Department of Pharmaceutical Engineering, Key Laboratory of Systems Bioengineering, Ministry of Education, School of Chemical Engineering and Technology, Tianjin University, Nankai District, Tianjin, China
Correspondence to: Zhe Liu, Department of Immunology, Tianjin Key Laboratory of Medical Epigenetics, 2011 Collaborative Innovation Center of Tianjin for Medical Epigenetics, Tianjin Research Center of Basic Medical Sciences, Tianjin Medical University, 22 Qixiangtai Road, Tianjin 300070, China. E-mail: email@example.com or Zhenyi Ma, Department of Biochemistry and Molecular Biology, School of Basic Medical Sciences, Tianjin Medical University, 22 Qixiangtai Road, Tianjin 300070, China. Fax: 86-22-83336533. E-mail: firstname.lastname@example.org
Autophagy is a lysosome-dependent process in which enzymatic degradation and recycling of cytosolic components occurred due to stressful conditions (Levine, 2007). It is initiated by forming a double-membrane structure, the autophagosome, which sequesters cytoplasmic proteins and organelles such as mitochondria or endoplasmic reticulum (ER). Autophagosomes then fuse with lysosomes, generating autophagolysosomes, which degrade the entrapped components to produce small molecules (e.g., amino acids, lipids, and nucleic acids). Autophagy plays important but controversial roles in pathophysiological processes. A basal level of autophagy prevents the accumulation of protein aggregates and damaged subcellular organelles, providing a more stable inner environment (Pyo et al., 2012). More importantly, autophagy can function as an important way to adapt to outer stresses, such as the extracellular matrix (ECM) detachment and starvation, by maintaining cellular homeostasis (Fung et al., 2008; Shang et al., 2011). Furthermore, autophagy can benefit the rapid growth of tumor cells by providing the small molecules (e.g., amino acids, lipids, and nucleic acids) for biosynthesis. In contrary, autophagy has been suggested to occur before the death of cancer cells that were subjected to drugs or gene disruption (Tsuchihara et al., 2009; Gonzalez et al., 2012). Accordingly, the regulatory mechanisms of autophagy are fundamental for the clinical treatment of tumor metastasis as well as those of anoikis.
The parenchymal cells require ECM adhesion to survive, to sustain their normal shape and polarity, and to function. When lost attachment from ECM or forced to suspend in a fluid environment, cells undergo anoikis, a detachment-induced apoptosis (Frisch and Francis, 1994; Frisch and Ruoslahti, 1997). The cells detached from their original proper sites would be eliminated, thus preventing dysplastic growth (Guadamillas et al., 2011). Hence, anoikis is essential in the maintenance of the physical condition of the whole organism and sustaining tissue homeostasis and development. For example, anoikis promotes the luminal clearance during mammary gland development, which to some extent explains the formation of our inner organ (Debnath, 2008). Moreover, a bulk of evidences has proved that anoikis resistance is intimately involved in the tumorigenesis and metastasis of various cancers, such as lung, colon, and prostate cancers (Horbinski et al., 2010; Broustas et al., 2012; Taddei et al., 2012; Vigil et al., 2012; Zheng et al., 2012). Therefore, the studies and summarizations of the regulatory mechanism(s) of anoikis are fundamental to the intervention of tumor metastasis.
The physiological and pathological roles of both anoikis and autophagy have been described in previous state-of-art reviews in detail (Bergmann, 2007; Chen and Karantza-Wadsworth, 2009). However, a systematic knowledge of the underlying regulatory mechanisms of how cells (especially solid tumor cells) decide to survive through autophagy or to die through anoikis still remains clouded. Anoikis and autophagy may be regulated by the similar stimuli and interact with each other. Herein, we focus on the integration of anoikis and autophagy in solid tumor cells and provide some insight for the regulation of the balance between anoikis and autophagy, which will shed a light on the antitumor outcome.
AUTOPHAGY RELATED SIGNALING PATHWAYS IN SOLID TUMORS
Autophagy can either benefit or harm the cells according to the biological contexts (e.g., cell type, energy metabolism level, redox state, and extracellular microenvironment). Debates over the dual roles of autophagy have lasted for years (Levine, 2006; Chen and Karantza-Wadsworth, 2009; Mathew and White, 2011; Wirawan et al., 2012). For example, the metabolic stress, such as nutrient deprivation, can affect autophagy level either ways (Tsuchihara et al., 2009; Lee et al., 2012). In one way, autophagy fulfills the requirement of material and energy to maintain the metabolic activity and cell viability (Mathew and White, 2011). It has also been reported that the inhibition of mammalian target of rapamycin complex 1 (mTORC1) induces autophagy as a cytoprotective mechanism by digesting subcellular organelles as the energy sources (Zhao et al., 2011). In the other way, the autophagic cell death (ACD) has also been observed in metabolic stressed cells. The Atg7 binding to p53 regulates the transcription of the gene encoding p21CDKN1A, a cell cycle inhibitor, thus blocking the starved mouse cells in the arrest phase (Lee et al., 2012). In addition to metabolic stress, the cellular redox state regulates autophagy both positively and negatively. Autophagy can selectively remove the damaged mitochondria, termed mitophagy, thus reducing the accumulation of reactive oxygen species (ROS) (Tolkovsky, 2009), whereas an excessive amount of ROS, especially the superoxide, effectively promotes the formation of autophagosomes. This negative feedback loop between ROS and autophagy sustains the cellular redox homeostasis and cell survival (Chen et al., 2009). In contrary, the accumulation of endogenous ROS causes the formation of autophagosomes which selectively degrade the catalases, in turn elevating the ROS level, and leading to the ACD (Yu et al., 2006).
Additionally, autophagy seems to be more complex when referring to cancer biology. The pro-survival and the pro-death mechanisms simultaneously exist in autophagic cancer cells, but only one of them is to be triggered by the context-dependent regulation of autophagy (Shen and Codogno, 2011; Denton et al., 2012). Herein, we demonstrate some representative examples of its controversial roles. On one hand, autophagy facilitates the survival and growth of the cancer cells under stressful conditions, such as serum starvation, low attachment, and hypoxia (Levine, 2007; Debnath, 2008; Li et al., 2011; Mohan et al., 2011; Hu et al., 2012a). Indeed, more evidences have arisen to support the hypothesis that autophagy contributes to cell survival circumstantially. The serum starvation induces autophagy and inhibits apoptosis in SH-SY5Y cells through the upregulation of NF-κB and Bcl-2 and downregulation of Bax and caspase 3, respectively (Mohan et al., 2011). Besides, lysosomal associated transmembrane protein LAPTM4B regulates the properties of lysosomal membrane. The enhancement of LAPTM4B significantly increases the autophagic flux and promotes breast tumor growth in vivo against the environmental metabolic stress (Li et al., 2011). The hypoxia-inducible factor-1α (HIF-1α)/AMP-activated protein kinase (AMPK) pathway mediates the hypoxia-induced autophagy in glioblastomas and prevents apoptotic cell death in vitro (Hu et al., 2012a). The in vitro three-dimensional (3D) cell culture model shows that autophagy can inhibit apoptosis and promote cell survival during anoikis and the lumen formation (Debnath, 2008). In hepatoblastoma (HB), a rare human liver cancer, the Beclin-1 and Atg5-dependent PI3K signaling pathway is supposed to be essential for the survival of HB cells in tolerance to chemotherapy (Chang et al., 2011).
On the other hand, autophagy strikingly limits the survival and growth of cancer cells. In resistant to the stressful tumor microenvironment, the autophagic cells digest misfolded proteins and damaged mitochondria to reduce, at least in part, the malignancy of cancer cells (Levine, 2007). Mutation of Atgs in numerous tumor types indicates that autophagy deficiency may positively correlate to tumorigenesis (Wirawan et al., 2012). The mutation of Beclin-1, a critical component that induces the formation of autophagosomes in mammalian cells, promotes the tumorigenesis through inhibiting autophagy. Beclin-1 is mono-allelically deleted in a majority of human sporadic breast cancer, while normal epithelial breast cells have higher levels of Beclin-1 than breast cancer cells (Qu et al., 2003). UVRAG, a Beclin-1-binding protein, is frequently frame-shift mutated in human cancers and enhances the binding of Beclin-1 and PI3K to promote autophagy and inhibit tumorigenesis (Liang and Jung, 2010). Several oncogenes such as K-Ras and Akt have been reported to negatively regulate autophagy, and to some extent propose the possibility that autophagy may adversely contribute to the malignant phenotype of cancer cells (Levine, 2006). The oncogenic H-RasV12 has been reported to mediate ACD through the induction of Noxa, a member of the Bcl-2 family, and Beclin-1, while silencing of Noxa and Beclin-1 rescues the clonogenic survival (Elgendy et al., 2011). Most recently, Chen et al. (2012) has reported an interaction between autophagy and the PI3K/mTOR pathway, which limits the proliferation of 3D-cultured cells by autophagy antagonizing the oncogenic PI3K transformation. The first identified tumor suppressor Rb has been reported to induce autophagy and ensure the stability of the genome and reduces tumorigenesis (Jiang et al., 2010). The cytoprotective enzyme Heme oxygenase-1 (HO-1) can reduce the level of autophagy and assist tumor cells to survive the chemotherapy (Banerjee et al., 2012). The apoptotic stimuli can induce ACD when Bax and Bak, two regulators of apoptosis, are double knockout. This process is mediated by the activation of c-Jun NH2-terminal kinase (JNK) as a downstream signal of autophagy (Shimizu et al., 2004). Most recently, the histone deacetylase 6 (HDAC6) has been addressed as a tumor suppressor by activating JNK/Beclin-1 pathway to induce ACD of hepatocellular carcinoma (Jung et al., 2012). Interestingly, autophagy has also been proved to play a role in tumor dormancy, which is critical to the recurrence of cancers (Gewirtz, 2009). The aplasia Ras homolog member I (ARHI), a tumor suppressor gene, causes ACD of human ovarian cancer cells in vitro. Considering the tumor microenvironment, however, ARHI-induced autophagy can be switched to a mechanism of tumor cell survival and dormancy. Hence, manipulating the ARHI-induced autophagy could be potentially beneficial to breast and ovarian cancers therapy (Lu et al., 2008; Zou et al., 2011). In summary, these researches and reviews indicate a strong implication that targeting anoikis resistance and autophagy in the clinical therapy may produce efficient approaches in the blockage of tumorigenesis and metastasis.
ANOIKIS RELATED SIGNALING PATHWAYS IN SOLID TUMORS
Anoikis resistance leads to the anchorage-independent (AI) growth and proliferation of cancer cells, promotes the cells to move to and colonize distal organs by blood circulation, and therefore facilitates the metastasis of solid tumor cells (Gassmann and Haier, 2008). Anoikis resistance in metastatic cancer is tightly regulated by diverse signaling pathways. Early researches have focused on the interactions between anoikis and known apoptosis signaling pathways. Wang et al. (2003) demonstrated that the cellular decision for anoikis of mammary epithelial cells occurs in the absence of caspases activation, but relies on the cleavage of Bid, a member of the Bcl-2 family. Bim, another Bcl-2 family member, is rather essential for anoikis in normal cells. BimEL and BimL, which are sequestered in the cytoskeleton-associated motor complex, can be released in response to apoptotic stimuli and translocated to mitochondria on detachment from the ECM. There Bim interacts with Bcl-XL and inhibits its pro-survival function (Reginato et al., 2003). Recently, a Bcl-2 modifying factor (BMF) has been identified to modulate the homeostasis of human intestinal epithelial cell by blocking the phosphorylation of Akt to induce anoikis (Hausmann et al., 2011).
Oncogenes including Ras, neurotrophic tyrosine kinase receptor (TrkB), phosphatidylinositol 3-kinase (PI3K)/Akt, integrin-linked kinase (ILK), and focal adhesion kinase (FAK) are important regulators in anoikis resistance (Attwell et al., 2000; McFall et al., 2001; Duxbury et al., 2004; Zheng et al., 2012). For example, activated Ras can impair anoikis through the downregulation of Bak and the upregulation of Bcl-XL, cIAP2, and XIAP (Rosen et al., 1998, 2000; Douma et al., 2004; Liu et al., 2005). It was also demonstrated that the promotion of the viability of detached malignant intestinal epithelial cells involves the sequential activation of Ras, RhoA, and protease Calpain, leading to the degradation of Beclin-1, a key regulator of autophagy (Yoo et al., 2010). Moreover, the mutation of Rb1 inactivates c-Raf/Erk and phosphorylates Akt at site serine 473 by forming mTOR complex 2 (mTORC2) in mouse embryonic fibroblasts (MEFs), leading to anoikis resistance (El-Naggar et al., 2009). Similarly, activation of the PI3K/Akt2 pathway mediates anoikis resistance in prostate carcinoma cells with reduced mitochondrial DNA (mtDNA) content (Moro et al., 2009). Overexpression of TrkB in ovarian cancer, pancreatic cancer, and neuroblastoma has been reported to be involved in anoikis resistance (Douma et al., 2004; Yu et al., 2008). The function of TrkB in anoikis resistance and epithelial-to-mesenchymal transition (EMT) depends on the activation of Akt, Twist-Snail, and Zeb1 pathways (Douma et al., 2004; Smit et al., 2009; Smit and Peeper, 2011). Integrins, a family of focal adhesion proteins locating both in plasma membranes of cells and in ECM, are essential in sustaining the adhesion of cells to the surrounding tissue. Lack of integrin ligation to ECM decreases the activities of ILK and FAK, subsequently impairs the downstream signals including activation of Rac1 and RhoA, and finally leads to anoikis (Zugasti et al., 2001; Xia et al., 2004; Ma et al., 2007). Consistently, the activation of ILK causes anoikis resistance by increasing the expression of Snail and lymphoid enhancer-binding factor-1 (LEF-1) (Becker-Santos et al., 2012). The phosphorylation of FAK by protein tyrosine kinase 6 (PTK6) also acts as an important mechanism in AI proliferation (Zheng et al., 2012). The ROS, which is constitutively high in metastatic cells, activates Src and further phosphorylates EGFR, subsequently leading to the phosphorylation of Akt and Erk, and finally, promoting the survival of cancer cells when losing ECM contact (Giannoni et al., 2008; Kim et al., 2012). The aforementioned researches suggest that the oncogenic regulations majorly positively contribute to the resistance of anoikis.
In addition, several tumor suppressors have also been reported to be associated with anoikis. Lines of evidences suggested that p53, an important tumor suppressor, is involved in anoikis. For instance, MCF-7 cells with overexpressed Caveolin-1 acquire the ability of anoikis resistance by inactivating p53 and elevating the expression of insulin-like growth factor-I (IGF-I) (Ravid et al., 2005). Functional salt-inducible kinase 1 (SIK1) links serine/threonine kinase 11 (LKB1), another tumor suppressor, to p53-dependent anoikis and blocks distal metastasis (Ji et al., 2007; Cheng et al., 2009). Reversely, anoikis can also affect the post-transcriptional modification of p53. For example, anoikis triggers the ubiquitination of p53/Mdm2 complex and the degradation of p53 in human primary ligament fibroblasts (Ghosh et al., 2010). The deregulation and/or degradation of E-cadherin, a tumor suppressor and marker of EMT, are tightly correlated to anoikis resistance and lead to oncogenic EMT (Kumar et al., 2011). Notably, p66Shc, an isoform of Shc adaptor proteins and a focal adhesion-associated protein, has been reported to confer anoikis by activation of RhoA in H69 and H209, two lines of human small cell lung cancer (Ma et al., 2010). Collectively, all these evidences illustrate that the anoikis resistance in solid tumor cells is tightly regulated by the oncogenic and the tumor suppressing signaling pathways. Furthermore, the tight regulation of anoikis by the oncogenes and tumor suppressors suggests the potential of clinical therapy targeting anoikis in solid tumor metastasis.
BALANCE BETWEEN AUTOPHAGY AND ANOIKIS RESISTANCE IN SOLID CANCER CELLS
Accumulating evidences show links between autophagy and anoikis resistance during the metastasis of solid tumor cells. When normal cells detach from the ECM, they undergo anoikis. Autophagy is also induced following ECM detachment (Debnath, 2008). During this process, anti-apoptotic signals are triggered in transformed cells, allowing the survival of these cells provided they reattach to the ECM in a timely manner (Fung et al., 2008). The cancer cells undergo malignant transformation, and can be transferred through blood or lymphatic circulation to a distant location. Autophagy may cooperate with the stimulation signals to bypass anoikis and promote metastasis (Kenific et al., 2010). In a 3D culture model of MCF-10A, Atg5 or Atg7 knockdown induced the luminal apoptosis, indicating that autophagy promotes epithelial cell survival during anoikis (Fung et al., 2008). However, the delicate balance between anoikis and autophagy, as well as the regulatory mechanisms are still unclear.
As indicated above, ROS is one of these node signals. ROS was suggested to stimulate cell survival during anoikis as early as several decades ago. The detachment of human umbilical vein endothelial cells (HUVECs) from the ECM leads to the Rac1-dependent production of ROS, which inhibits anoikis through sequential activation of Src and EGFR/Erk1/2 (Szatrowski and Nathan, 1991). Until recently, ROS has been suggested to regulate the energy level under stressful conditions and promote autophagy to resist anoikis (Chen et al., 2009). Epithelial cells detached from the ECM produce cytotoxic ROS and induce autophagy to survive. Such opposing effects coordinately decide the fate of detached cells. The endoplasmic reticulum kinase (PERK) enhances autophagy to block anoikis and promotes the survival. Consistently, the levels of phosphorylated PERK and autophagosome marker microtubule-associated protein 1 light chain 3 beta (MAP1LC3B or LC3B) is increased in human breast ductal carcinoma in situ (DCIS), compared with normal breast tissues (Avivar-Valderas et al., 2011). Except for ROS, other tumor microenvironmental stresses such as the nutrients deprivation also stimulate the detachment and migration of malignant cancer cells. Therefore, metabolism related mechanisms may involve in the coordination of anoikis and autophagy. Autophagy facilitates the glycolysis as a core metabolic function which is required in the oncogenic Ras mediated AI proliferation and transformation (Lock et al., 2011). AMPK, maintaining the cellular ATP level, is significantly activated during detachment, thus promoting autophagy, although the precise upstream signaling pathway is still unclear (Fung et al., 2008).
Integrins, as anoikis regulating protein, are involved in the signaling pathways of autophagy (Lock and Debnath, 2008). EGFR is downregulated in epithelial cells detaching from ECM, to reduce mTOR and promote autophagy (Reginato et al., 2003), while the EGF withdrawal can also effectively induce autophagy during anoikis of breast epithelial cells (Fung et al., 2008). In addition, downregulation of mTOR is also related to the inhibition of the activity of FAK, an essential component of adhesion-related signaling and anoikis (Gan et al., 2006). Moreover, the oncogene Akt also suppresses the anoikis as well as the autophagy in tumorigenesis through the phosphorylation of Beclin 1 and the sequential formation of an autophagy-inhibitory Beclin 1/14-3-3/vimentin intermediate filament complex (Wang et al., 2012). Detachment induced phosphorylation of eukaryotic initiation factor 2 alpha (eIF2a) and PERK activation are associated with the autophagosome formation (Kouroku et al., 2007).
Autophagy plays multiple functions according to the biological contexts (e.g., the cellular energy metabolism, extracellular microenvironment, and the detachment state). As shown in Fig. 1, autophagy and anoikis can function as important ways to adapt to outer stresses in normal epithelial cells by maintaining cellular homeostasis. Importantly, autophagy may also serve as a pro-metastatic function to promote tumor cell survival and metastasis by bypassing anoikis. As accumulating evidences have suggested a delicate balance between autophagy and anoikis, researchers realize that autophagy may benefit the clinical treatment of tumor as complementary strategies.
TARGETING ANOIKIS RESISTANCE AND AUTOPHAGY IN CANCER THERAPY
The screen and identification of the therapeutic agents targeting anoikis resistance and autophagy in malignant solid tumors are emerging as a hot area in cancer medical research. The signaling pathways involved in bypassing anoikis in solid tumors are blocked to increase the susceptibility of the cells to anoikis. For example, ROS mediates the downregulation of Bcl-2 and sensitizes anoikis induced by curcumin, a major bioactive ingredient of turmeric (Pongrakhananon et al., 2010). Screening of chemicals shows that Anisomycin sensitizes cells to anoikis by decreasing FLIP, an inhibitor of caspase 8 (Mawji et al., 2007). Combination of WZ4002, ABT-263, and trichostatin A (TSA), inhibitors of EGFR, Bcl-2, and histone deacetylase, respectively, decreases the viability of suspended lung adenocarcinoma more effectively (Sakuma et al., 2012).
To date, it remains largely unclear how tumor cells are regulated to trigger the ACD or survival. Nevertheless, drugs targeting autophagy is still an emerging field of cancer therapy. Considering the important but complex role of autophagy in cancer, autophagy has been regarded as a complementary target to treat tumor. In fact, several anticancer drugs, such as EGFR inhibitor cetuximab (Li and Fan, 2010), HDAC inhibitors vorinostat as well as OSU-HDAC42 (Liu et al., 2010) have been demonstrated to induce autophagy (Yang et al., 2011). The specific mechanism of these drugs may be involved in excessive autophagy leading to cell death which are effectively enhanced by pharmacological inhibitors of autophagy, including the early stage inhibitors 3-methyladenine (Kondo and Makita, 1997), wortamannin and LY294002, and late stage inhibitors chloroquine, hydroxychloroquine (HCQ), bafilomycin A1 and monensin (Yang et al., 2011). Inhibition of caspases, the apoptosis executor, results in the elevation of autophagy markers LC3B-II accumulation. Meanwhile, inhibition of the late phase autophagy by chloroquine can sensitized metastatic cells to flavonoid luteolin treatment and induce apoptosis (Verschooten et al., 2012). Oridonin can also induce prostate cancer PC-3 cells ACD (Ye et al., 2012). In contrary, the administration of multikinase inhibitor sorafenib promotes the autophagosome formation and increases the probability of the hepatoma cells survival, suggesting that complementary strategies are urgently needed to expand the clinical application of sorafenib (Shimizu et al., 2012). For example, BO-1051, an N-mustard linked with a DNA-affinic molecule, has been reported to induce apoptosis and autophagy in hepatocellular carcinoma. However, the induction of autophagy can provide metabolic substance to counteract the apoptosis of cancer cells. The inhibition of autophagy facilitates the anticancer effect of BO-1051 (Chen et al., 2011). Combination therapy of the synthetic retinoid N-(4-hydroxyphenyl) retinamide and the flavonoid apigenin can induce apoptosis and inhibit autophagy in human malignant neuroblastoma cells subjected to serum starvation by upregulating Bax and downregulating Bcl-2 to trigger mitochondrial-dependent caspase-3 activity (Mohan et al., 2011). Another plant extract anthocyanin shows antioxidant and anticancer activities by scavenging hydrogen oxide and inducing apoptosis in hepatocellular carcinoma. The early-phase autophagy inhibitor 3-methyladrine enhances its therapeutic activity (Longo et al., 2008).
CONCLUDING REMARKS AND PERSPECTIVES
Over the past years, anoikis resistance and autophagy have been understood in pathological and physiological conditions. However, the pathways leading to activation of anoikis and autophagy are still far from the total understanding. Recent works have demonstrated that the microRNAs and epigenetic regulation of anoikis and autophagy in malignant solid tumors are attractive research fields with important physiologic consequences (Oh et al., 2008; Parsons et al., 2009; Penna et al., 2011; Yang et al., 2011; Gammoh et al., 2012). Although the “inside-outside-in” hypothesis does much to integrate anoikis and autophagy areas of study, a number of broad questions remain. For instance, how hematopoietic cells, which are both anchorage independent and anoikis insensitive, defend against anoikis during their development? What is the molecular machinery to regulate anoikis resistance and autophagy? How metastatic cells determine die or survive? However, no matter what role it plays in tumor survival, autophagy would be considered as a drug target based on the significant molecules essential for anoikis. Also, anoikis resistance inhibition by complementary approaches limits metastasis, determining the prognosis of cancer patients, due to the better inhibition of dissemination of the disease. Traditionally, the basal levels of autophagy keep survival pathway and enhanced autophagy facilitates reversible stress such as drug resistance during cancer treatments (Hu et al., 2012b). However, different types of cell death such as apoptosis, ACD, and necrosis as well as the unknown death response can also be triggered by death stimuli (Table 1). Complementary approaches that impair anoikis resistance and autophagy might limit metastasis formation. Therefore, elucidating mechanisms that integrate autophagy and anoikis resistance in tumor cells is a fascinating research area of cancer metastasis with substantial therapeutic potential.
Table 1. Cell Survival and Death Types based on Autophagic Flux and Efflux
Basal levels of autophagy keep survival pathway and enhanced autophagy facilitates reversible stress such as drug resistance. However, other types of cell death such as apoptosis, autophagic cell death, and necrosis as well as the unidentified death response(s) can also be triggered by death stimuli. Consequently, the interplay between autophagy and cell death pathways is extremely complex and the decision taken by cancer cells will vary under different circumstances.
Basal survival signals
Basal level of autophagic flux
Basal level of autophagic efflux
Reversible stress signals
Enhanced autophagic flux
Enhanced autophagic efflux
Survival under reversible stress
Lethal stress signals
Enhanced autophagic flux
Decreased autophagic efflux
Cell death with autophagy
Lethal stress signals
Decreased autophagic flux
Enhanced autophagic efflux
Unidentified death response
Lethal stress signals
No autophagic flux change
No autophagic efflux change
Apoptosis or necrosis
The authors acknowledge Lance S. Terada in the Department of Internal Medicine, University of Texas Southwestern Medical Center at Dallas for his continuous support and many stimulating discussions.