Cell Survival and Cell Death at the Intersection of Autophagy and Apoptosis: Implications for Current and Future Cancer Therapeutics

Abstract

Autophagy and apoptosis are functionally distinct mechanisms for cytoplasmic and cellular turnover. While these two pathways are distinct, they can also regulate each other, and central components of the apoptosis or autophagy pathway regulate both processes directly. Furthermore, several upstream stress-inducing signaling pathways can influence both autophagy and apoptosis. The crosstalk between autophagy and apoptosis has an integral role in pathological processes, including those related to cancer, homeostasis, and aging. Apoptosis is a form of programmed cell death, tightly regulated by various cellular and biochemical mechanisms, some of which have been the focus of drug discovery efforts targeting cancer therapeutics. Autophagy is a cellular degradation pathway whereby cells recycle macromolecules and organelles to generate energy when subjected to stress. Autophagy can act as either a prodeath or a prosurvival process and is both tissue and microenvironment specific. In this review we describe five groups of proteins that are integral to the apoptosis pathway and discuss their role in regulating autophagy. We highlight several apoptosis-inducing small molecules and biologics that have been developed and advanced into the clinic and discuss their effects on autophagy. For the most part, these apoptosis-inducing compounds appear to elevate autophagy activity. Under certain circumstances autophagy demonstrates cytoprotective functions and is overactivated in response to chemo- or radiotherapy which can lead to drug resistance, representing a clinical obstacle for successful cancer treatment. Thus, targeting the autophagy pathway in combination with apoptosis-inducing compounds may be a promising strategy for cancer therapy.

Introduction

Autophagy and apoptosis are two self-destructive processes that recycle cytoplasmic organelles and entire cells, respectively. While these pathways are distinct, they can affect each other, and dysregulation of either autophagy or apoptosis influences numerous pathological processes.¹

Macroautophagy (hereafter referred to as autophagy) facilitates the proteolytic degradation of cytosolic proteins and organelles in lysosomes to generate energy and remove harmful components from cells.² The process is set in motion by the formation of double-membrane vesicles, known as autophagosomes, which fuse with lysosomes to form autolysosomes.³ Autophagy is generally considered a cell-protective response that can also act as a prodeath process in a context-dependent manner.⁴⁻⁶ Malignant cancers often have elevated autophagy activity and have a functional role in tumorigenesis, tumor evasion, and resistance.⁷ The clinically approved antimalarial drug hydroxychloroquine (HCQ) is an autophagy inhibitor that has shown promising results at high doses in combination with standard of care (SOC) cancer chemotherapeutics in clinical trials.⁸ However, these results were inconsistent and HCQ has significant cardiac and retinal toxicity.⁹ Thus, there is intense interest in developing novel therapeutic agents that specifically target autophagy and this was recently reviewed by our group.¹⁰

Apoptosis is a form of programmed cell death that is a vital component of various processes including cell turnover, proper development and functioning of the immune system, embryonic development, and drug-induced cell death.¹¹ Programmed cell death can be activated in two ways: through the extrinsic or intrinsic pathways. This results in degradation of chromosomal DNA, cytoskeletal and nuclear proteins, the formation of apoptotic bodies, and finally uptake by phagocytic cells. Apoptosis is an essential process during development and for maintaining the homeostasis of cell populations in tissues,¹¹ but it is often dysregulated in malignant cells, and the evasion of apoptosis is a hallmark of cancer.^12,13 As cancer cells divide and proliferate, normal control of cell death is impaired and tumor formation occurs. As such, researchers have been working on developing drugs against several targets in the apoptosis pathway. These include the B-cell lymphoma 2 (Bcl-2) family of proteins, the tumor suppressor p53, the inhibitor of apoptosis (IAP) proteins, the tumor necrosis factor receptor (TNFR) superfamily, and the cysteine-dependent aspartate-directed proteases (caspases).

Several therapeutics targeting apoptosis have entered clinical trials for the treatment of solid or hematologic cancers and have shown varying effects in patients. Venetoclax, a Bcl-2 inhibitor, has received FDA approval for hematologic malignancies,¹⁴ and Debio 1143, an IAP antagonist, has recently been granted breakthrough therapy designation for head and neck cancer.¹⁵ Similar to other apoptosis targeting drugs, however, some cancer patients acquire resistance to these therapeutics. Thus, combination treatment with other anticancer agents may be required to improve the overall outcome for patients with cancer. Apoptosis targeting drugs have been studied extensively in combination with other anticancer agents including chemotherapy, radiation therapy, and targeted small-molecule inhibitors. Moreover, recent studies have suggested that IAP antagonists could enhance the efficacy of immune checkpoint inhibitors (ICIs) for immunotherapy.^16,17 Herein, we focus on the crosstalk between the apoptosis and autophagy pathways, highlighting promising combination strategies for these two approaches for the treatment of cancer.

Apoptosis

Apoptosis can be activated through the extrinsic or intrinsic pathways (Figure 1). The extrinsic pathway is initiated by extracellular “death” ligands that bind to the corresponding receptors including TNF/TNFR1, FS-7-associated surface antigen (Fas) ligand (FasL)/Fas receptor (FasR), and TNF-related apoptosis-inducing ligand (TRAIL)/ TRAIL receptor 1/2 (also known as death receptor (DR) 4/5).¹⁸ Upon binding to their ligand the TNF receptors form a trimer and recruit cytoplasmic adaptor proteins via the death domains to form the death-induced signaling complex (DISC).¹⁹ TNFR1 and DR3/4 recruit the TNFR1-associated death domain (TRADD), which then recruits Fas-associating protein with death domain (FADD) and receptor-interacting protein kinase (RIPK),^20,21 while FasR directly recruits FADD.²² FADD associates with procaspase-8 via dimerization, resulting in the activation of caspase-8 and caspase-10.^23,24 Once caspase-8 and -10 are activated, apoptosis is carried out by activating the effector caspases-3 and -7 and cleaving BID into truncated BID (tBID).²⁵ By binding to caspase-8 and FADD, c-FLIP (Cellular FLICE (FADD-like IL-1β-converting enzyme)-inhibitory protein) can inhibit caspase-8 and -10 activation.²⁵ Caspase-8 activation can be further suppressed by cIAP1 or cIAP2, which associates with TNFR-associated factor (TRAF)1/2 at the DISC.²⁶ The death receptors also activate the JNK and the nuclear factor kappa B (NF-κB) signaling pathways²⁷ and RIPK1/3 mediated necroptotic cell death (Figure 1).²⁸

Figure 1.

Schematic of the cell death pathways. The extrinsic apoptosis pathway is set in motion by the TNF ligands (e.g., TNF, TRAIL, FasL) binding to the TNF receptor superfamily, which recruits TRADD and FADD to form the DISC complex and activates caspase-8. Caspase-8 activation is regulated by TRAF1/2, cIAP1, cIAP2 and c-FLIP. The Bcl-2 family of proteins regulates the intrinsic pathway. The Bcl-2 family of proteins include BH3-only proteins, BIM, BID, BIK, NOXA, and PUMA (purple); prosurvival proteins, Bcl-2 and Bcl-xl (yellow); and prodeath proteins, BAX and BAK (dark blue). Activation of the intrinsic pathway results in the formation of BAX/BAK channels and MOMP, resulting in the release of cyto c (green) and SMAC (red) from the mitochondria. Cytosolic cyto c forms the apoptosome complex with APAF1 and pro-caspase-9, activating caspase-9. Caspase-8 can amplify the intrinsic pathway by cleaving BID into tBID. As illustrated, IAP proteins (light blue) and SMAC also regulate apoptosis. Initiator caspases activate the executioner caspases-3, -6, and -7, which carry out apoptosis. The TNF receptors can also activate the NF-κB signaling pathway and recruit RIPK1 and RIPK3 to activate mixed lineage kinase domain-like (MLKL) and implement the necroptosis pathway in a caspase-independent manner.

The intrinsic pathway senses intracellular damage and is initiated by the Bcl-2 family of proteins.²⁹ The Bcl-2 protein family consists of prosurvival, prodeath, and Bcl-2 homology (BH)3-only proteins. Prosurvival Bcl-2 proteins (Bcl-2, Bcl-XL, Mcl-1) bind to and inhibit death proteins Bcl-2-associated X-protein (BAX) and Bcl-2 homologous antagonist/killer (BAK). To initiate apoptosis, BH3-only proteins (PUMA, BID, BIM, BAD, BIK, NOXA) bind to prosurvival Bcl-2 proteins allowing unbound BAX and BAK to oligomerize and result in mitochondrial outer membrane permeabilization (MOMP). This results in the formation of channels that release cytochrome c (cyto c) and the second mitochondria-derived activator of caspases (SMAC) into the cytosol.³⁰⁻³² Mitochondrial damage and cyto c release are further mediated by tBID.³³ Cytosolic SMAC perpetuates the apoptotic signal by competing with caspases for binding to the IAP proteins.³¹ Cytoplasmic cyto c associates with apoptotic protease-activating factor 1 (APAF1) and caspase-9 to form the caspase-9-activating apoptosome.^34,35 DNA damage can activate proapoptotic BH3-only proteins, p53 upregulated modulator of apoptosis (PUMA), or caspase-2 in a p53-dependent manner.³⁶

The extrinsic and intrinsic apoptosis pathways converge to activate executioner caspases-3, -6, and -7 by initiator caspases-2, -8, -9, and -10.³⁷ The caspases are a group of cysteine proteases that cleave proteins at aspartic acid residues and carry out apoptosis.^38,39 Nascent active caspases are regulated by the IAPs.⁴⁰ There are eight IAP proteins, NAIP, cIAP1, cIAP2, XIAP, survivin, BRUCE (apollon), ML-IAP (livin) and ILP-2, that are ascribed to the IAP family by the presence of one or more baculovirus IAP protein repeat (BIR) domain. XIAP is the best characterized IAP protein and can directly inhibit capsase-3, -7, and -9 via its BIR2 and BIR3 domains. cIAP1 and cIAP2 inhibit caspase-8 activation by binding to TRAF1/2 at the DISC.²⁶ Elevated levels of other IAP proteins (survivin, apollon, ML-IAP) have also been shown to modulate the apoptosis pathway. Survivin has been associated with evasion of apoptosis through stabilizing XIAP,⁴¹ and BRUCE has been shown to modulate SMAC and caspase-9 levels. Cytoplasmic SMAC regulates apoptosis by binding to the IAP proteins in a manner that displaces the caspases. In addition, ML-IAP regulates cytoplasmic SMAC levels through its RING domain, facilitating the ubiquitination and degradation of itself and SMAC. Finally, apoptosis is carried out by the effector caspases and results in the degradation of chromosomal DNA and cytoskeletal and nuclear proteins, formation of apoptotic bodies, and lastly uptake by phagocytic cells.

Autophagy

The autophagy pathway is spatially and temporally regulated by multiple molecular components (Figure 2a).³ Autophagy is initiated by a complex comprising UNC51-like kinases 1 or 2 (ULK1/2), the FAK family kinase-interacting protein of 200 kDa (FIP200), and autophagy-related genes (Atg)13 and Atg101.⁴² The ULK1 kinase initiation complex is negatively regulated by the mechanistic target of rapamycin complex 1 (mTORC1) and positively regulated by AMP-activated protein kinase (AMPK), which are both cellular energy and stress sensors.^42,43 Downstream the vacuolar protein sorting 34 (Vps34)-beclin1 complex drives the nucleation of the isolation membrane.^44,45 Autophagy is set in motion by the formation of an isolation membrane known as the phagophore. Nucleation facilitates the formation of a double-membrane compartment called the autophagosome that compartmentalizes targeted proteins and organelles.³ Two essential ubiquitin-like frameworks are activated, the Atg5-Atg12 conjugation step and the microtubule-associated protein light chain 3 (LC3) or Atg8 processing step. Mammals have multiple Atg8 homologues (LC3A, LC3B, LC3C, GABARAP, GABARAPL1, and GABARAPL2) involved in the autophagy pathway.⁴⁶ The protease Atg4 cleaves LC3 into LC3-I and subsequently Atg5 and Atg7 mediate the conjugation of phosphatidylethanolamine (PE) to LC3-I, forming LC3-II, and direct localization to the autophagosome.⁴⁷ The scaffold protein p62 (or SQSTM1) traffics ubiquitinated protein aggregates to the autophagosomes and directly binds to LC3-II.⁴⁸ Finally, the autophagosome either may initially fuse with an endosome before depositing its cargo into the lysosome or may directly fuse with the lysosome to form an autolysosome. Autolysosome maturation is propagated by several lysosomal proteases that hydrolyze proteins, lipids, and nucleic acids.³

Figure 2.

Autophagy pathway and crosstalk between autophagy and apoptosis. (a) Components of the autophagy pathway are illustrated. The pathway is initiated by the ULK1/2 complex (red), which consists of ULK1 or ULK2, Atg13, FIP200, and Atg101. AMPK positively and mTORC1 negatively regulates the ULK1/2 complex. The Vps34-beclin1 complex (green) consists of Vps34, beclin1, p150, Atg14, and PI3K-III. The Vps34-beclin1 complex regulates the formation of the isolation membrane. Expansion of the phagophore is modulated by two ubiquitin-like frameworks: the Atg12 system (purple), Atg12, Atg7, Atg5, Atg10, and Atg15, and the Atg8 system (yellow), LC3, Atg4, Atg5, and Atg7. Atg2 (cyan) functions in autophagosome formation. At later stages, the autophagosome fuses with the lysosome to form the autolysosome. Here components are degraded and released back into the cell. (b) Mitophagy regulates mitochondria outer membrane permeabilization (MOMP). Mitophagy removes damaged or unwanted mitochondria, which increases the threshold for MOMP and cyto c release. Cytosolic cytochrome c activates apoptosis and can inhibit mitophagy by activating caspase-8. (c) Formation of intracellular death-induced signaling complexes at the autophagosome membranes modulates in crosstalk between autophagy and apoptosis. Atg5-Atg12-Atg16L bound to the autophagosomal membrane forms a complex with FADD and caspase-8 to induce apoptosis. Similarly, LC3-II and p62 can bind caspase-8 and initial programmed cell death.

Autophagy-Dependent Cell Death

Autophagy-dependent cell death (ADCD) defines a form of regulated cell death that strictly depends on the components of the autophagy machinery rather than the change in autophagy accommodating apoptosis or necroptosis.^1,49,50 Autophagy is a cytoprotective process that responds to cell stress, but under certain conditions enhanced autophagic flux results in ADCD and the suppression of autophagy prevents cellular demise.⁵¹⁻⁵³ Much of the evidence for ADCD comes from the study of Drosophila melanogaster. In these studies, suppression of essential Atg genes was shown to prevent midgut degradation, while inactivation of caspases did not affect tissue development.^54,55 In the Drosophila salivary glands, both the apoptotic as well as the autophagy machinery are required for cell degradation during development.^56,57 In mammals, Shimizu et al. showed that mouse embryonic fibroblasts (MEF) from apoptosis resistant double-knockout Bax/Bak mice were able to undergo cell death when exposed to the DNA damage-inducing chemotherapeutic etoposide.⁵⁸ This was attributed to an increase in autophagosomes and was further suppressed by genetically or chemically inhibiting autophagy. Further validation of ADCD in mammals was demonstrated using Atg5/Bax/Bak triple-knockout mice.⁵⁹ Embryos from these mice also showed significantly delayed interdigital loss and brain malformation compared to wild-type mice.^58,59 In the A549 lung cancer cell line, resveratrol treatment resulted in ADCD, displaying similar sustained induction of autophagy flux that was attenuated by knocking down certain autophagy genes. GRB2-associated-binding protein 1 (GBA1) was shown to positively regulate resveratrol-dependent autophagic cell death, did not activate the apoptosis or necroptosis pathway, and was not affected by treatment with a pan-caspase inhibitor, z-VAD-FMK.⁶⁰ This compensatory role of autophagy was also shown to be the main process for cell death by glucocorticoids in hematological malignancies.^61,62 Autosis is a specific form of ADCD, whereby elevated autophagy flux and cell death are both dependent on the plasma membrane Na⁺/K⁺-ATPase.⁶³ These studies suggest that cytotoxic drugs can induce cell death in an autophagy-dependent manner when the apoptotic or necroptotic pathway is not functional. Additional in vivo mammalian studies are still required to verify these observations and validate the therapeutic potential of ADCD.

Crosstalk between Apoptosis and Autophagy

Autophagy has been shown to regulate apoptosis and vice versa; however, the exact nature of this interaction is unclear. Some connections occur upstream of the autophagic and apoptotic pathways where stress-elicited signaling pathways regulate both processes. p53, a potent activator of apoptosis, can also induce autophagy through increasing the expression of damage-regulated autophagy modulator (DRAM), a target gene for p53.⁶⁴ Similarly, activation of the PI3 kinase/Akt pathway, which is a well-known pathway that inhibits apoptosis, also inhibits autophagy.⁶⁵ Furthermore, oncogenes, including myc,⁶⁶ and major signaling pathways, such as the Ras-PI3K-mTOR axis,⁶⁷ can simultaneously increase and/or decrease both autophagy and apoptosis.

Proteins that are themselves central components of the apoptosis or autophagy pathway regulate both processes directly.⁶⁸ Components of the autophagy and apoptosis machinery frequently intersect and cross-regulate one another at the mitochondria (Figure 2b). Intrinsic apoptosis is initiated by MOMP and the release of cytochrome c. Mitophagy is a specific form of autophagy that selectively clears damaged mitochondria.^69,70 Autophagy induction (e.g., by nutrient deprivation, sulforaphane treatment, etc.) can result in mitochondria elongation, inhibiting the release of cytochrome c⁷¹, which prevents apoptosis and sustains cell viability.⁷² Apoptosis induction following cytochrome c release activates caspase-8 which cleaves several Atg proteins and inactivates autophagy.⁷³⁻⁷⁵ This can enhance apoptosis by further promoting the release of proapoptotic factors from mitochondria.⁷⁶ Alternatively, in mitochondrial apoptosis deficient cells, autophagy induction can eventually result in stress-induced cell death through caspase-8 activation.⁷⁷ Furthermore, Atg12 has been shown to interact with apoptosis proteins at the mitochondria; however, the effect of this interaction on cell fate is still unclear.^78,79

Another recently identified platform for crosstalk of the two signaling pathways is the autophagosomal membrane. Several groups have demonstrated the formation of intracellular DISCs (iDISC) that recruit procaspase-8 and FADD onto autophagosomes (Figure 2c).⁸⁰⁻⁸⁶ Caspase-8 is recruited to the phagophore by FADD, which interacts with the Atg5-Atg12-Atg16L complex^80,83,85, or by binding to p62 and LC3.^81,82,86 This complex usually forms independently from death receptor signaling and requires LC3-positive autophagic membranes. Knockdown of Atg5 or Atg7 or treatment with the autophagy inhibitor 3-MA suppresses the formation of the iDISC.^80,81,84 Depending on the cellular conditions, recruited caspase-8 will either be activated at the phagophore through oligomerization and enhance cell death,⁸⁰⁻⁸⁶ or the mature autophagosomes will fuse with the lysosome, degrade caspase-8, and block apoptosis.⁸⁷⁻⁸⁹ According to Tang et al. this switch is regulated by c-FLIP and NF-kB signaling on expanding autophagosomal membranes.⁸⁴ Under nutrient-rich conditions NF-kB signaling upregulates c-FLIP to suppress caspase-8 activation while starvation suppresses NF-kB translocation allowing caspase-8 activation and initiation of apoptosis.⁸⁴

Several additional mechanisms of cross-regulation between autophagy and apoptosis have been identified, allowing the same protein to participate in these opposing processes. Atg5 has been demonstrated to regulate autophagy and enhance apoptosis.⁹⁰ Atg5 can be cleaved by the noncaspase protease calpain, allowing truncated Atg5 to associate with Bcl-xl at the mitochondria, resulting in cytochrome c release and caspase activation.⁹⁰ Betin et al. showed that Atg4D is cleaved by caspase-3 which stimulates the processing of γ-aminobutyric acid receptor-associated protein-like 1 (GABARAP-L1) and autophagosome formation. Interestingly, both genetic knockdown of Atg4D and overexpression of cleaved Atg4D resulted in increased cytotoxicity, suggesting that Atg4D is an important factor at the regulatory interface between autophagy and apoptosis.⁹¹ Caspase-10 has also been shown to inhibit autophagy by cleaving Bcl-2-associated transcription factor 1 (BCLAF1) and thus preventing ADCD in multiple myeloma.⁹² Furthermore, Han et al. demonstrated that caspase-9 forms a complex with Atg7, which inhibits the catalytic activity of caspase-9 and enhances LC3-II processing and autophagosome formation.⁹³ This suggests that linking apoptosis genes to the autophagy process regulates cell death pathways and vice versa.

Necroptosis is a caspase-independent process defined as “a modality of regulated cell death triggered by perturbations of extracellular or intracellular homeostasis that critically depends on MLKL, RIPK3, and RIPK1”.⁵⁰ Necroptosis has been shown to crosstalk with the apoptosis and autophagy pathways. On autophagosomal membranes, RIPK1 forms a complex with FADD:caspase-8:Atg5:Atg12:Atg16L.⁸⁰ Under normal cell media conditions, active caspase-8 inhibits autophagy by blocking RIPK1 and cells die through apoptosis. In the absence of caspase-8, active RIPK1 can instruct cells to die through necroptosis rather than through ADCD.⁸⁰ One study found that RIPK3 directly activated AMPK, which also resulted in phosphorylation of ULK1 and beclin1.⁹⁴ Furthermore, TNF-induced necroptosis has been shown to block the lysosomal degradation of autophagosomes possibly through dysregulating the SNARE complex.⁹⁴ In addition, there is evidence that blocking apoptosis with the caspase inhibitor z-VAD-FMK results in cells dying through necroptosis rather than ADCD as z-VAD-FMK also inhibits lysosomal proteases and thereby autophagy.^95,96

While many studies show that autophagy is upregulated in response to cancer therapy, which then acts as an adaptive response to promote tumor cell survival, under other circumstances, autophagy induction can further promote cells to undergo cell death. One explanation for this may be that modulation of the autophagy pathway results in changes that are cell type- and context-dependent. Alternatively, commonly used pan-caspase inhibitors used to block the apoptosis pathway, as well as autophagy assays used to measure autophagosome accumulation, have several limitations and this complicates the interpretation of results.

Bcl-2 Proteins and BH3-Mimetics

The Bcl-2 protein family closely regulates the balance between cell survival and cell death.⁹⁷ These prosurvival proteins (Bcl-2, Bcl-XL, Mcl-1) promote cell survival by inhibiting the proapoptotic proteins BAX and BAK, preventing BAX/BAK from oligomerizing, forming pores, and permeabilizing the outer mitochondrial membrane. The BH3-only proteins (BIM, PUMA, BID, BIK, and NOXA) initiate programmed cell death in response to stress by binding to prosurvival proteins or by activating the effectors of apoptosis.⁹⁸ Characterization of the functions and structural interactions between proapoptotic and prosurvival Bcl-2 family members resulted in the development of small-molecule mimetics (BH3-mimetics) that kill cancer cells by targeting prosurvival Bcl-2 members (Table 1). To date, six BH3-mimetic drugs have reached the clinic. One of them, the Bcl-2 inhibitor, venetoclax/ABT-199,⁹⁹ is currently approved for patients with chronic lymphocytic leukemia (CLL) or small lymphocytic lymphoma (SLL)¹⁴ and may become the front-line therapy for acute myeloid leukemia (AML) in the future.^100−102 However, leukemic cells can acquire resistance to ABT-199 and other BH3-mimetics,^103,104 suggesting that combination therapy with other antileukemic agents may be required to reduce or overcome the resistance.

Table 1. Selected Apoptosis Drugs and Their Effects on Autophagy.

Bcl-2 proteins primarily regulate the autophagy pathway by interacting with a key protein beclin1, which forms the beclin1–Vps34 core complex to induce autophagy (Figure 3a).¹⁰⁵ Bcl-2 was initially identified to bind to beclin1 in yeast¹⁰⁶ and later in mammalian cells to inhibit beclin1-dependent autophagy.¹⁰⁷ Beclin1 has a BH3 domain that mediates the Bcl-2–beclin1 complex formation during low nutrient levels and downregulates autophagy during starvation.¹⁰⁷ Beclin1 can also bind to Bcl-XL and Mcl-1.¹⁰⁵ Proapoptotic proteins such as BAX decrease autophagy activity by enhancing caspase-mediated cleavage of beclin1 and preventing beclin1 from interacting with Vps34.¹⁰⁸ BH3-only proteins also regulate autophagy. For example, BIK displaces nutrient-deprivation autophagy factor-1 (NAF-1) from Bcl-2 at the endoplasmic reticulum and initiates autophagy.¹⁰⁹ Beclin1 phosphorylation at the BH3 domain by kinases further regulates its interaction with Bcl-2/Bcl-XL. Death-associated protein kinase (DAPk) phosphorylates beclin1 at T119, promoting beclin1 dissociation from Bcl-XL and autophagy induction.¹¹⁰ Alternatively, serine/threonine kinase 4 (STK4) phosphorylates beclin1 at the Thr108 site and suppresses autophagy by enhancing the interaction between beclin1 and Bcl-1/Bcl-XL.^111,112 Additional mechanisms by which beclin1 regulates autophagy and apoptosis have been extensively reviewed by others.^105,113

Figure 3.

Components of the apoptosis pathway modulate the autophagy pathway. (a) Bcl-2 proteins module the Vps34–-beclin1 complex. Beclin1 has a BH3 domain allowing it to bind to Bcl-2 proteins. Prosurvival Bcl-2 proteins (Bcl-2, Bcl-XL, and Mcl-1) can form a complex with beclin1, preventing the formation of the Vps34–beclin1 complex and downregulating autophagy. BH3-only proteins (BIK, BID, BIM, and PUMA) indirectly positively regulate autophagy. The proapoptotic protein, BAX, negatively regulates autophagy by enhancing caspase-mediated cleavage of beclin1 and preventing it from interacting with Vps34. (b) p53 can both positively and negatively modulate the autophagy pathway depending on its cellular localization. Cytoplasmic p53 negatively regulates autophagy by binding to FIP200, resulting in the destabilization of the ULK1/2 complex. Nuclear p53 positively regulates autophagy by inducing the transcription of AMPK, mTOR, and DRAM. (c) IAP proteins interact with various components of the autophagy machinery. Survivin downregulates autophagy by inhibiting the Atg12–Atg5–Atg16 complex and Atg7. Apollon positively regulates autophagy initiation through the upstream kinases AMPK and ULK1. Apollon also inhibits the fusion between autophagosomes and lysosomes. Similarly, XIAP can positively or negatively modulate the autophagy pathway via beclin1 and LC3 expression, respectively. (d) The death domain of FADD has been shown to upregulate autophagy under certain cellular circumstances, while under different cellular conditions, elevated autophagy has been shown to degrade DR4 and DR5.

BH3-mimetics have been reported to be implicated in regulating the autophagy pathway (Table 1). Recent studies have focused on ABT-199, a selective inhibitor of Bcl-2, ABT-263, and ABT-737, inhibitors of Bcl-2, Bcl-XL, and Bcl-w, and pan-Bcl-2 inhibitors, obatoclax, and gossypol.¹¹⁴ Most notably, studies in primary follicular lymphoma cells resistant to ABT-199 show an increase in autophagy activity assessed by a decrease in SQSTM1/p62 levels and an increase in LC3 I/II processing.^115,116 Treatment of multiple myeloma or colon cancer cells by the ABT-199 analogues ABT-737 and ABT-263, respectively, similarly increased autophagy activity.^81,117 BH3-mimetics, ABT-199, ABT-737, and ABT-263, competitively inhibit the interaction between beclin1 and Bcl-2/Bcl-XL, freeing beclin1 from inhibition and activating autophagy.¹¹⁸ Moreover, combination treatment with the autophagy inhibitor chloroquine (CQ) further enhances the ability of ABT-263 to induce apoptosis⁸¹ and can be attenuated by knocking down essential autophagy-related proteins. Bortezomib resistant multiple myeloma (MM) cells with low BIM expression were also able to overcome resistance when treated with ABT-737 and CQ.¹¹⁹ These studies suggest that autophagy contributes to mitigating the cytotoxicity of BH3-mimetics.¹¹⁴ The pan-Bcl-2 inhibitor obatoclax (GX15-070) appears to induce autophagy in various cancer cells independently of beclin1 as knockdown of beclin1 did not attenuate obatoclax-induced autophagy.^120,121 However, the mechanism by which obatoclax induces autophagy is not fully understood, and questions regarding whether obatoclax-induced autophagy promotes or inhibits cell death are also under discussion.¹¹⁴ On one hand, obatoclax-induced autophagy has been linked to cell death,¹ as genetic or pharmacological blockade of autophagy was found to inhibit cell death.¹¹⁴ Alternatively, inhibition of autophagy was demonstrated to enhance obatoclax-induced cell death, supporting a prosurvival function of autophagy.¹²² Furthermore, obatoclax has also been shown to inhibit late-stage autophagy independently of beclin1, Atg7, and Atg12.¹²³ Gossypol, another pan-Bcl-2 inhibitor, has also been shown to induce autophagy and apoptotic cell death.^124,125 In one study, cotreatment of gossypol with autophagy inhibitors potentiated cell death, suggesting that gossypol-induced autophagy is cytoprotective.¹²⁵ Others have made the antithetical observation that genetic knockdown of beclin1 and/or Atg5 reduced gossypol-induced cell death.¹²⁴ More recently, genetic deletion of Mcl-1 in the nervous system has been shown to induce autophagy,¹²⁶ suggesting that Mcl-1 selective inhibitors may also activate autophagy.

BH3-mimetics are promising therapeutics for the treatment of hematologic malignancies, and all small molecules we reviewed appear to upregulate autophagy. Genetic knockdown experiments demonstrate that Bcl-2 proteins intersect with autophagy by interacting with beclin1, suggesting that inhibition of Bcl-2 proteins requires free beclin1 to initiate autophagy. Interestingly, not all BH3-mimetics required beclin1 to increase autophagy as beclin1 knockdown did not always attenuate inhibitor-induced autophagy. This opens two possibilities: (1) Bcl-2 proteins are able to regulate autophagy in a not yet understood beclin1-independent manner, or (2) Bcl-2 inhibitors are binding to off-target proteins involved in regulating autophagy.

p53 and MDM2-p53 Antagonists

Tumor protein p53 is a master regulator of genomic stability, cell death, proliferation, and metabolism and is the most well-characterized tumor suppressor protein.¹²⁷ It is a stress-responsive transcription factor, and upon activation, p53 can promote cells to undergo apoptosis by inducing proapoptotic genes including PUMA, BAX, and BIM.¹ p53 has been extensively studied in the context of cancer, and several compounds that reactivate the tumor suppressor protein have been developed.¹²⁸ Murine double minute 2 (MDM2) tightly regulates and inhibits the activity of p53. Blocking the MDM2–p53 interaction with small molecules can reactivate the tumor-suppressive functions of p53.^129,130 Several small-molecule MDM2 inhibitors have been successfully designed and developed (Table 1), and seven of these compounds have been advanced into human clinical trials as novel anticancer therapeutics.¹²⁸ It is estimated that p53 is mutated or deleted in ∼50% of human cancers and can vary significantly between cancer types.¹³¹ Thus, the ability of these MDM2–p53 antagonists to induce apoptosis is restricted to WT p53 cancers and generally these compounds do not fully eliminate the cancer cells.¹²⁹ Consequently, MDM2 antagonists are increasingly being tested in combination with other therapeutics in a search for synergy that would enhance the efficiency of these drugs to eliminate cancer.

The role of p53 in the regulation of autophagy appears to be dependent on its cellular localization, and the mechanism is not yet fully understood (Figure 3b).¹³² Tasdemir et al. found that cytoplasmic, but not nuclear, p53 suppresses autophagy, and knockout, knockdown, or pharmacologic inhibition of p53 induces canonical autophagy signaling.¹³³ This enhanced autophagy helped cancer cells survive under stress. Cytoplasmic p53 inhibits autophagy by coaggregating with FIP200, resulting in the destabilization of ULK1. This blocks the formation of the ULK1–FIP200–Atg13–Atg101 complex and prevents autophagosome formation.^134,135 Furthermore, induction of autophagy by starvation or rapamycin treatment stimulates proteasomal degradation of p53 and a further increase in autophagy.¹³³ Alternatively, nuclear p53 has been shown to activate autophagy by regulating expression of the AMPK/mTOR pathway^136,137 and is a key regulator of the lysosomal protein DRAM.⁶⁴

Several groups have looked at the ability of MDM2–p53 antagonists to modulate autophagy in various cancer models. Nutlin-3a¹³⁸ and MI-63¹³⁹ have both been shown to induce p53-dependent autophagy in hematologic malignancies.^140,141 This p53-dependent increase in autophagy appears to be proapoptotic as genetic or pharmacological inhibition of autophagy impaired apoptosis. CP-31398¹⁴² and RITA (reactivation of p53 and induction of tumor cell apoptosis)¹⁴³ were also shown to induce autophagy; however, in pancreatic cancer cells, cotreatment with autophagy inhibitors further sensitized these cells to apoptotic cell death, suggesting a prosurvival autophagic response in this context.¹⁴⁴ Furthermore, nutlin-3a synergizes with several anticancer therapeutics to induce autophagy and apoptosis.^145−147 Some studies suggest that nutlin-3a modulates autophagy in a context-dependent manner. Duan et al. compared cancer cells sensitive to, or resistant to, nutlin-3a-induced apoptosis and found that in resistant cells nutlin-3a promoted prosurvival autophagy while in sensitive cells nutlin-3a inhibited autophagy. Additionally, cotreatment with autophagy inhibitors and nutlin-3a could sensitize the nutlin-3a resistant cancer cells to undergo cell death.¹⁴⁸

It has been proposed that autophagy suppresses p53 in some cancers, facilitating cell survival and tumorigenesis.^149−152 Therefore, cotreatment with MDM2 inhibitors and autophagy inhibitors could hyperactivate WT p53, more effectively induce apoptosis in cancer cells, and impede disease progression. However, WT p53 cancers make up only ∼50% of human cancers, which limits the clinical utility of MDM2–p53 antagonists and would require patient stratification.

IAP Proteins and Antagonists

The family of IAP proteins is involved in blocking and attenuating programmed cell death pathways, predominantly through modulation of the caspase cascade.^153,154 IAP proteins cIAP1, cIAP2, XIAP, and ML-IAP are directly involved in the regulation of apoptosis.¹⁵⁵ IAPs are often up-regulated in cancers¹⁵⁶ and are believed to underlie the resistance of many malignant cells to chemotherapeutics.^157,158 IAP antagonists, or SMAC mimetics, have received considerable attention as potential therapeutic agents for the treatment of cancer (Table 1),^159,160 with several compounds under clinical evaluation for solid and hematologic cancers.¹⁶¹ Clinical data from these studies showed that SMAC mimetics are well tolerated, display evidence of target engagement, and have a reasonable safety profile.¹⁶² However, previous SMAC mimetics have demonstrated limited clinical efficacy as single agents, with the majority of patients showing no response to the drugs.¹⁶² More recently, there have been some advances in the clinical efficacy of IAP antagonists, with the breakthrough therapy destination approval for Debio 1143 for front line treatment of head and neck cancer,¹⁵ however, there is still a need for safe and effective IAP antagonists that synergize with other anticancer drugs.

Besides their antiapoptotic role, IAP proteins have recently been linked with modulating the autophagy pathway (Figure 3c).¹⁶³ A study using hepatocellular carcinoma tissue specimens found that XIAP expression was inversely associated with LC3 expression.¹⁶⁴ Furthermore, knockdown studies by Huang et al. demonstrated that XIAP negatively regulated the autophagy pathway through MDM2–p53 signaling in vitro and in vivo.¹⁶⁵ Alternatively, XIAP and cIAP1 have also been shown to positively regulate beclin1-dependent autophagy expression via the NF-kB pathway.¹⁶⁶ Survivin has been suggested to modulate autophagy by interacting with the Atg12–Atg5–Atg16L protein complex to inhibit autophagosome formation and the elongation process.^167−169 Recent studies suggest that BRUCE may also play a role in autophagy, potentially at the stage of fusion between autophagosomes and lysosomes to form autolysosomes.¹⁶³ However, this appears to be unclear as some studies found that BRUCE knockdown negatively regulated autophagy¹⁷⁰ by interacting with GABARAP and GABARAPL1, and syntaxin 17¹⁷¹ or through the upstream kinases AMPK-ULK1.¹⁷² Alternatively, BRUCE may degrade LC3-I and thus negatively regulate autophagy.¹⁷³

Recent studies have also evaluated the ability of IAP antagonists to modulate autophagy. APG-1387, a bivalent IAP antagonist that is currently under clinical evaluation for cancer and has been demonstrated to induce autophagy in ovarian cancer cells.¹⁷⁴ Furthermore, cotreatment with autophagy inhibitors and APG-1387 further promoted cell death.¹⁷⁴ Similarly, WX20120108, a monovalent IAP antagonist, increased autophagy activity in HeLa and MDA-MB-231 cells. Both compounds appear to increase autophagy activity as cells start to undergo cell death with an increase in compound concentration and incubation time.¹⁷⁵ Ding et al. suggested that WX20120108 relies on activating the ROS-Foxo3 pathway to induce autophagy.¹⁷⁵ The clinical candidate GDC-0152 also induced autophagy in HL-60 cells. However, rather than acting as a cell survival mechanism, autophagy further increased apoptosis, and cotreatment with autophagy inhibitors reduced GDC-0152-induced apoptosis.¹⁷⁶ LCL-161 and birinapant were evaluated in an autophagy flux assay and were both shown to inhibit autophagy by blocking the fusion between autophagosomes and lysosomes.¹⁷⁷ YM155, a survivin suppressor, was shown to upregulate autophagy and induce cell death and DNA damage in an autophagy-dependent manner.^178,179

There is evidence that IAP proteins and antagonists modulate autophagy; however, the mechanism is not fully understood and may be context-dependent. While most studies demonstrate that IAP antagonists elevate both autophagy and apoptosis in certain contexts, in some cells and tissues IAP antagonists have been shown to inhibit autophagy. In the future a better understanding of the functions of IAPs may be necessary to enable us to fine-tune the therapeutic efficacy of IAP antagonists and attenuate possible side effects.

Death Receptor Ligands

The TNFR superfamily constitutes a key component of the extrinsic apoptosis pathway. Death receptor (DR)4 and DR5 are part of the TNFR family and contain “death domains” that can initiate apoptosis.¹⁸⁰ The death receptors recruit cytoplasmic adaptor protein TRADD, FADD, and procaspase-8 to form the DISC upon binding to the ligands Apo2 ligand (Apo2L) or TRAIL. This results in the cleavage and activation of initiator caspases-8 and -10, which initiate apoptosis by activating the executioner caspases.¹⁸⁰ Apo2L/TRAIL and molecules mimicking their functions have emerged as promising candidates for cancer therapy. This is based on their ability to induce apoptotic cell death in cancerous cells without causing significant toxicity to healthy cells.¹⁸¹ The recombinant human soluble protein Apo2L/TRAIL (dulanermin) and its derivatives were the first attempt to target DR4/5, and while they showed promising efficacy and safety in preclinical models, they did not achieve sufficient overall therapeutic benefit clinical trials.¹⁸² Monoclonal antibodies targeting DR4 or DR5 have also been in development, seven of which have entered clinical studies. The DR4/DR5 monoclonal antibodies were well tolerated and relatively safe in patients, but only achieved stable disease progression and no improved response rates were noted as single agents.¹⁸²

Components of the extrinsic death receptor pathway have also been associated with controlling autophagy (Figure 3d). As mentioned above, TRAIL has a unique ability to induce apoptotic cell death in cancer cells while sparing healthy cells. A study has suggested that this resistance to undergo cell death in normal epithelial cells may be through the upregulation of cytoprotective autophagy.¹⁸³ This was proposed to be mediated by the death domain of FADD or by AMPK signaling. During malignant transformation, cancer cells may become resistant to the death receptor ligands TNF and TRAIL. This resistance is suggested to be through autophagic degradation of DR4 and DR5,^184,185 as TRAIL-resistant cancer cells have shown elevated autophagy activity and pharmacologic or genetic inhibition of autophagy resulted in the death of these cells.^186,187 Alternatively, resistant glioma cells have been shown to overcome this resistance and undergo cell death in response to chemotherapeutic agents by activating autophagy-mediated cell death.¹⁸⁸

Prostate cancer cells treated with dulanermin were shown to have elevated autophagy flux activity in TRAIL-resistant but not in TRAIL-sensitive cell lines. Cotreatment with the autophagy inhibitor CQ further enhanced dulanermin-induced cell death in TRAIL-resistant prostate cancer cells.¹⁸⁶ HW1, a human single-chain fragment variable (scFv) antibody against DR5, was shown to induce autophagy and cell death as a single agent in both TRAIL-resistant and TRAIL-sensitive cells. Furthermore, HW1-induced cell death appeared to be through autophagy-mediated cell death, since knockdown of the autophagy protein beclin1 or Atg7 greatly attenuated HW1-induced cell death.¹⁸⁹

Unlike the other apoptosis-inducing drug targets, limited research has been reported on the role of DR antibodies in autophagy. As previously mentioned, death receptors form DISCs in response to TRAIL/TNF-α treatment at the plasma membrane, which results in caspase-8 activation. While it is known that caspase-8 modulates autophagy by cleaving several autophagy proteins, the extent by which DR antibody-dependent caspase activation affects autophagy remains to be elucidated.

Caspase Inhibitors

The caspases are a group of cysteine proteases that are central to initiating and executing the programmed cell death pathway. In mammals, there are seven major caspases that are involved in apoptosis and are classified as initiators (caspase-2, -8, -9, and -10) or executioners (caspase-3, -6, and -7). There is a third group of caspases (caspase-1, -4, and -5) involved in carrying out pyroptosis, which is an inflammatory type of regulated cell death.⁵⁰ Caspases exist as inactive precursors and are activated by oligomerization, cleavage, and/or dissociation from endogenous inhibitors allowing them to cleave their substrates at specific aspartate-containing recognition sites.¹⁹⁰

Unlike the other discussed apoptosis proteins, caspase inhibitors have been used to prevent cell death in various pathologies, including liver disease. In addition, caspase inhibitors serve as useful tool compounds to study the programmed cell death pathways and the crosstalk with the autophagy pathway. It is unsurprising that caspases also recognize and cleave many autophagy proteins.¹⁹⁰ A study by Norman et al. showed that in vitro Atg3 is cleaved by caspases-3, -6, and -8, beclin1 is cleaved by caspase-3 and -6, p62 is cleaved by caspase-6 and -8, while Atg7, Atg9, and Atg4 homologues can be cleaved by caspase-3.¹⁹¹ This has been validated by other groups in cellular models, and in most cases, cleavage of autophagy proteins by caspases results in suppression of autophagy, and the homeostatic balance shifts toward apoptosis.^190,192,193 Nevertheless, caspase-mediated cleavage of autophagy proteins does not always result in waste products; rather, several Atg-fragments have been shown to acquire new properties, which differ from their initial full-size isoforms and consequently affect apoptosis and autophagy in numerous ways.^76,91 On the other hand, caspases have also been found to promote autophagy under certain contexts.^91,93 Notably, autophagy is also able to regulate apoptosis by targeting caspases for degradation in the autolysosome.^88,190

Several caspase inhibitors have been developed as research tools and as antiapoptotic therapeutics for liver disease.¹⁹⁴ The FDA approved pan-caspase inhibitor IDN-6556 has been shown to inhibit autophagy resulting in cell death by necroptosis in mouse dermal fibroblasts.¹⁹⁵ Peptide-like pan-caspase inhibitors including z-VAD-FMK appear to modulate autophagy in a context-dependent manner. Caspase-mediated cleavage of beclin1 inactivates autophagy, which can be inhibited by z-VAD-FMK.^76,91,196 Similarly, blocking Atg3 cleavage by caspase-8 with z-VAD-FMK has been shown to increase autophagy in cells.⁷⁵ Under certain cellular contexts z-VAD-FMK appears to induce ROS-accumulation and upregulate autophagy, which results in cell death via ADCD or necroptosis.^80,84,197 Alternatively, others have shown that z-VAD-FMK blocks autophagy by inhibiting an off-target lysosomal protease, cathepsin B.⁹⁶ This would argue that when apoptosis is attenuated by z-VAD-FMK cells undergo necroptosis, refuting published examples of ADCD and highlighting the need for more vigorous experiments when studying autophagosome accumulation.

Implications for Current and Future Cancer Therapeutics

In this review, we discussed the crosstalk between apoptosis and autophagy, while focusing on several small-molecules and biologics that target the apoptosis pathway. Several drugs that modulate apoptosis have been developed and advanced into the clinic in recent years by targeting the Bcl-2 protein family, IAP proteins, death receptors, p53, or caspases. However, these anticancer therapeutics generally lack single-agent efficacy and must be combined with other treatments in clinical trials. In vitro studies using a large panel of cell lines have shown that some cancer cells are more predisposed or “primed” to undergo cell death in response to treatment with apoptosis-inducing compounds while other cancer cells are completely resistant to treatment.¹⁹⁸ The majority of studies described in this review demonstrate that autophagy is elevated in response to treatment with apoptosis-inducing compounds. While apoptosis occurs at the same time or is potentially preceded by autophagy, it is unclear to what extent autophagy mediates cell death.

To date, only a handful of autophagy inhibitors (CQ/HCQ, verteporfin, and clarithromycin) have entered clinical trials for the treatment of cancer.¹⁹⁹ CQ, and its analogue HCQ, are late-stage autophagy inhibitors that block the fusion between autophagosomes and lysosomes and are being used in several clinical trials in the treatment of various cancers (ClinicalTrials.gov).¹⁹⁹ Verteporfin is FDA approved for macular degeneration and was found to inhibit autophagy in a screen of off-patent agents.²⁰⁰ Verteporfin has been shown to increase the efficacy of some anticancer agents when used in combination in various cancer models^201−203 and is currently in clinical trials for the treatment of prostate and pancreatic cancer (NCT03067051, NCT03033225). Clarithromycin, an antibiotic used for the treatment of bacterial infections, has also been demonstrated to attenuate autophagy²⁰⁴ and is currently being investigated in ongoing clinical trials for various hematologic malignancies (NCT04302324, NCT04063189). Moreover, several novel small-molecule autophagy inhibitors are in preclinical development, including ULK1/2 inhibitors (SBI-0206965,²⁰⁵ MRT68921²⁰⁶), Vps34 inhibitors (VPS34-IN1,²⁰⁷ SB02024,²⁰⁸ SAR405,²⁰⁹ PIK-III²¹⁰), Atg4B inhibitors (NSC185058,²¹¹ UAMC-2526²¹²), and lysosomal inhibitors (bafilomycin A1,²¹³ Lys05²¹⁴). Preclinical studies have demonstrated that some of these autophagy inhibitors can induce apoptotic cell death as single agents in various cancer types and could show promise as combination treatments with anticancer therapeutics.^{205,215−218}

In many cases, autophagy mitigates the cytotoxicity induced by proapoptotic drugs, and suppression of autophagy by pharmacological inhibition or by genetic knockdown of Atg genes accelerates apoptotic cell death.^215,216 Autophagy inhibition has also been proposed to prime cells for apoptotic cell death in response to anticancer agents by increasing Forkhead box protein O3a (FOXO3a) activity and upregulating the expression of PUMA.^219,220 In vivo studies in mice have shown markedly improved survival rates due to tumor burden when combining antiapoptotic compounds with CQ compared with either drug alone.²²⁰ This suggests a prominent role for autophagy as a prosurvival response and raises the question of whether autophagy inhibitors could synergize with apoptosis drugs and provide therapeutic benefit to patients with cancer. HCQ and verteporfin have shown promising clinical results with multiple chemotherapies and radiation therapies,¹⁹⁹ however, they have not yet been tested in combination with apoptosis targeting compounds in clinical trials. Thus, in principle, it would be feasible to investigate whether administering HCQ or verteporfin in combination with apoptosis-inducing therapeutics would improve treatment outcomes for cancer patients. In addition, novel autophagy inhibitors targeting ULK1/2, Vps34, or Atg4B could also show promise by synergizing with apoptosis therapeutics as they may have fewer off-target effects and could show efficacy at lower doses compared to HCQ.

On the other hand, apoptotic drug-induced autophagy may under different cellular contexts further promote cell death and inhibiting autophagy would suppress cell death. Loss of the essential autophagy gene beclin1 promotes tumorigenesis in mice and this has also been found to occur in human cancers.^12,221 Furthermore, p62 is eliminated through autophagy, but in the absence of autophagy, p62 accumulation results in tumor growth.^222,223 Paradoxically, this suggests that autophagy is a tumor-suppression mechanism, and inhibition of the pathway in combination with proapoptosis drugs would worsen the outcome. Several studies described in this review have demonstrated in vitro and in vivo that autophagy inhibition reverses the effects of apoptosis-inducing compounds. This would suggest that in certain contexts activating autophagy may be a better therapeutic approach in combination with apoptosis-inducing drugs.

In order to improve future cancer therapy, it is essential to investigate the combined effect of autophagy inhibitors and proapoptosis therapeutics not just in tumors and cancer cells but also in different cellular contexts, such as immune cells. Immunotherapies act on the immune system and enhance its ability to recognize, target, and destroy cancer cells.^224,225 Immune checkpoint inhibitors (ICIs) have shown remarkable responses in clinical trials across various cancers.²²⁶ There is much focus on novel immunotherapies, such as cancer vaccines and chimeric antigen receptor (CAR) T cell therapies, as well as investigating combination treatments with immunotherapies and other modalities.²²⁴ Autophagy and apoptosis exhibit demonstrable crosstalk in T cells to regulate survival and proliferation, and components of both pathways have been shown to regulate the immune system.^80,227−229 Moreover, small-molecule proapoptotic drugs have been shown to upregulate major histocompatibility complex (MHC)-I, sensitize cancer cells to T-cell-dependent killing, add to checkpoint inhibitor efficacy in vivo, and are currently being investigated in clinical trials in combination with immunotherapies.^16,17 Autophagy inhibitors have been shown to regulate immune cell activation and response and also demonstrate promise as a therapeutic approach for improving cancer immunotherapy.^227,230

Clinical trials have demonstrated that targeting the apoptosis pathway can be safe and offer benefit to patients with cancer. Recent clinical studies with the autophagy inhibitor HCQ have also shown safety and efficacy in cancer patients. However, both apoptosis-inducing drugs and autophagy inhibitors generally lack single-agent efficacy, and patients often acquire resistance to the therapy over time. Numerous in vitro and in vivo studies have shown enhanced efficacy at inducing cell death in cancer cells when cotreating proapoptosis drugs with autophagy inhibitors. Further, investigating the combinatorial effect of proapoptosis therapeutics and autophagy inhibitors could provide a novel approach to improve future cancer therapy and offer significant benefit to patients.

Glossary

Abbreviations

HCQ

hydroxychloroquine

SOC

standard of care

Bcl-2

B-cell lymphoma 2

IAP

inhibitor of apoptosis

TNF

tumor necrosis factor

TNFR

tumor necrosis factor receptor

FASR

FS-7-associated surface antigen receptor

TRALR

TNF-related apoptosis-inducing ligand receptor

DISC

death-induced signaling complex

tBID

truncated BID

BH3

Bcl-2 homology 3

BAX

Bcl-2-associated X-protein

BAK

Bcl-2 homologous antagonist/killer

MOMP

mitochondrial outer membrane permeabilization

SMAC

second mitochondria-derived activator of caspases

TARF

TNFR-associated factor

APAF1

apoptotic protease-activating factor 1

BIR

baculovirus IAP protein repeat

PUMA

p53 upregulated modulator of apoptosis

MLKL

mixed lineage kinase domain-like

ULK

UNC51-like kinases

FIP200

FAK family kinase-interacting protein of 200 kDa

ATG

autophagy-related genes

mTORC1

mechanistic target of rapamycin complex 1

AMPK

AMP-activated protein kinase

Vps34

vacuolar protein sorting 34

LC3

microtubule-associated protein light chain 3

PE

phosphatidylethanolamine

ADCD

autophagy-dependent cell death

MEF

mouse embryonic fibroblasts

GBA1

GRB2-associated-binding protein 1

RIPK1

receptor interacting serine/threonine kinase 1

FADD

Fas-associating protein with death domain

IL

interleukin

DRAM

damage-regulated autophagy modulator

iDISC

intracellular DISC

c-FLIP

cellular FLICE (FADD-like IL-1β-converting enzyme)-inhibitory protein

FADD-like

IL-1β-converting enzyme-inhibitory protein

NF-kB

nuclear factor kappa B

GABARAP-L1

gamma-aminobutyric acid receptor-associated protein-like 1

BCLAF1

Bcl-2-associated transcription factor 1

CLL

chronic lymphocytic leukemia

SLL

small lymphocytic lymphoma

AML

acute myeloid leukemia

NAF-1

nutrient-deprivation autophagy factor-1

DAPK

death-associated protein kinase

STK4

serine/threonine kinase 4

MM

multiple myeloma

MDM2

murine double minute 2

RITA

reactivation of p53 and induction of tumor cell apoptosis

BIR

baculovirus IAP protein repeat

TRADD

TNFR1-associated death domain

Apo2L

Apo2 ligand

TRAIL

TNF-related apoptosis-inducing ligand

scFv

single-chain fragment variable

FOXO3a

forkhead box protein O3a

CAR

chimeric antigen receptors

MHC

major histocompatibility complex;

Author Contributions

This manuscript was written through contributions of all authors. All authors have given approval to the finalBcl-2 inhibitors are binding to off-target proteins version of the manuscript.

This work was supported by the Epstein Family Foundation, a Sanford Burnham Prebys National Cancer Institute Cancer Center Support Grant P30 CA030199, a Larry L. Hillblom Foundation Grant 2019-A-005-NET, and a 2019 Pancreatic Cancer Action Network Translational Research Grant 19-65-COSF to NDPC. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

The authors declare no competing financial interest.