathies), a neurodegenerative or neurological disease or disorder ²(e.g.,²Alzheimer's disease, Parkinson's disease, Huntington's disease, retinitis ²pigmentosa, spinal ²muscular atrophy, various forms of cerebellar degeneration, and / or amyotrophic ²lateral ²sclerosis), a liver disease or disorder, a kidney disease or disorder, a ²metabolic disease or ²disorder, an immunological disorder (e.g., organ transplant rejection, ²arthritis, lupus, IBD, ²Crohn's disease, asthma, multiple sclerosis, and / or diabetes), ischemia (e.g., ²stroke, ²myocardial infarction and / or reperfusion injury), infertility, a blood ²disorder (e.g., fanconi ²anemia, aplastic anemia, thalassemia, congenital neutropenia, and / or ²myelodysplasia), renal ²hypoxia, diabetes, hepatitis, asthma and AIDS.²

[0020] Also ²provided is a method for measuring inactive monomeric BAX levels in a²tissue sample from a subject comprising contacting the sample with any of the ²synthetic Fab²<DP=12>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-12-²antibodies disclosed herein and measuring the amount of antibody bound to the ²sample, ²wherein the amount of bound antibody is indicative of the level of inactive ²monomeric BAX ²in the tissue sample.²

[0021] ²Further provided is a method for screening for small molecules that bind to an ²N-²terminal activation site of BAX comprising contacting BAX with any of the ²synthetic Fab ²antibodies disclosed herein in the presence and in the absence of a small ²molecule and ²measuring the amount of antibody that binds to BAX, wherein a decrease in the ²amount of ²antibody binding to BAX in the presence of the small molecule indicates that ²the small ²molecule is a candidate for binding to the N-terminal activation site of BAX. ²It is also ²possible that the small molecule binds to a site on BAX that causes allosteric ²conformational ²changes that cause a decrease in the amount of antibody binding to BAX.²

[0022] The ²method can comprise a competitive ELISA screening assay. In an²embodiment of the methods described herein, the methods are useful for ²identifying ²therapeutic cell death inhibitors. In an embodiment of the methods described ²herein, the ²methods are useful for identifying therapeutic cell death activators.²

[0023] In an ²embodiment of the methods described herein, the small molecule has a²molecular weight of 2000 daltons or less. In an embodiment of the methods ²described herein, ²the small molecule has a molecular weight of 1500 daltons or less. In an ²embodiment of the ²methods described herein, the small molecule has a molecular weight of 1000 ²daltons or less. ²In an embodiment of the methods described herein, the small molecule has a ²molecular ²weight of 800 daltons or less. In an embodiment of the methods described ²herein, the small ²molecule has a molecular weight of either 2000, 1500, 1000, 800, 700, 600, 500 ²or 400 ²daltons or less. In an embodiment of the methods described herein, the small ²molecule is a ²small organic molecule.²

[0024] Still ²further provided is a method of predicting the ability of a cancer therapy to²promote apoptotic cell death comprising contacting a cancerous tissue sample ²with any of the ²synthetic Fab antibodies disclosed herein, with and without the cancer ²therapy, and ²measuring the amount of antibody that binds to the tissue sample, wherein a ²decrease in the ²amount of bound antibody with the cancer therapy indicates that the cancer ²therapy promotes ²activation of BAX and apoptotic cell death, and the amount of bound antibody ²without the ²cancer therapy can predict the capacity of the cancer therapy to promote ²activation of BAX ²and apoptotic cell death in the cancerous tissue.²<DP=13>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-13-²

[0025] In the ²methods disclosed herein, the synthetic Fab antibody can be labeled with a²detectable label, such as a fluorescent label or a radioactive label. The ²antibody can be ²biotinylated.²

[0026] As ²used herein, "BAX" is BCL-2-associated X-protein. In an embodiment, the²BAX is mammalian. In a preferred embodiment, the BAX is a human BAX. In an ²embodiment, the BAX comprises consecutive amino acid residues having the ²following ²sequence:²MDGSGEQPRGGGPTSSEQIMKTGALLLQGFIQDRAGRMGGEAPELALDPVPQDASTKKLSECLKRIGD ²ELDSNMELQRMIAAVDTDSPREVF FRVAADMESDGNENWGRVVALFYFASKLVLKALCTKVPEL I RT I ²MGW TLDFLRERLLGW I QDQGGWDGLL SYFGTPTWQTVT FVAGVLTASLTIWKKMG (SEQ ID NO:57, ²human).²

[0027] All ²combinations of the various elements described herein are within the scope of²the invention unless otherwise indicated herein or otherwise clearly ²contradicted by context.²100281 This ²invention will be better understood from the Experimental Details, which²follow. However, one skilled in the art will readily appreciate that the ²specific methods and ²results discussed are merely illustrative of the invention as described more ²fully in the claims ²that follow thereafter.²EXPERIMENTAL DETAILS²Introduction²100291 ²Synthetic antibody technology was applied to BCL-2 family proteins to probe²novel conformations and functional regions on the surface of BAX. Synthetic ²antibody ²discovery utilizes diversity targeted to the antibody complementarity ²determining regions ²(CDRs) that is encoded by designed, synthetic oligonucleotides and screened by ²phage or ²yeast display (22,23). Since library construction and selection is performed ²entirely in vitro, ²the state of the target can be controlled through the antibody discovery ²process. Therefore, ²the specificity of the output antibodies can be tuned to user-specified ²stringency (24,25). ²Using a phage display screen, synthetic antibodies were identified that bind ²with high affinity ²and specificity to BAX. Although BAX can be expressed and purified readily, ²and studied by ²NMR and other biophysical techniques, it is prone to aggregation under some ²conditions ²(such as detergents). This aggregation phenomenon is presumably linked to its ²function of ²forming oligomeric pores in the mitochondrial membrane, but presents a ²formidable ²challenge in raising conformation-specific antibodies using techniques ²requiring ²immunization. Thus, there was a strong rationale for utilizing in vitro ²selection.²<DP=14>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-14-²

[0030] ²Antibodies were identified that can be used as structural and biochemical ²probes²to dissect key regulatory mechanisms and conformations of BAX (26,27). ²Fourteen novel ²synthetic antibody fragments (Fabs) were identified that specifically target ²BAX. These ²synthetic Fabs have no significant homology in the CDR sequences suggesting a ²diversity of ²molecular interactions with BAX. They bind to BAX with nanomolar affinities ²and occupy ²overlapping binding sites on BAX. Additionally, the synthetic Fabs inhibit BAX ²in assays ²using liposomal membranes or isolated mitochondria. Further analysis using ²isolated ²mitochondria suggest that the Fabs bind to cytosolic BAX and inhibit its ²ability to translocate ²and insert onto the mitochondria outer membrane. Structural studies using NMR ²and ²hydrogen / deuterium exchange mass spectrometry showed that a representative ²synthetic Fab ²(3G11) forms a stoichiometric and stable complex with monomeric and inactive ²BAX, with a ²binding site that involves residues of helices al / a6 and of the al-a2 loop. ²Therefore, binding ²of the synthetic Fabs overlaps with the N-terminal activation site of BAX ²suggesting a novel ²mechanism of BAX inhibition through direct competition with the BAX activation ²process. ²These Fabs provide new tools for probing BAX activity in an unparalleled ²manner and ²provide a strategy for therapeutic inhibition of BAX in disease.²Materials and Methods²100311 ²Production of Recombinant BAX - Human full-length wild type BAX, truncated²a9 BAX (BAX AC) and BAX mutants in pTYB1 vector (New England Biolabs) were ²expressed in BL21 CodonPlus (DE3)-RIPL + E. coli strain and purified as ²previously ²described (9, 10).²

[0032] Fab ²protein expression - The phage display vectors for the 14 identified clones²were converted to Fab protein expression vectors by insertion of a stop codon ²and a 6-²Histidine tag upstream of the P3 gene fusion. Fab proteins were expressed ²periplasmically in ²E. co / i BL21(DE3) (New England Biolabs, Ipswich, MA) by growth in low-²phosphate media ²at 30 C for 18-22 h. The cells were harvested by centrifugation and lysed by ²using the ²BugBuster lysis reagent according to the manufacturer's instructions (Novagen, ²Madison, ²WI). The lysate was subjected to centrifugation, and the supernatant was ²applied to a nickel ²column (Ni-NTA resin, Qiagen, Valencia, CA) pre-equilibrated with Tris-²buffered saline ²(IBS), pH 7. The Fab-bound Ni-NTA column was then washed with 20-column volume ²1'13S ²buffer with 20 mM imidazole. Next, the protein was eluted with TBS buffer with ²250 mM ²imidazole. The eluent was dialyzed into PBS at pH 8.0, and applied to a ²Protein A affinity ²column (beads from Pierce Thermo Scientific, Rockford, IL) for further ²purification. The²<DP=15>²-15-²Fab-bound beads were washed with PBS, pH 8.0 (15 column volumes), and the Fab ²proteins ²were eluted with 100 mM glycine, pH 2.0. The eluted Fab proteins were ²immediately ²neutralized to ¨ pH 7 using 1 M Tris buffer, pH 8.0. Fractions containing the ²Fab proteins were ²buffer-exchanged into PBS, pH 7.0 and were used directly in following binding ²and activity ²assays, or flash-frozen and stored at 80 C for future use The Fab protein ²concentrations were ²determined by measuring the absorbance at 280 nm. The extinction coefficients ²at 280 nm were ²calculated from the Fab protein sequences using ExPASy.²100331 Origin of sequence of the CDR regions - The antibody framework of ²all the BAX ²antibodies is a common scaffold based on HerceptinTM (mAb 4D5), which is a ²humanized ²variant of a mouse anti-Her2 monoclonal antibody developed by Genentech (South ²San ²Francisco, CA). The complementarity determining regions (CDRs), which bind the ²antigen ²and provide the mAb specificity, contain wholely "synthetic" (artificial) ²sequences. The nature ²of these synthetic CDR sequences reflects the engineered nature of the phage ²antibody library ²from which the antibodies were obtained: Koellhoffer et al., Chembiochem, ²2012, 13, 2549.²

[0034] Sequences of antibodies expressed in cells - The BAX antibodies ²were converted to ²single chain variable fragments (scFvs) and cloned into cell expression ²plasmids (discussed ²below). DNA sequences for the scFv open reading frames are indicated below for ²different ²Fabs:²3E8, SEQ ID NO:86²GGATCCGACATCCAGATGACCCAGTCCCCAAGCTCCCTGAGCGCATCCGTGGGCGATAGGGTGACCATCA ²CAT GCAG G GCAT C T CAGA GCGT G TC TAGC GCAGT GGCATGGTACCAG CAGAAGC CAGG CAAG ²GC CC C TAA ²GCTGCTGATCTACAGCGC CTCCTCTCTGTATTCCGGAGTGCCTTCTCGGTTCTCCGGCAGCCGGAGCGGA ²ACC GACT TTACCCTGACAATCAGCTCCCTGCAGCCAGAGGATTTCGCCACATACTATTGCCAGCAGTCTA ²GC TAC TC C CTGA T CACCT T TGGC CAGGGCACAAAGG TG GAGA T CAAG GGAGGAGGCAG ²CGGAGGAGG C T C ²CGGAGGCGGCTCTGAGGT GCAGC TGGTGGAGAGCGGAGGAGGACTGGTGCAGCCTGGAGGCAGCCTGAGG ²CTGTCCTGTGCAGCATCT GGCTTCAACCTGTCTTACTATAGCATGCACTGGGTGCGCCAGGCACCAGGCA ²AGGGCCTGGAGTGGGTGGCCTCCATCTCTCCCTACTATGGCTACACCTACTATGCCGACTCTGTGAAGGG ²C C G GT TCACAAT CAGC GC C GATACC TC CAAGAACACAG CC TAT CTGCAGAT GAATAGC ²CTGAGGGCAGAG ²GACACCGCAGTGTACTAT TGTGCCAGAGGCGGCGCCTACTAT T TTGGCTACTATGGCAGCGGCTCCTACG ²CCATGGATTATTGGGGCCAGGGCACCCTGGTGACAGTGTCCTCTTAATCTAGA²3G11, SEQ ID NO:87²GGATCCGACATC CAGATGACC CAGAGC C CAAG C T CC C T GAGC G CATC C GTGGGC GATAGGGT ²GACCAT CA ²CAT GCAG G GCAT C T CAGA GCGT G TC TAGC GCAGT GGCATGGTACCAG CAGAAGC CAGG CAAG ²GC CC C TAA ²GCTGCTGATCTACTCCGC CTCCTCTCTGTATAGCGGCGTGCCTTCCCGGTTCTCCGGCAGCCGGAGCGGA ²AC C GACT T TACC C T GACAATCAG CT CC C T GCAGC C T GAGGAT T TC GC CACATAC TAT ²T GC CAGCAGT GGA ²GCTTCGGCCCAATCACCT TTGGCCAGGGCACAAAGGTGGAGATCAAGGGAGGAGGCTCTGGAGGAGGCAG ²CGGAGGCGGCTCCGAGGT GCAGC TGGTGGAGT CCGGCGGCGGC CTGG TGCAGCCAGGAGGCT CTCTGAGG ²CTGAGCTGTGCCGCCTCC GGCT T CAACATCTC CTACTATTCTATGCACTGGGTGCGCCAGGCACCAGGCA ²AGGGCCT GGAGT GGGTGG CCTCCATCTACCCC TATTCTAGCT C CACC TACTATGCCGACTCT GTGAAGGG ²²C C G GT TTACAAT C T CT GC C GATACCAGCAAGAACACAG CC TAC CT GCAGAT GAATAGC CT ²GAGGGCAGAG ²GACACCGCAGTGTACTAT TGTGCCAGATCTAGCGCCATGGATTATTGGGGCCAGGGCACCCTGGTGACAG ²TGTCCTCTTAATCTAGA²Date Recue / Date Received 2023-03-20²<DP=16>²-16-²3H1, SEQ ID NO:88²GGATCCGACATCCAGATGACCCAGAGCCCAAGCTCCCT GAGC G CATC C GTGGGCGATAGGGT GACCAT CA ²CAT GCAGGGCATCTCAGAGCGTGTCTAGCGCAGTGGCATGGTACCAGCAGAAGCCAGGCAAGGCCCCTAA ²GCT GCTGATCTACAGCGC CTCCT CTCTGTATTCCGGAGTGCCT TCTCGGTTCTCCGGCAGCCGGAGCGGA ²ACC GACT T TACCCTGACAATCAGCTCCCTGCAGCCAGAGGATT TCGCCACATACTATT GCCAGCAGCACT ²C C TACCC CATCAC C TT TG GCCAG GGCACAAAG GT GGAGATCAAGGGAGGAGGCAGCGGAGGAGGCT C ²C GG ²AGGCGGC T CTGAGGTGCAGCTGGTGGAGTCCGGAGGAGGACTGGTGCAGCCTGGAGGCAGCC TGAGGCTG ²TCC TGTGCAGCATCTGGC TTCAACATCTCTTACTCTAGCATCCACTGGGTGCGCCAGGCACCAGGCAAGG ²GCC TGGAG TGGG T GGC CT CTATC TACTCCTATTCTGGCAGCACCTACTATGCCGACAGCGTGAAGGGCCG ²GT T TACAATCAG C GCC GA TAC C T ²CCAAGAACACAGCCTATCTGCAGATGAATTCCCTGAGGGCAGAGGAC ²AC C GCAG T GTAC TATT GT GCCAGATACGGCGCCATGGATTATT GGGGCCAGGGCACCC TGGT GACAG ²T GT ²CC T CTTAATCTAGA²2B1, SEQ ID NO:89²GGATCCGACATCCAGATGACCCAGTCCCCAAGCTCCCT GAGC G CATC C GTGGGCGATAGGGT GACCAT CA ²CAT GCAGGGCATCTCAGAGCGTGTCTAGCGCAGTGGCATGGTACCAGCAGAAGCCAGGCAAGGCCCC TAA ²GC T GCTGATCTACAGCGC CTCCT CTCTGTATAGCGGCGTGCCATCCCGGTTCTCCGGCAGCCGGAGCGGA ²ACC GACT T TACCCTGACAATCAGCTCCCTGCAGCCCGAGGATT TCGCCACATACTATT GCCAGCAGTACT ²CCGGCTCT GGCCACTATC TGATCACCTTTGGCCAGGGCACAAAGGTGGAGATCAAGGGAGGAGGCTC TGG ²AGGAGGCAGCGGAGGCGGCTCCGAGGTGCAGC TGGTGGAGTCCGGCGGCGGCCTGGTGCAGCCTGGAGGC ²TCT CTGAGGCTGAGCT GT GCAGCATCCGGCTTCAACAT CTAC T CTAGCTCCATGCACT GGGTGCGCCAGG ²CAC CAGG CAAGG GC CTGGAGTGG GTGGC CAGCAT CTCTAGCT C CTCTAGCTACACCTC ²TTATGCCGACAG ²CGT GAAGGGCCGGTTTACAATCT CCGCCGATACCTCTAAGAACACAGCCTATCTGCAGATGAATTCCCTG ²AGGGCAGAGGACACCGCAGTGTACTATTGTGCCAGAGGCTAC T GGTACTATTGGGCCT GGTGGGCCAGCG ²CCATGGAT TATT GGGGCCAGGGCACCCTGGTGACAGTGTCCTC TTAATCTAGA²2D9, SEQ ID NO:90²GGATCCGACATCCAGATGACCCAGAGCCCAAGCTCCCT GAGC G CATC C GTGGGCGATAGGGT GACCAT CA ²CAT GCAGGGCAT CTCAGAGCGTGTCTAGCGCAGTGGCATGGTACCAGCAGAAGCCAGGCAAGGCCCC TAA ²GC T GCTGATCTACAGCGC CTCCT CTCTGTATTCCGGAGTGCC T TCTCGGTTCTCCGGCAGCCGGAGCGGA ²AC C GACT T TACCCTGACAATCAGCTCCCTGCAGCCAGAGGATT TCGCCACATACTATT GCCAGCAGTGGT ²CC CACTAT CTGATCACCT TTGGCCAGGGCACAAAGGTGGAGAT CAAGGGAGGAGGCAGCGGAGGAGGCTC ²CGGAGGCGGCTCTGAGGT GCAGC TGGTGGAGTCCGGAGGAGGACTGGTGCAGCCTGGAGGCAGCCTGAGG ²CTGTCCT GTGCAGCATCT GGCTT CAACCTGTACTATTACTCTATGCACTGGGTGCGCCAGGCACCAGGCA ²AGGGCCT GGAGT GGGTGG C CTC CATCT C T CCCAGCTAC GGCTATACCAGCTAC GCCGACTCC GT ²GAAGGG ²CCGGTTCACAATC TCT GC CGATACCAGCAAGAACACAGCCTAT CTGCAGAT GAATTCC CTGC GGGCC GAG ²²GACACCGCCGTGTATTAC TGTGCCAGATCTAGCTTTTATTAC TATGCCCTGGATTACT GGGGCCAGGGAA ²CCC TGGTGACAGTGTCCT CTTAATCTAGA .²

[0035] ²Cloning and procedures for in cell expression - The characterized mAbs were²cloned to mammalian expression vectors (pCDNA3.1 or pMSCY PIG: Niro TRES GFP) ²in a ²reduced Fab form constructed by linear two copies of single chain variable ²Fragments (scFv) ²to ensure proper folding and expression. pCDNA3.1 is available from Addgene ²(Cambridge, ²MA) and GenScript (Piscataway, NJ) used it to clone the Fab sequences. ²Memorial Sloan ²Kettering Cancer Center (New York, NY) subcloned the Fab sequences from ²pCDNA3.1 to ²pMSCV. To easily detect the expression of Fabs and use them in pull-down ²experiments, 6 ²Myc tag sequences were included as an N-terminal tag to each scFv sequence ²during the ²cloning process. Transient expression of five Fabs in Mouse Embryo Fibroblasts ²(MEF) ²demonstrated that all Fabs are well expressed at similar levels_ For more ²efficient transfection, ²Fab plasmids were transduced with retroviral transfection and stable Fab-²Date Reeue / Date Received 2023-03-20²<DP=17>²-17-²expressing clones were selected using GFP as a selection marker. Co-²immunoprecipation ²experiments using anti-my c agarose beads pulled down the Fabs and ²demonstrated their ²capacity to bind BAX in cells using anti-BAX Western analysis.²100361 Direct binding ELISAs - BAX (1 jig per well) was first immobilized ²on 96-well ²EIA / RIA plate (Coming Incorporated, Corning, NY) at room temperature for 1 h ²or at 4 C for ²overnight. PBS containing 3% BSA was used to block the wells after BAX ²immobilization ²(incubation for 1.5 h at room temperature). The Fabs were diluted into PBS ²buffer (pH 7.0), ²applied to the wells, and incubated for 1 h at room temperature. The plates ²were then washed ²with PBS and incubated for 1 h with horseradish peroxidase / anti-FLAG M2 ²antibody conjugate ²(binding to the FLAG tag at the C-terminus of Fab light chain). The wells were ²washed with ²PBS, developed with 3,3',5,5'-tetramethylbenzidine (TMB) substrate, and ²quenched with 0.5 ²M H2504. The absorbance at 450 nm was determined. The data were fit to a ²standard four ²parameter logistic equation by using GraphPad Prism (GraphPad Software, La ²Jolla, CA). The ²half-maximal binding (ECH) values were obtained from the inflection point of ²the curve.²100371 Competitive BindingELISAs - Six Fab clones (3H1, 3G11, 3E8, 2A5, ²2D9 and 2B1) ²were chosen for the competition ELISAs. Two different approaches were used In ²the first ²approach, non-biotinylated Fabs were used to compete with the binding of ²biotinylated 3G11 ²(b3G11). To immobilized BAX on the 96-well EIA / RIA plate (Corning ²Incorporated, Corning, ²NY), a mixture of 50 nM b3G11 and increasing amount of each of the six ²selected Fabs were ²added. After washes, only the binding of b3G11 was monitored using horseradish ²²peroxidase / streptavidin conjugate and TMB substrate. In the second approach, ²purified Fab ²proteins were used to compete with Fabs displayed at the phage surface. ²Similarly, to ²immobilized BAX on the 96-well EIA / RIA plate (Corning Incorporated, Coming, ²NY), a ²mixture of Fab-displayed phage (with a titer of ¨1012 pfu / mL) and 100 nM of ²each of the six ²selected Fab proteins were added. Only the binding of Fab-phage complex ²binding was ²monitored using horseradish peroxidase / anti-M13 antibody conjugate and TMB ²substrate.²100381 Biolayer Inteiferometry - The forteBio BLItz system (Pall ²Corporation, Menlo ²Park, CA) was used to determine the binding kinetics and affinity between BAX ²and Fab ²proteins. Ni-NTA biosensors (Pall Corporation, Menlo Park, CA) were used for ²initial Fab ²protein immobilization, which was followed by BAX association and dissociation ²interaction ²analysis. 10 j.tg / mL Fab was used for immobilizing at pH 7. For each Fab ²protein, at least three ²different BAX concentrations, varying from 30 nM to 1.2 M, were used, and ²subsequently ²global fitting was used to generate the ka (association rate constant), kd ²(dissociation rate ²constant) and ICD (equilibrium dissociation constant) values.²Date Recue / Date Received 2023-03-20²<DP=18>²-18-²

[0039] Size-exclusion Chromatography - Superdex 75 10 / 300 GL and 200 ²10 / 300 GL (GE ²Healthcare) columns were used for size exclusion chromatography of recombinant ²BAX and ²Fab-BAX complex. Proteins were injected in columns equilibrated with a buffer ²containing 20 ²mM HEPES pH 72, and 150 mM KC1.²100401 Liposome Permeabilization Assay - Liposomes were composed of the ²following ²molar percentages of lipids (Avanti Polar Lipids): phosphatidylcholine, 48%; ²phosphatidylinositol, 10%; phosphafidylethanolamine, 28%; dioleoyl ²phosphatidylserine, ²10%; and tetraoleoyl cardiolipin, 4% and were loaded with ANTS / DPX (Molecular ²Probe) ²upon extrusion. Fabs at the indicated concentrations in 96-well format ²(Corning), with and ²without BAX (400 nM) or BAX (400nM) and tB1D (30 nM) and then Liposomes were ²added ²(10 pL from 1 mL lipid stock) in assay buffer (10 mM HEPES, pH 7,200 mM KC1, ²and 1 mM ²MgCl2) to a fmal volume of 100 1. ANTS / DPX release was quantified based on ²the increase ²in fluorescence intensity that occurs when the ANTS fluorophore is separated ²from the DPX ²quencher upon release from the liposomes into solution. Fluorescence (Xex = ²355 nm and Xem ²= 520 nM) was measured over time at 30 C using a Tecan Infinite M1000 plate ²reader. The ²percentage release of ANTS / DPX at 90 mM was calculated as percentage release = ²((F ¨ ²F0) / (F100 ¨ FO)) x 100, where FO and F100 are baseline and maximal ²fluorescence, ²respectively. 1% TritonTm treatment is used to determine the maximum amount of ²liposomal ²release per assay, and this value sets the 100% value for the kinetic curve ²figures whereas the ²calculated percent release value for the bar graphs was calculated by using ²the percent release ²value of tBid-induced BAX mediated ANT / DPX release as 100%.²100411 BAX trcmslocation assay - Mitochondria from liver of Balcl- mice ²(0.75mg / m1) were ²resuspended in experimental buffer (125 mM KCl, 10 mM Tris-MOPS [pH 7.41, 5 mM ²²glutamate, 2.5 mM malate, 1 mM KPO4, 0.1mM EGTA-Tris [pH 7.4]) and treated ²with the ²indicated concentrations of 3G11 recombinant BAX (200 nM) and Mid (100nM), ²singly and ²in combination, and incubated at room temperature for 20 min. The supernatant ²fractions were ²isolated by centrifugation at 5500xg for 10 min and the mitochonclrial pellets ²resuspended and ²washed with 0.1M sodium carbonate (pH 11.5) for 20 min, centrifuged at ²13,000xg for 10 min ²at 4 C, and then solubilized in 1% Triton X-100 / PBS for 1 h at 4 C. ²Mitochondrial supernatant ²and pellet fractions were separated by 4-12% NuPageTM (Invitrogen) gels and ²analyzed by ²immunoblotting with anti-BAX antibody (Cell Signaling Cat. 2772).²Date Recue / Date Received 2023-03-20²<DP=19>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-19-²

[0042] ²Mitochondrial release assay - Mitochondria from liver of Bail" mice (1.5mg / m1)²were resuspended in experimental buffer (125 mM KC1, 10 mM Tris-MOPS [pH 7.4], ²5 mM ²glutamate, 2.5 mM malate, 1 mM KPO4, 0.1mM EGTA-Tris [pH 7.4]) and treated ²with the ²indicated concentrations of 3G11 recombinant BAX (400 nM) and tBid (30nM), ²singly and in ²combination, and incubated at room temperature for 60 mM. The supernatants ²were isolated ²by centrifugation at 5500xg for 10 mM and the mitochondrial pellets ²solubilized in 1% Triton ²X-100 / PBS. Mitochondrial supematant and pellet fractions were separated by 4-²12% NuPage ²(Invitrogen) gels and analyzed by immunoblotting with anti-cytochrome c ²antibody (BD ²Biosciences Cat. 556433).²

[0043] ²Western Blotting - Samples from the mitochondria and translocation assay were²electrophoretically separated on 4-12% NuPage (Invitrogen) gels, transferred ²to mobilon-FL ²PVDF membranes (Millipore) and subjected to immunoblotting. For visualization ²of proteins ²with Odyssey Infrared Imaging System (LI-COR Biosciences) membranes were ²blocked in ²PBS containing 3% milk powder. Primary BAX antibody (Cell Signaling 2772S) was ²²incubated overnight at 4 C in a 1:1,000 dilution. After washing, membranes ²were incubated ²with an IRdye800-conjugated goat anti-rabbit IgG secondary antibody (LI-COR ²Biosciences) ²in a 1:10,000 dilution. Proteins were detected with Odyssey Infrared Imaging ²System.²

[0044] NMR ²samples and spectroscopy - Uniformly 15N-labeled full-length human BAX²was generated as previously described. Protein samples of BAX and BAX-3G11 ²mixtures at ²indicated concentrations were prepared in 25 mM sodium phosphate, 50 mM NaCl ²solution at ²pH 6.0 in 5% D20. Correlation 1H-15N HSQC and 1H-15N TROSY spectra were ²acquired ²at 25 C on a Brulcer 600 MHz NMR spectrometer equipped with a cryogenic probe, ²²processed using Topsin, and analyzed with CCPNMR. BAX wild type cross-peak ²assignments were applied as previously reported (9).²

[0045] ²Hydrogen / Deuterium exchange mass spectrometry - Prior to hydrogen-deuterium²exchange experiments, the quench condition for best sequence coverage of BAX ²was ²optimized as previously described (28). Briefly, 3 l of stock solution of BAX ²at 1.0 mg / rnl ²was mixed with 9 pl of H2O buffer (8.3 mM Tris, 150 mM NaCl, in H20, pH7.2) at ²0 C and ²then quenched with 18 pl of ice cold quench solutions of 0.8% formic acid, 16% ²Glycerol, ²and GdnHC1 at final concentrations of 0.05 M, 0.5 M, 1.0 M and 2.0 M. The ²quenched ²samples were frozen on dry ice and then subjected to an immobilized pepsin ²column ²(1x20mm, 30 mg / ml porcine pepsin (Sigma)) for online digestion for 40 sec. ²The resulting ²peptides were collected on a C18 trap (Michrom MAGIC C18AQ 0.2x2) and ²separated using ²a reversed phase C18 column (Michrom MAGIC C18AQ 0.2x50 3um 200A) with a 30 ²min²<DP=20>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-20-²linear gradient of 0.046% (v / v) trifluoroacetic acid, 6.4% (v / v) acetonitrile ²to 0.03% (v / v) ²trifluoroacetic acid, 38.4% (v / v) acetonitrile. The effluent was directed into ²an OrbiTtrap ²Elite mass spectrometer (Thermo-Fisher Sci. Inc) for MS analysis. The ²instrument was ²operated in positive ESI mode and the resolution of the instrument was set at ²60,000. ²Proteome Discoverer software (Thermo Fisher Sci. Inc.) was used to identify ²the sequence of ²the resulting peptides. The optimal quench condition with the best coverage ²map of BAX ²(0.08M GuHC1 in 0.8% formic acid) was used for subsequent functionally ²deuterated studies. ²Hydrogen-deuterium exchange reactions were initiated by diluting 3 1.11 of pre-²chilled protein ²stock solution (free BAX, 1 mg / ml, or antibody-bound BAX, 2mg / m1) into 9 ill ²D20 buffer ²(8.3 mM Tris, 150 mM NaC1, in D20, pDREAD 7.2). The samples were incubated at ²0 C ²for 10 sec, 100 sec and 1000 sec. The exchange reaction was terminated by the ²addition of ²18 pl of optimized quench solution at 0 C and samples were immediately frozen ²on dry ice ²and stored at -80 C. In addition, un-deuterated samples and equilibrium-²deuterated control ²samples were also prepared as previously described (29). The deuterated ²samples were then ²loaded onto above instrument for DXMS analysis. The centroids of the isotopic ²envelopes of ²un-deuterated, functionally deuterated, and equilibrium deuterated peptides ²were measured ²using HDExaminer, and then converted to corresponding deuteration levels with ²corrections ²for back-exchange (30).²100461 Structure calculations - Structure calculations were performed with ²Crystallography and NMR system solve (CNS) within the HADDOCK web server using ²the ²prediction interface (31). HADDOCK calculations generated models of the ²complex that are ²in agreement with experimental distance restraints and have optimal ²electrostatic and van der ²Waals interactions based on a combination of molecular dynamics and energy ²minimization. ²HADDOCK docking was performed using the BAX NMR structural ensemble imported ²directly from the PDB (PDB ID: 1F16). The 3G11 structural model was generated ²using ²PIGS software based on structural homology modeling of known Fab structures of ²the same ²family that have different amino acid composition of CDR regions (32). ²Calculations were ²performed with ambiguous interaction restraints (AIR) derived from the ²hydrogen deuterium ²exchange mass spectrometry data and residues of the CDR regions of 3G11. For ²BAX AIR ²calculations, only residues that exhibited significant protection of hydrogen ²deuterium ²exchange upon titration and solvent accessibility over 50% as determined by ²the program ²NACCESS (Hubbard and Thornton, 1993) were defined as active residues; passive ²residues ²were automatically assigned by the HADDOCK web interface as those residues ²surrounding ²the active residues. For 3G11, AIR restraints were assigned based on residues ²involved in the²<DP=21>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-21-²CDRH1, CDRH2, CDRH3, and CDRL3. Specifically, active residues for BAX included ²²residues 19-25 and 123-142; active residues for 3G11 were defined as Heavy ²Chain: 26-32 ²(Loop 1), 52-54 (Loop 2), 96-101 (loop 3) and Light Chain; 141-148, 165-168, ²and 206-211. ²In each HADDOCK structure calculation, 1000 orientations / structures of the ²complex were ²generated by rigid-body docking energy minimization of the individual ²structures. The 188 ²lowest energy structures were semi-flexibly refined in torsion angle space and ²then refined in ²explicit solvent. 132 structures formed the most populous cluster (HADDOCK ²score = -140.5 ²12.1 and Z-score = -1.7) with RMSD from the overall lowest energy structure ²(1.9 1.1 A) ²calculated on the backbone (CA, C, N, 0, P) atoms of all residues involved in ²intermolecular ²contact using a 10 A cut-off. In all cases 3G11 was localized to N-terminal ²trigger site. ²Ribbon diagrams and molecular models were depicted using PYMOL (Schrodinger).²Results²100471 ²Discovery of 14 Novel Synthetic Antibodies Targeting Monomeric BAX - A²restricted diversity phage antigen-binding fragment (Fab) library ("Library ²F") was used to ²select a panel of Fabs that bind to monomeric BAX (9). Library F was designed ²to contain ²mostly binomial Tyr / Ser diversity in CDR-H1 and H2, and expanded diversity, ²encoding the ²nine amino acids Tyr / Ser / Gly / Ala / Phe / Trp / His / Pro / Val in a 5 / 4 / 4 / 2 / 1 / 1 / 1 / 1 / 1 ²ratio, at CDR-²L3 and H3 (33). In addition, the CDR-L3 and H3 loops vary in size. The ²expanded diversity ²in CDR-L3 and H3 in Library F was designed to mimic the distribution of amino ²acids in ²functional CDR-H3 segments of natural antibodies. Library F was screened ²against the ²monomeric BAX, which that was purified after utilizing chitin-column affinity ²purification ²and size exclusion chromatography (10). After 2-3 rounds of stringent ²selection, a panel of 14 ²BAX-specific Fabs with diverse CDR sequences was obtained (Fig. 1A). These ²clones were ²isolated from two distinct panning regimes, one in which biotinylated BAX was ²immobilized ²onto streptavidin-coated wells, and another in which BAX was coated directly ²on the wells. ²The properties of the selected clones were similar from either screening ²regime and thus ²further analyzed as a single conglomerated panel. Sequence comparison between ²the 14 Fabs ²revealed high diversity of residues and loop sizes in many positions with no ²apparent ²sequence homology in CDR segments from any two clones, demonstrating the ²diverse modes ²of interaction with BAX.²100481 Fabs ²were expressed and purified and the half-maximal binding titers (EC50) for²BAX using ELISA were determined. The ECRI values ranged from high affinity 2.3 ²nM ²(2B1) to moderate affinity 250 nM (2A5) with most Fabs having single or double ²digit nM²<DP=22>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-22-²nanomolar EC50 (Fig. 1A). Select Fabs 3G11, 3E8, 2A5 and 2B1 were evaluated ²for binding ²kinetics to monomeric BAX using Biolayer Interferometry (Fig. 1B). The binding ²affinities ²(K0 values) and ranking of the Fab proteins based on binding correlate with ²the EC50 from ²ELISA. In both assays, Fabs 3G11, 3E8 and 2B1 were the highest-affinity ²binders to BAX, ²while Fab 2A5 was the lowest-affinity binder (Fig. 1A, 1B). Hence, an array of ²14 high-²affinity antibodies against monomeric BAX was identified.²100491 To ²explore if the Fabs engaged overlapping or distinct epitopes on BAX,²competitive ELISA was performed using six of the clones (3H1, 3G11, 3E8, 2A5, ²2D9 and ²2B1) whose EC50 values cover the entire affinity range. A competitive ELISA ²was performed ²in which Fab 3G11 was first biotinylated (b3G11) and binding of b3G11 to BAX ²in the ²presence of other non-biotinylated Fab proteins was analyzed. All Fabs except ²Fab 2A5 ²demonstrated competitive inhibition of b3G11 binding with IC5os ranging from ²nM to i.tM ²(Fig. 1C). Fab 2A5 does not compete b3G11 binding likely because it is a much ²weaker ²binder than 3G11. Furthermore, a study was also conducted of how binding of ²phage-²expressed Fabs (0) can be competitively inhibited by free Fab proteins. The ²data showed that ²phage-expressed Fab binding was inhibited to varying degrees in the presence ²of 100 nM Fab ²protein for many of the clones and that, in each case, the phage-expressed Fab ²for a particular ²clone could be inhibited by its own Fab protein (Fig. 1D). Again, Fab 2A5, the ²lowest affinity ²clone (EC50 = 250 nM) had little effect on Fab phage binding for all the other ²clones, but all ²the other Fabs proteins could significantly inhibit Fab 2A5 phage binding. On ²the contrary, ²Fabs 3E8 and 2B1, the highest affinity clones (EC5os = 2.6 nM and 2.3 nM, ²respectively) ²showed the strongest inhibition on all Fab phage binding. Therefore, the ²competitive ELISA ²data matches predictions based on ELISA half-maximal binding titers and KD ²values. These ²results indicate that, to a large extent, these Fabs have overlapping epitopes ²on the BAX ²surface with a range of affinities, but many in the low nanomolar range.²100501 ²Synthetic Antibodies Inhibit BH3-triggered BAX activation and MOM? ¨ To²examine how binding to BAX by Fabs proteins modulates BAX function, liposomal ²assays ²were performed that explicitly evaluate how BH3-triggered BAX forms a pore in ²a ²membrane environment made from similar lipid composition of mitochondria ²without the ²contribution of other mitochondria' factors (34). Selected Fabs have the ²potential to activate ²BAX by engaging one of the activation sites of BAX either at its N-terminal or ²C-terminal ²surface (10,15,35,36) or inhibit BAX activation by inhibiting the BAX binding ²surface of the ²activator protein tBID or conformational changes associated with BAX ²activation ²(12,15,16,20,37). Therefore, the capacity of the Fabs to either activate BAX ²or inhibit BAX²<DP=23>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-23-²activation induced by tBID was examined. None of the 14 Fabs had an effect on ²the ²liposomal integrity alone; neither does an unrelated vascular-endothelial ²growth factor ²(VEGF)-specific Fab (YADS1, negative control). Furthermore, none of the Fab ²proteins ²activated BAX and induced liposomal release (Fig. 2A). However, when tBID and ²BAX are ²combined with liposomes, liposomal release is robust as expected (Fig. 2A, ²2B). In contrast, ²high affinity Fabs when combined with tBID and BAX inhibited tBID-triggered ²BAX ²activation significantly, or completely, at 21.IM, in all cases, except with ²the exception of the ²lower affinity clone Fab 2A5 and the negative control YADS1 (Fig. 2B). Three ²of the highest ²binding Fabs 3G11, 3E8 and 2B1 yielded dose-responsive and time-dependent ²inhibition of ²liposomal release whereas the weaker binder 2A5 had no effect even at 21.1M ²(Fig. 2C-2F).²

[0051] To ²examine whether Fabs can inhibit BAX activation in the presence of²mitochondrial membranes loaded with anti-apoptotic BC1-2 or other ²mitochondrial proteins, ²mouse liver mitochondria were isolated from Balc- / - mice to perform a ²mitochondrial release ²assay. Combination of tBID and BAX was exposed to several doses of Fab 3G11, ²which ²demonstrated dose-responsive inhibition of mitochondrial cytochrome c release ²induced by ²activated BAX, as assessed by separation of the supernatant and mitochondrial ²fractions and ²western analysis (Fig. 3A). Moreover, using the mitochondria assay, BAX ²localization was ²determined to examine whether inhibition by Fab 3G11 was due to prevention of ²BAX ²translocation to the membrane or blocking of the membrane integration and ²oligomerization ²on the outer mitochondrial membrane. Fab 3G11 dose-responsively inhibited the ²capacity of ²BAX for mitochondrial translocation suggesting that 3G1I's high affmity ²binding to the ²monomeric BAX either competes binding of tBID or prevents conformational ²changes of ²BAX required for its mitochondrial translocation (Fig. 3B). Taken together, ²these data ²indicate that these Fab proteins inhibit BH3-triggered BAX activation and MOMP ²by ²restraining mitochondrial membrane translocation of BAX.²100521 Fab ²3G11 Protein Forms a Stable and Stoichiornetric Complex with BAX - To²investigate the effect of synthetic Fab proteins on the structure of BAX, the ²stability and ²stoichiometry of several Fab-BAX complexes were studied upon mixing the ²protein at ²various concentrations and analyzing by size-exclusion chromatography (SEC) ²(data not ²shown). Based on gel SEC and SDS-PAGE analysis, binding of the Fab 3G11 to BAX ²results ²in a stable 1:1 stoichiometric complex. 3G11 was selected for further ²structural analysis ²because of its high affinity and low dissociation constant (Fig. 1B). Next, ²structural effects of ²the 3G111 were analyzed based on 1H-15N HSQC spectra of full length BAX ²monomer at ²several doses of 3G11. The HSQC spectra are in agreement with the 3G11 and BAX ²forming²<DP=24>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-24-²a stable complex, with a 1:1 stoichiometry and in slow exchange on the NMR ²time scale as ²evidenced by the dose-dependent loss of intensity of the cross peaks of the ²monomeric BAX ²and lack of new chemical shifts upon 3G11 titration. Despite the 3G11-BAX ²complex ²formation evidenced by the NMR titration, it is not possible to observe the 1H-²15N cross ²peaks of the complex in the 1H-15N HSQC spectra due to the size of the complex ²(70 KDa).²

[0053] Fab ²3G11 Protein Binds the N-terminal Surface of BAX to Prevent BAX²Activation - To further analyze the Fab-BAX interaction, the changes on the ²BAX structure ²were analyzed by measuring the solvent accessibility and hydrogen-deuterium ²exchange of ²the backbone amide hydrogens using hydrogen-deuterium exchange mass ²spectrometry ²(HXMS). First, the deuterium exchange of unbound BAX was analyzed in solution, ²which ²underlined the different deuterium exchange rates for exposed or unfolded ²regions (N-²terminus, al-a2 loop) and more solvent-protected or structured regions ²including a2, a3-a4, ²a5, a6-a8. Upon 3G11-BAX complex formation at stoichiometric levels, deuterium ²²exchange, sample digestion, preparation and analysis was performed in the same ²conditions ²as with free BAX. Interestingly, HXMS analysis of the Fab-bound BAX in ²solution ²highlighted significant solvent protection in helix al and a6 and the al-a2 ²loop while other ²regions of the BAX structure had little to no change in the deuterium ²incorporation upon ²interaction with the 3G11. Furthermore, HXMS showed a modest increase in ²solvent ²accessibility for residues in helices a7, a8 and partially a9 and a2 that are ²found at the C-²terminal surface of BAX. Binding of the 3G11 Fab to full-length BAX (BAX WT) ²and to the ²C-terminal helix a9-truncated BAX (BAX AC) was determined with similar ²dissociation ²constant KD (Table 1), suggesting that major contacts of 3G11 occur elsewhere ²from the a9 ²or the canonical hydrophobic groove of BAX.²

[0054] The ²HXMS analysis suggests that the binding region of the 3G11 is localized on²the N-terminal surface of the BAX structure and overlaps with the 1) N-²terminal trigger site ²of BAX (al / a6) that controls a series of conformational changes upon BH3 ²domain ²activation (11-12) and 2) the binding epitope (residues 12-24) that is ²recognized by the 6A7 ²antibody only on the conformational active BAX (Fig. 4). Interestingly, the al-²a2 loop, ²whose displacement was determined to be essential for the initiation of ²conformational ²changes upon BH3-triggered BAX activation (12), is protected by the Fab ²binding ²interaction. Thus, the binding interaction of 3G11 to BAX is localized to the ²N-terminal ²surface of BAX, overlapping with the N-terminal trigger site and preventing ²conformational ²changes that lead to activation of the monomeric BAX.²<DP=25>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-25-²

[0055] ²Although recent structural studies based on peptides from the cytomegalovirus²protein vMIA (38) and BCL-2 (20) proteins suggested potential sites that ²intervene with ²BAX activation, interestingly, none of these peptides directly block the N-²terminal activation ²site of the soluble BAX or the C-terminal activation site of the mitochondrial ²associated ²BAX. A protein-protein structure calculation approach (39) was used to further ²confirm this ²novel interaction of 3011 bound to the N-terminal surface of BAX. HADDOCK ²structure ²calculations were performed using ambiguous interaction restraints between BAX ²residues ²determined by HXMS to be most protected from solvent upon 3G11 binding ²(residues of the ²al, a6 and the al-a2 loop) and residues of the CDR regions present in the 3011 ²protein ²sequence. Furthermore, calculations included an NMR structural ensemble of BAX ²with ²different loop conformations. Based on these calculations, the 3G11 binds more ²favorably in ²a direct interaction with N-terminal trigger site of BAX and the closed ²conformation of al-a2 ²loop (10,12). 3G11 protein makes interactions of hydrophobic and hydrophilic ²nature with ²the solvent-exposed hydrophobic residues and polar / charged residues of the BAX ²trigger site ²covering a large interface surface of (1640A).²

[0056] To ²validate the 3G11 direct interaction with the N-terminal trigger site, it was²tested whether mutations on BAX can disrupt binding to 3G11 and affect its ²inhibitory ²activity of 3G11 in liposomal ANTS / DPX release assays. The 3G11-BAX structural ²model ²was used. Residues on BAX were selected that form contacts with residues of ²3G11 using a ²3A cutoff. Residues K21 in helix al and R134 in helix a6 have previously been ²shown to ²interact with the stapled BIM BH3 peptide that activates BAX through the N-²terminal trigger ²site (10). Further analysis of the structural model showed that R134 residue ²forms hydrogen ²bonds with four residues (S30, Y31, Y32, S33) in 3011 CDR-H1 loop while the ²1(21 residue ²is predicted to form one hydrogen bond interaction with residue S143 in 3G11. ²Consistently ²with the predicted contributions of each residue to the interaction with 3G11 ²Fab, liposomal ²ANTS / DPX release experiments showed that the inhibition effect of 3G11 on BAX ²WT ²activation by tBID is weakened with the presence of the 1(21E mutation, but it ²is was ²completely abolished with the R134E mutation (Fig. 5A-5E). Likewise, the ²double mutation ²R134E / K21E also abolished the capacity of 3G11 to inhibit BAX activation (Fig. ²SD, 5E). ²The inability of 3011 to inhibit activation of these BAX mutants is consistent ²with the ²decreased affinity of 3G11 to these BAX mutants as determined by ELISA binding ²²experiments (Fig. SF). Thus, the data suggest that 3011 binds and blocks ²access to the N-²terminal trigger site of BAX and therefore directly prevents BAX activation ²and its ²associated conformational changes. In accordance, mapping the predicted BAX ²interacting²<DP=26>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-26-²residues based on the 3G11-BAX structural model and residues of the trigger ²site that interact ²with the stapled BIM BH3 peptide as determined by NMR studies demonstrates an ²extensive ²overlap (10). Taken together, these results suggest a novel mechanism of BAX ²inhibition and ²demonstrate the capacity of the reported Fabs as structural and functional ²probes of BAX.²Discussion²

[0057] ²Decision of cellular life or death through the mitochondrial apoptotic ²pathway, in²physiological or disease conditions, is mainly controlled by the interactions ²among the BCL1-²2 family proteins. Activation of pro-apoptotic BAX is essential for apoptosis ²to proceed ²through mitochondrial dysfunction and caspases activation. Here, synthetic ²antibody ²technology was harnessed to generate high affinity BAX binding antibody ²fragments (Fabs). ²The screen identified at least 14 different Fabs with sequence diversity in ²CDR regions that ²bind with nM affinity to BAX. Interestingly, the competitive ELISAs confirmed ²that the ²identified Fab proteins bind to overlapping regions of the BAX surface. ²Selected Fabs bind ²full length BAX, which represents cytosolic BAX (9). It was demonstrated that ²3G11 binds ²C-terminal truncated BAX, which mimics the mitochondrial-associated BAX (15). ²Mitochondrial inserted BAX oligomerizes and undergoes dramatic conformational ²changes ²of the monomeric BAX to such extent that the N-terminal surface of BAX is not ²available for ²binding to Fabs (13,15,16,40). Therefore, selected Fabs are expected to be ²conformational ²specific for cytosolic and mitochondrial-associated BAX and inhibit both ²conformations from ²proceeding along the BAX activation process.²100581 All of ²the identified BAX-binding Fabs inhibit BAX activation triggered by pro-²apoptotic tBID in liposomal and mitochondrial assays. Further investigation of ²the ²mechanism, using Fab 3G11protein, indicated that the Fabs bind to the N-²terminal surface of ²BAX without causing significant conformational changes on BAX. HADDOCK ²calculations ²and mutagenesis show that 3G11 binds to an extended surface on BAX that ²includes the N-²terminal trigger site (helices al / a6), which BIM, BID, and PUMA pro-apoptotic ²BH3 helices ²bind to trigger BAX activation (10,11,36,38,41). The data indicate that 3G11 ²competitively ²inhibits tBID-mediated BAX activation by blocking access to the N-terminal ²trigger site and ²preventing N-terminal conformational changes associated with BAX activation ²(11,36). ²Indeed, 3G1 1-binding prevents mitochondrial translocation of BAX, which ²requires ²significant conformational changes and integration into the membrane ²(13,15,16,40).²

[0059] BAX is ²shown to have two different activation sites depending on its cytosolic or²mitochondrial associated conformations; the N-terminal BH3 pocket (trigger ²site) and the C-²<DP=27>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-27-²terminal BH3 pocket. Although several proteins have been reported to directly ²bind BAX and ²inhibit its activation, only two other studies reported structural evidence of ²the binding ²interaction on the surface of BAX. A stapled helical peptide of the BH4 domain ²of BCL-2 ²protein binds to a cleft formed by select residues of the al -a2 loop, a2, a3, ²a5 and a6 (20). A ²helical peptide of the cytomegalovirus protein vMIA binds to a distinct site ²at the same side ²of the BAX structure that includes the loops of a3-a4 and a5-a6 (38). In both ²cases, the ²peptides bind to a geographically distinct site that have no overlap with ²either of the N-²terminal or C-terminal activation site. Therefore, these mechanisms of BAX ²inhibition reflect ²allosteric mechanisms that suppress conformational changes upon BAX ²activation. The ²present data indicate that the direct inhibition mechanism of BAX by 3G11 Fab ²through ²interaction with the N-terminal activation site is a feasible and effective ²mechanism of BAX ²inhibition.²100601 ²Pharmacological targeting of BAX whether to either promote or inhibit its²activation has been proposed to be a promising therapeutic strategy. However, ²discovery of ²pharmacological modulators of BAX has been challenging due to limited insights ²or lack of ²appropriate probes to use for small molecule discovery. Recently, small ²molecule BAX ²activators that bind to the BAX trigger site have been identified using a ²competitive binding ²assay of the stapled BIM BH3 peptide activator that binds to the same site ²(42). Likewise, the ²application of synthetic antibodies to BAX provides a novel opportunity to use ²Fabs as ²probes for screening small molecule libraries using a competitive binding ²assay between the ²identified inhibitory Fabs and BAX. Such small molecule inhibitors that bind ²directly to the ²BAX trigger site will be effective for development of therapeutics.²Table 1. Binding kinetics and thermodynamics of Fab 3G11 with BAX WT and BAX ²AC.²BAX kc,. (M-1 s-1) ______ koff (S-1) Kro (nM) ²BAX WT (3.0 1.5) x 105 (1.1 0.1) x 10'2 38 12²BAX (2.9 0.5) x 105 (1.8 0.1) x le 65 13 ²REFERENCES²1. Fuchs, Y., and Steller, H. (2011) Programmed cell death in animal ²development and disease. Cell 147, 742-758²2. Danial, N. N., and Korsmeyer, S. J. (2004) Cell death: critical control ²points. ²Cell 116, 205-219²<DP=28>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-28-²3. Youle, R. J., and Strasser, A. (2008) The BCL-2 protein family: opposing ²²activities that mediate cell death. Nat. Rev. Mol. Cell Biol. 9, 47-59²4. Chipuk, J. E., Moldoveanu, T., Llambi, F., Parsons, M. J., and Green, D. ²R. ²(2010) The BCL-2 family reunion. Mol. Cell 37, 299-310²5, Hsu, ²Y. T., Wolter, K. G., and Youle, R. J. (1997) Cytosol-to-membrane²redistribution of Bax and Bc1-X(L) during apoptosis. Proc. Nat. Acad. Sci. USA ²94, 3668-²3672²6. Wolter, K. G., Hsu, Y. T., Smith, C. L., Nechushtan, A., Xi, X. G., and ²Youle, ²R. J. (1997) Movement of Bax from the cytosol to mitochondria during ²apoptosis. J. Cell ²Biol. 139, 1281-1292²7. Walensky, L. D., and Gavathiotis, E. (2011) BAX unleashed: the ²biochemical ²transformation of an inactive cytosolic monomer into a toxic mitochondria! ²pore. Trends ²Bioch. Sci. 36, 642-652²8. Westphal, D., Kluck, R. M., and Dewson, G. (2014) Building blocks of the ²²apoptotic pore: how Box and Bak are activated and oligomerize during ²apoptosis. Cell Death ²Diff. 21, 196-205²9. Suzuki, M., Youle, R. J., and Tjandra, N. (2000) Structure of Box: ²coregulation of dimer formation and intracellular localization. Cell 103, 645-²654²10. Gavathiotis, E., Suzuki, M., Davis, M. L., Pitter, K., Bird, G. H., ²Katz, S. G., ²Tu, H. C., Kim, H., Cheng, E. H., Tjandra, N., and Walensky, L. D. (2008) BAX ²activation is ²initiated at a novel interaction site. Nature 455, 1076-1081²11. Kim, H., Tu, H. C., Ren, D., Takeuchi, 0., Jeffers, J. R., Zambetti, G. ²P., ²Hsieh, J. J., and Cheng, E. H. (2009) Stepwise activation of BAX and BAK by ²tBID, BIM, ²and PUMA initiates mitochondrial apoptosis. Mol. Cell 36, 487-499²12. Gavathiotis, E., Reyna, D. E., Davis, M. L., Bird, G. H., and Walensky, ²L. D. ²(2010) BH3-triggered structural reorganization drives the activation of ²proapoptotic BAX. ²Mol. Cell 40, 481-492²13. Gahl, R. F., He, Y., Yu, S., and Tjandra, N. (2014) Conformational ²rearrangements in the pro-apoptotic protein, Bax, as it inserts into ²mitochondria: a cellular ²death switch. J. Biol. Chem. 289, 32871-32882²<DP=29>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-29-²14. Lovell, J. F., Billen, L. P., Bindner, S., Shamas-Din, A., Fradin, C., ²Leber, B., ²and Andrews, D. W. (2008) Membrane binding by tBid initiates an ordered series ²of events ²culminating in membrane permeabilization by Bax. Cell 135, 1074-1084²15. Czabotar, P. E., Westphal, D., Dewson, G., Ma, S., Hockings, C., ²Fairlie, W. ²D., Lee, E. F., Yao, S., Robin, A. Y., Smith, B. J., Huang, D. C., Kluck, R. ²M., Adams, J. M., ²and Colman, P. M. (2013) Bax crystal structures reveal how BH3 domains ²activate Bax and ²nucleate its oligomerization to induce apoptosis. Cell 152, 519-531²16. Bleicken, S., Jeschke, G., Stegmueller, C., Salvador-Gallego, R., ²Garcia-Saez, ²A. J., and Bordignon, E. (2014) Structural model of active Box at the ²membrane. Mol. Cell ²56, 496-505²17. Sattler, M., Liang, H., Nettesheim, D., Meadows, R. P., Harlan, J. E., ²Eberstadt, M., Yoon, H. S., Shuker, S. B., Chang, B. S., Minn, A. J., ²Thompson, C. B., and ²Fesik, S. W. (1997) Structure of Bc1-xL-Bak peptide complex: recognition ²between ²regulators of apoptosis. Science 275, 983-986²18. Ding, J., Mang, Z., Roberts, G. J., Falcone, M., Miao, Y., Shao, Y., ²Zhang, X. ²C., Andrews, D. W., and Lin, J. (2010) Bc1-2 and Box interact via the BH1-3 ²groove-BH3 ²motif interface and a novel interface involving the BH4 motif J. Biol. Chem. ²285, 28749-²28763²19. Ding, K., Wang, H., Wu, Y., Zhang, L., Xu, J., Li, T., Ding, Y., Zhu, ²L., and ²He, J. (2015) Rapamycin protects against apoptotic neuronal death and improves ²neurologic ²function after traumatic brain injury in mice via modulation of the mTOR-p53-²Bax axis. J. ²Surg. Res. 194, 239-247²20. Barclay, L. A., Wales, T. E., Garner, T. P., Wachter, F., Lee, S., ²Guerra, R. ²M., Stewart, M. L., Braun, C. R., Bird, G. H., Gavathiotis, E., Engen, J. R., ²and Walenslcy, L. ²D. (2015) Inhibition of Pro-apoptotic BAX by a noncarionical interaction ²mechanism. Mol. ²Cell 57, 873-886²21. Czabotar, P. E., Lessene, G., Strasser, A., and Adams, J. M. (2014) ²Control of ²apoptosis by the BCL-2 protein family: implications for physiology and ²therapy, Nat. Rev.. ²Mol. Cell Biol. 15, 49-63²22. Sidhu, S. S., and Fellouse, F. A. (2006) Synthetic therapeutic ²antibodies. Nat. ²Chem. Biol. 2, 682-688²<DP=30>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-30-²23. Paduch, M., Koide, A., Uysal, S., Rizk, S. S., Koide, S., and ²Kossiakoff, A. A. ²(2013) Generating conformation-specific synthetic antibodies to trap proteins ²in selected ²functional states. Methods 60, 3-14²24. Fellouse, F. A., Esaki, K., Binalan, S., Raptis, D., Cancasci, V. J., ²Koide, A., ²Jhurani, P., Vasser, M., Wiesmann, C., Kossialcoff, A. A., Koide, S., and ²Sidhu, S. S. (2007) ²High-throughput generation of synthetic antibodies from highly functional ²minimalist phage-²displayed libraries. J. Mol. Biol. 373, 924-940²25. Koellhoffer, J. F., Chen, G., Sandesara, R. G., Bale, S., Saphire, E. ²0., ²Chandran, K., Sidhu, S. S., and Lai, J. R. (2012) Two synthetic antibodies ²that recognize and ²neutralize distinct proteolytic forms of the ebola virus envelope ²glycoprotein. Chembiochem ²13, 2549-2557²26. Koerber, J. T., Thomsen, N. D., Hannigan, B. T., Degrado, W. F., and ²Wells, ²J. A. (2013) Nature-inspired design of motif-specific antibody scaffolds. Nat. ²Biotech. 31, ²916-921²27. Gao, J., Sidhu, S. S., and Wells, J. A. (2009) Two-state selection of ²conformation-specific antibodies. Proc. Nat. Acad. Sci. USA 106, 3071-3076²28. Marsh, J. J., Guan, H. S., Li, S., Chiles, P. G., Tran, D., and Morris, ²T. A. ²(2013) Structural insights into fibrinogen dynamics using amide ²hydrogen / deuterium ²exchange mass spectrometry. Biochem. 52, 5491-5502²29. Li, S., Tsalkova, T., White, M. A., Mei, F. C., Liu, T., Wang, D., ²Woods, V. ²L., Jr., and Cheng, X. (2011) Mechanism of intracellular cAMP sensor Epac2 ²activation: ²cAMP-induced conformational changes identified by amide hydrogen / deuterium ²exchange ²mass spectrometry (DXMS). J. Biol. Chem. 286, 17889-17897²30. Zhang, Z., and Smith, D. L. (1993) Determination of amide hydrogen ²exchange by mass spectrometry: a new tool for protein structure elucidation. ²Prot. Science 2, ²522-531²31. de Vries, S. J., van Dijk, M., and Bonvin, A. M. (2010) The HADDOCK web ²²server for data-driven biomolecular docking. Nat. Prot. 5, 883-897²32. Marcatili, P., Olimpieri, P. P., Chailyan, A., and Tramontano, A. ²(2014) ²Antibody structural modeling with prediction of immunoglobulin structure ²(PIGS), Nat Prot. ²9, 2771-2783²<DP=31>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-31-²33. Persson, H., Ye, W., Wernimont, A., Adams, J. J., Koide, A., Koide, S., ²Lam, ²R., and Sidhu, S. S. (2013) CDR-H3 diversity is not required for antigen ²recognition by ²synthetic antibodies. J. Mol. Biol. 425, 803-811²34. Yethon, J. A., Epand, R. F., Leber, B., Epand, R. M., and Andrews, D. ²W. ²(2003) Interaction with a membrane surface triggers a reversible ²conformational change in ²Box normally associated with induction of apoptosis. J. Biol. Chem. 278, 48935-²48941²35, ²Edwards, A. L., Gavathiotis, E., LaBelle, J. L., Braun, C. R., Opoku-Nsiah, K.²A., Bird, G. H., and Walenslcy, L. D. (2013) Multimodal interaction with BCL-2 ²family ²proteins underlies the proapoptotic activity of PUMA BH3. Chem. & Bio1.20. 888-²902²36. Leshchiner, E. S., Braun, C. R., Bird, G. H., and Walenslcy, L. D. ²(2013) ²Direct activation of full-length proapoptotic BAK Proc. Nat. Acad. Sci. USA ²110, E986-995²37. Ding, J., Mooers, B. H., Zhang, Z., Kale, J., Falcone, D., McNichol, ²J., Huang,²B., Zhang, X. C., Xing, C., Andrews, D. W., and Lin, J. (2014) After embedding ²in ²membranes antiapoptotic Bcl-XL protein binds both Bc1-2 homology region 3 and ²helix 1 of ²proapoptotic Box protein to inhibit apoptotic mitochondrial permeabilization. ²J. Biol. Chem. ²289, 11873-11896²38. Ma, J., Edlich, F., Bermejo, G. A., Norris, K. L., Youle, R. J., and ²Tjandra, N. ²(2012) Structural mechanism of Box inhibition by cytomegalovirus protein yMIA. ²Proc. Nat. ²Acad. Sci. USA 109, 20901-20906²39. Dominguez, C., Boelens, R., and Bonvin, A. M. (2003) HADDOCK: a ²protein-protein docking approach based on biochemical or biophysical ²information. J. Amer. ²Chem. Soc. 125, 1731-1737²40. Annis, M. G., Soucie, E. L., Dlugosz, P. J., Cruz-Aguado, J. A., Penn, ²L. Z., ²Leber, B., and Andrews, D. W. (2005) Bax forms multispanning monomers that ²oligomerize ²to permeabilize membranes during apoptosis. EMBO J. 24, 2096-2103²41. Tsai, C. J., Liu, S., Hung, C. L., Thong, S. R., Sung, T. C., and ²Chiang, Y. W. ²(2015) BAX-induced apoptosis can be initiated through a conformational ²selection ²mechanism. Structure 23, 139-148²42. Gavathiotis, E., Reyna, D. E., Bellairs, J. A., Leshchiner, E. S., and ²Walenslcy, ²L. D. (2012) Direct and selective small-molecule activation of proapoptotic ²BAX. Nat. ²Chem. Biol. 8, 639-645²<DP=32>²CA 02998115 2018-03-08²WO 2017 / 044308 ²PCT / US2016 / 048508²-32-²43. ²Billard, C. (2012) Development of Noxa-like BH3 tnimetics for apoptosis-²based therapeutic strategy in chronic lymphocytic leukemia. Mol. Cancer Res. ²10, 673-676²

Claims

<DP=1>²-33-²What is claimed is:²1. A ²synthetic fragment antigen-binding (Fab) antibody that specifically binds to ²an²N-terminal activation site of BCL-2-associated X-protein (BAX) and inhibits ²BAX activation,²wherein the synthetic fragment antigen-binding (Fab) antibody is selected from ²the group ²consisting of²3E8 Fab, having a light chain comprising a CDR L1 region comprising the amino ²acid ²sequence SVSS (amino acids 28-31 of SEQ ID NO:59), a CDRL2 comprising the ²amino acid ²sequence YSASS (amino acids 49-53 of SEQ ID NO:59), and a CDRL3 region ²comprising the ²amino acid sequence QSSYSLI (SEQ ID NO:5), a CDRH1 region comprising the amino ²acid ²sequence LSYYSM (SEQ ID NO:6), a CDRH2 region comprising the amino acid ²sequence ²SISPYYGYTY (SEQ ID NO:7), and a CDRH3 region comprising the amino acid ²sequence ²RGGAYYFGYYGSGSYAMD (SEQ ID NO:8);²3G9 Fab, having a light chain comprising a CDR L1 comprising the amino acid ²sequence ²SVSS (amino acids 28-31 of SEQ ID NO: 63), a CDRL2 comprising the amino acid ²sequence ²YSASS (amino acids 49-53 of SEQ ID NO: 63), and a CDRL3 region comprising the ²amino acid ²sequence QHYYYSPWPI (SEQ ID NO:17), a CDRH1 region comprising the amino acid ²sequence LYSYYI (SEQ ID NO:18), a CDRH2 region comprising the amino acid ²sequence ²SISPYYSSTY (SEQ ID NO:19), and a CDRH3 region comprising the amino acid ²sequence ²RSSYSYAGMD (SEQ ID NO:20); and²2C11 Fab, having a light chain comprising a CDR Ll comprising the amino acid ²sequence SVSS (amino acids 28-31 of SEQ ID NO: 79), a CDRL2 comprising the ²amino acid ²sequence YSASS (amino acids 49-53 of SEQ ID NO: 79), and a CDRL3 region ²comprising the ²amino acid sequence QSYVSPI (SEQ ID NO:21), a CDRH1 region comprising the ²amino acid ²sequence ISSYYI (SEQ ID NO:22), a CDRH2 region comprising the amino acid ²sequence ²SISSYYSSTY (SEQ ID NO:23), and a CDRH3 region comprising the amino acid ²sequence ²RVSYGHAYVGYSSGMD (SEQ ID NO:24).²2. The synthetic Fab antibody of claim 1, wherein the antibody maintains BAX ²in its ²inactive, monomeric form.²Date Recue / Date Received 2023-12-15²<DP=2>²-34-²3. The synthetic Fab antibody of claim 1, wherein the antibody binds to ²residues of ²helices al and a6, and to residues of a1-a2 loop of BAX.²4. The synthetic Fab antibody of claim 1, wherein the antibody binds to BAX ²with a ²half-maximal binding (EC50) affinity of 2 nM-70 nM.²5. The synthetic Fab antibody of claim 1, wherein the antibody blocks the ²interaction of ²BAX with a BAX activating partner.²6. The synthetic Fab antibody of claim 5, wherein the BAX activating paitner ²is selected ²from the group consisting of tBID, PUMA, BIM and NOXA.²7. The synthetic Fab antibody of claim 1, wherein the synthetic Fab antibody ²is ²conjugated to a therapeutic agent.²8. The synthetic Fab antibody of claim 1, wherein the synthetic Fab antibody ²is ²conjugated to an agent that facilitates transport across a cell membrane.²9. The synthetic Fab antibody of claim 1, wherein the synthetic Fab antibody ²is labeled ²with a fluorescent label or a radioactive label.²10. A synthetic fragment antigen-binding (Fab) antibody that specifically ²binds to an N-²terminal activation site of BCL-2-associated X-protein (BAX) and inhibits BAX ²activation, ²wherein the synthetic fragment antigen-binding (Fab) antibody is selected from ²the group ²consisting of:²3H4 Fab, having a light chain comprising a CDR Ll region, a CDRL2 region, and ²a ²CDRL3 region and comprising the amino acid sequence of SEQ ID NO:69, wherein ²the CDRL3 ²region comprises QSWYYSYPI (SEQ ID NO:25), and a heavy chain CDRH1 region ²comprising ²the amino acid sequence LSYSSM (SEQ ID NO:26), a CDRH2 region comprising the ²amino ²acid sequence SISSYYSYTS (SEQ ID NO:27), and a CDRH3 region comprising the ²amino acid ²sequence RYYGYGGGID (SEQ ID NO:28);²Date Recue / Date Received 2023-12-15²<DP=3>²-35-²2A6 Fab, having a light chain comprising a CDR Ll region, a CDRL2 region, and ²a ²CDRL3 region and comprising the amino acid sequence of SEQ ID NO:75, wherein ²the CDRL3 ²region comprises QSAGGYPLI (SEQ ID NO:29), and a heavy chain CDRH1 region ²comprising ²the amino acid sequence IYYSSM (SEQ ID NO:30), a CDRH2 region comprising the ²amino ²acid sequence SISPYSSYTS (SEQ ID NO:31), and a CDRH3 region comprising the ²amino acid ²sequence RSFGYGWAFD (SEQ ID NO:32); and²3H1 Fab, having a light chain comprising a CDR Ll region, a CDRL2 region, and ²a ²CDRL3 region and comprising the amino acid sequence of SEQ ID NO:67, wherein ²the CDRL3 ²region comprises QHSYPI (SEQ ID NO:33), and a heavy chain CDRH1 region ²comprising the ²amino acid sequence ISYSSI (SEQ ID NO:34), a CDRH2 region comprising the amino ²acid ²sequence SIYSYSGSTY (SEQ ID NO:35), and a CDRH3 region comprising the amino ²acid ²sequence RYGAMD (SEQ ID NO:36).²11. The synthetic Fab antibody of claim 10, wherein the antibody maintains BAX ²in its ²inactive, monomeric form.²12. The synthetic Fab antibody of claim 10, wherein the antibody binds to ²residues of ²helices al and a6, and to residues of al-a2 loop of BAX.²13. The synthetic Fab antibody of claim 10, wherein the antibody binds to BAX ²with a ²half-maximal binding (EC5o) affinity of 2 nM-70 nM.²14. The synthetic Fab antibody of claim 10, wherein the antibody blocks the ²interaction ²of BAX with a BAX activating partner.²15. The synthetic Fab antibody of claim 14, wherein the BAX activating partner ²is ²selected from the group consisting of tBID, PUMA, BIM and NOXA.²16. The synthetic Fab antibody of claim 10, wherein the synthetic Fab antibody ²is ²conjugated to a therapeutic agent.²Date Recue / Date Received 2023-12-15²<DP=4>²-36-²17. The synthetic Fab antibody of claim 10, wherein the synthetic Fab antibody ²is ²conjugated to an agent that facilitates transport across a cell membrane.²18. The synthetic Fab antibody of claim 10, wherein the synthetic Fab antibody ²is ²labeled with a fluorescent label or a radioactive label.²Date Re cue / Date Received 2023- 12-15²