Autophagy in 5-Fluorouracil Therapy in Gastrointestinal Cancer: Trends and Challenges.

OBJECTIVE: 5-Fluorouracil (5-FU)-based combination therapies are standard treatments for gastrointestinal cancer, where the modulation of autophagy is becoming increasingly important in offering effective treatment for patients in clinical practice. This review focuses on the role of autophagy in 5-FU-induced tumor suppression and cancer therapy in the digestive system. DATA SOURCES: All articles published in English from 1996 to date those assess the synergistic effect of autophagy and 5-FU in gastrointestinal cancer therapy were identified through a systematic online search by use of PubMed. The search terms were "autophagy" and "5-FU" and ("colorectal cancer" or "hepatocellular carcinoma" or "pancreatic adenocarcinoma" or "esophageal cancer" or "gallbladder carcinoma" or "gastric cancer"). STUDY SELECTION: Critical reviews on relevant aspects and original articles reporting in vitro and/or in vivo results regarding the efficiency of autophagy and 5-FU in gastrointestinal cancer therapy were reviewed, analyzed, and summarized. The exclusion criteria for the articles were as follows: (1) new materials (e.g., nanomaterial)-induced autophagy; (2) clinical and experimental studies on diagnostic and/or prognostic biomarkers in digestive system cancers; and (3) immunogenic cell death for anticancer chemotherapy.
RESULTS: Most cell and animal experiments showed inhibition of autophagy by either pharmacological approaches or via genetic silencing of autophagy regulatory gene, resulting in a promotion of 5-FU-induced cancer cells death. Meanwhile, autophagy also plays a pro-death role and may mediate cell death in certain cancer cells where apoptosis is defective or difficult to induce. The dual role of autophagy complicates the use of autophagy inhibitor or inducer in cancer chemotherapy and generates inconsistency to an extent in clinic trials.
CONCLUSION: Autophagy might be a therapeutic target that sensitizes the 5-FU treatment in gastrointestinal cancer.

INTRODUCTION

The antimetabolite 5-fluorouracil (5-FU)-based combination therapies have been standard treatments for many patients diagnosed with gastrointestinal cancer in the past decades. However, resistance to 5-FU together with its usage has become a common issue, and this has been recognized as a cause of cancer therapy failure. The resistance to anticancer drugs can be attributed to a wide variety of mechanisms including tumor cell heterogeneity, drug efflux, and other periods of tumor microenvironment stress-induced genetic or epigenetic alterations as a cellular response to drug exposure.[12] Among these mechanisms, the adaptation of tumor cell to anticancer drug-induced microenvironment stresses is a vital cause of chemotherapy resistance.

Macroautophagy (hereafter denoted simply as autophagy) is a cell survival pathway involving the degradation of cytoplasmic constituents, and the recycling of adenosine triphosphate and essential building blocks for the maintenance of cellular biosynthesis during nutrient deprivation or metabolic stress.[3] For tumor cells, autophagy is a “double-edged sword” since it can be either protective or damaging, and the effects may change during tumor progression.[45] The dual role of autophagy in tumor development remains unclear. Current evidence supports the idea that autophagy eliminates damaged organelles and recycle macromolecules, thus functioning as a tumor suppressive mechanism, particularly during malignant transformation and carcinogenesis.[678] However, in established tumors, cancer cells may need autophagy for cytoprotection to cope with their hostile microenvironments such as nutrient deprivation, hypoxia, the absence of growth factors, and the presence of chemotherapy or some targeted therapy mediated resistances to anticancer therapies.[910] Consequently, the combination of autophagy inhibitors with chemotherapy drug has become more attractive in cancer therapy. Most studies have indicated that 5-FU-treatment-induced autophagy of cancer cells in vivo,[111213] and inhibiting autophagy potentiated the anticancer effects of 5-FU. Inhibitory effect of chloroquine (CQ) and its derivative hydroxychloroquine (HCQ) on autophagy in preclinical models and their safety in clinical trials have been approved by the Food and Drug Administration (FDA); it might be possible to treat certain cancer types without the need for phase I studies.

Here, the association between autophagy and 5-FU chemotherapy in various gastrointestinal cancer is summarized, the mechanisms of autophagy in 5-FU chemotherapy are reviewed, and the emerging questions of their promising potential as therapeutic targets for the treatment of gastrointestinal cancer are also highlighted.

AUTOPHAGY PARADOX IN THERAPEUTIC PURPOSES IN CANCER

The pioneer work by Liang et al. embraced the discovery that one copy of the Beclin-1 gene is deleted in some specimens of human breast, ovarian, and prostate tumors[14] suggesting that autophagy may play an anti-tumor role in tumorigenesis. During the following two decades, a large number of autophagy-related genes were found at a reduced expression level or even totally lost in certain types of cancer cells,[151617181920] supporting the conclusion that basal autophagy may act as a cellular housekeeper to eliminate damaged organelles and recycle macromolecules, and thus protect against cell transformation in the early phase of tumorigenesis. Later, as tumors grew, existing evidence highlighted an indispensable role for autophagy in tumorigenesis.[21] In solid tumors, prior to angiogenesis, autophagy defection induces long-term and chronic inflammation in cancer cells undergoing a continuous low-level of necrosis. Alternatively, autophagy-competent cancer cells could survive this nutrient-limited and low oxygen microenvironment by activating autophagic pathways with both no death and no proliferation. This ability to cope with stress is also useful to cancer cells that disseminate and metastasize.[22] Hence, the paradox leads to a similar contradictory response of autophagy in tumor following anticancer treatments. On one hand, autophagy is activated as a protective mechanism to mediate the acquired resistant phenotype of some cancer cells during chemotherapy. On the other hand, autophagy may also function as a death executioner to induce autophagic cell death (a form of physiological cell death that is contradictory to apoptosis). Accordingly, two therapeutic strategies were currently used in the clinical trials: One was to inhibit the cytoprotective function of autophagy to improve the killing efficacy of chemotherapy drugs or resensitize the chemoresistant tumor cells to drugs; the other was to induce autophagic cell death in the apoptosis-defective tumor cells, which showed high resistance to apoptosis by activating autophagic pathways.

AUTOPHAGY-MEDIATED CHEMORESISTANCE TO 5-FLUOROURACIL IN GASTROINTESTINAL CANCER

Over the past several years, the selection of chemotherapeutic regimens has expanded greatly due to the development of molecular targeted therapy.[23] Among varieties of those drugs, 5-FU remains the most popular and has been widely used for gastrointestinal cancer for about 40 years.[24] However, the resistance to 5-FU which might result in therapy failure has become a common clinical issue in the treatment of patients with such disease. Regarding the chemoresistance, 5-FU treatment also induces autophagic responses in multiple types of gastrointestinal cancer cells [Figure 1].[252627282930] So far, the molecular mechanisms of 5-FU-induced autophagy remain poorly defined. Many studies have examined the synergistic effect of autophagy and 5-FU in colorectal cancer, hepatocellular carcinoma (HCC), pancreatic adenocarcinoma, esophageal cancer, gallbladder carcinoma (GBC), and gastric cancer [Table 1]; some hold great promise and are currently being investigated within the context of phase I and phase II clinical trials [Table 2].

Figure 1

Autophagy is considered a key mechanism in the development of resistance to 5-fluorouracil. 5-Fluorouracil-based combination therapies are standard treatments for many patients diagnosed with various gastrointestinal tumors. Since autophagy is a mechanism of chemoresistance to 5-fluorouracil, several inhibitors of autophagy, or interference of certain genes will promote sensitivity to 5-fluorouracil in gastrointestinal cancer.

Table 1

Autophagy in response to 5-FU in different types of gastrointestinal cancer

Cell lines (cancer type)	Mediating autophagy methods (target)	Regulating mechanisms	References
HT-29 (colorectal cancer)	CQ (lysosome)	p21Cip1, p27Kip1, and CDK2	[31]
DLD-1 (colon cancer)	CQ (lysosome)	p27, p53, CDK2, and cyclin D1	[32]
Colon 26 (colon cancer)	CQ (lysosome)	Bad and Bax	[13]
HCT116, HT-29 (colon cancer)	3-MA (PI3K III), CQ (lysosome), RNAi (Beclin-1, Atg5)	Bcl-2/JNK pathway	[33]
HT-29, colon 26 (colon cancer)	3-MA (PI3K III)	Bcl-xL, cytochrome c/caspase-3/PARP pathway	[34]
HCT116, DLD-1 (colon cancer)	3-MA (PI3K III), RNAi (Atg7)	Bcl-xL, p53-AMPK-mTOR	[11]
HaCaT and HCT116 (colon cancer)	SB203580 and RNAi (p38MAPK)	MAP2K, MAPK kinase-3, and MAPK kinase-6	[35]
SMMC-7721, Hep3B, HepG2 (HCC)	3-MA (PI3K III), CQ (lysosome), RNAi (Beclin-1)	Unknown	[12]
HepG2, SMMC7721 (HCC)	Pifithrin-α and RNAi (P53)	ROS	[28]
PANC-1, BxPC-3 (pancreatic adenocarcinoma)	CQ (lysosome) and wortmannin (PI3K/PLK1)	Unknown	[27]
OE21, KYSE450, OE19, OE33 (esophageal cancer)	RNAi (Beclin-1, Atg7)	Unknown	[30]
GBC-SD, SGC-996 (gallbladder carcinoma)	3-MA, CQ, RNAi (Atg5, Atg7)	Unknown	[26]
SGC7901 (gastric cancer)	RNAi (PI3K III)	Unknown	[36]
SNU-5 (gastric cancer)	3’UTR luciferase reporter (Beclin-1)	MiR-30	[37]
SGC-7901 (gastric cancer)	Bafilomycin A1 (vacuolar H+ATPases)	Unknown	[38]

CQ: Chloroquine; ROS: Reactive oxygen species; HCC: Hepatocellular carcinoma.

Table 2

Examples of clinical trials involving chloroquine or hydroxychloroquine for the treatment of gastrointestinal cancer

Condition	HCQ combined therapy	Phase	Clinical trial ID
Liver cancer	TACE	I/II	NCT02013778
Advanced solid tumors	Vorinostat	I	NCT01023737
Colorectal cancer	Vorinostat	II	NCT02316340
Advanced solid tumors, melanoma, prostate or kidney cancer	MK0394 (Akt inhibitor)	I	NCT01480154
Pancreatic cancer	Proton or Photon beam radiation therapy and capecitabine	II	NCT01494155
Pancreatic cancer	Gemcitabine/abraxane	I/II	NCT01506973
Advanced or metastatic cancer	Sirolimus/vorinostat	I	NCT01266057
Pancreatic cancer	Gemcitabine hydrochloride and paclitaxel albumin-stabilized nanoparticle formulation	II	NCT01978184
Colorectal cancer	Fluorouracil, leucovorin calcium, oxaliplatin, and bevacizumab	I/II	NCT01206530
Colorectal cancer	Bevacizumab and combination chemotherapy	II	NCT01006369
Metastatic solid tumors	Temsirolimus	I	NCT00909831
Refractory or relapsed solid tumors	Sorafenib	I	NCT01634893

The content of Table 2 was from http://www.cancer.gov/clinicaltrials. HCQ: Hydroxychloroquine; TACE: Transarterial chemoembolization.

Colorectal cancer

5-FU is a cornerstone in chemotherapy of advanced colorectal cancer;[39] improved combinations of 5-FU with irinotecan; or oxaliplatin have progressively increased tumor response as well as the median survival time of patients with unresectable tumor.[40] Previous studies have demonstrated that inhibition of autophagy augments anticancer effects of 5-FU in colorectal cancer,[1331] and autophagy responds to 5-FU through the regulation of Bcl-2 and Bcl-xL.[3433] Bcl-2 inhibits autophagy and negatively regulates the autophagy-promoting Beclin-1-VPS34 complex by binding to the BH3 domain of Beclin-1.[4142] To date, many small molecule BH3 mimetics have been designed to inhibit the anti-apoptotic Bcl-2 proteins and induce apoptosis. However, most of them failed to exhibit antitumor effects in the preclinical and clinical trials,[43] suggesting the induction of autophagic cell death might be better suited at present to the strategies focusing on the inhibition of anti-apoptotic Bcl-2 proteins for overcoming 5-FU resistance.[44]

Recently, the p38MAPK signaling pathways were found to play a critical role in controlling the balance between apoptosis and autophagy in response to 5-FU. The genotoxic stress-induced by 5-FU is mediated by ataxia telangiectasia mutated, and ataxia telangiectasia and Rad3 related proteins, which also promote the activation of the signaling axis, MAPK kinase 6/3-p38MAPK-p53 driven apoptosis.[35] Another mechanism that may participate in the 5-FU-induced autophagy response is p53-AMPK-mTOR pathway.[1145] 5-FU chemotherapy causes genotoxic stress and then increases p53 expression in colon cancer cells; p53 positively regulates autophagy by activation of AMPK, and subsequent inhibition of mTOR, a process that requires TSC1/2.[4647] Pharmacologic interference with these interactions might provide a novel therapeutic strategy targeting colorectal cancer cells with high 5-FU treatment resistance. In fact, the combination of oxaliplatin/bevacizumab with HCQ is currently being investigated in clinic trials [Table 2].

Hepatocellular carcinoma

Over the past decades, the surgical operation has been the most effective therapeutic strategy for HCC patients at early stages,[48] but most patients reach an advanced stage for the first diagnosis of HCC and lose the opportunity of surgical resection. In those patients with advanced HCC, chemotherapy is mostly ineffective with a low response rate.[49] It has been revealed that suppression of autophagy enhances oxaliplatin-induced cell death[50] while combining it with bevacizumab markedly inhibits the growth of HCC.[51] Moreover, the combination of CQ with sorafenib (a potent multikinase inhibitor that has been recognized as the standard systemic treatment for patients with advanced HCC-based on the results of Study of Heart and Renal Protection trial)[52] can generate more ER stress-induced cell death in HCC both in vivo and in vitro.[53]

Several genes and signal pathways contribute to autophagy-mediated chemoresistance in HCC. Recent research revealed that p53 contributes to cell survival and chemoresistance in HCC under nutrient-deprived conditions by modulating autophagy activation.[28] Blocking p53 leads to impaired activation of autophagy, increased nutrient starvation, and 5-FU-induced cell death in nutrient-deprived HCC accompanied by a remarkable increase in the reactive oxygen species (ROS) generation and mitochondrial damage.

Activation of Mek/Erk signaling could activate autophagy in tumor cells.[54] Recently, linifanib has been reported to inhibit PDGFR-β and its downstream Akt/mTOR and Mek/Erk signal pathways and activate autophagy in HCC cells, which contributes to their survival both in vitro and in vivo.[55] Several other mechanisms triggering autophagy have also been investigated. For instance, Zhou et al. reported that autophagy inhibits chemotherapy-induced apoptosis through downregulation of Bad and Bim in HCC cells.[56] JNK-Bcl-2/Bcl-xL-Bax/Bak pathway and SMAD2 signaling have also been determined as contributors to autophagy of HCC.[5758]

Pancreatic adenocarcinoma

Autophagy plays a cytoprotective role in response to chemotherapy in pancreatic cancer cells lines, PANC-1, and BxPC-3.[27] In a recent study, genistein potentiates the antitumor effect of 5-FU by inducing apoptosis and autophagy in MIA PaCa-2 human pancreatic cancer cells and their derived xenografts.[59] Furthermore, in phase I/II clinical trial, preoperative inhibition of autophagy with HCQ and gemcitabine in patients with pancreatic adenocarcinoma is safe, well-tolerated, and effective.[60] However, a contradictive study showed that HCQ monotherapy achieved inconsistent autophagy inhibition and demonstrated negligible therapeutic efficacy,[61] which might be because the use of HCQ with concurrent chemotherapy may obviate the need for complete autophagy inhibition in tumors, but the exact mechanisms explaining the inconsistency in those clinic trials are yet to be determined.

Inhibition of autophagy with CQ promotes apoptotic cell death in response to inhibition of the PI3K-mTOR pathway in pancreatic adenocarcinoma both in vitro and in vivo.[62] Activation of PI3K results in sequential AKT and mTOR activation, ultimately suppressing autophagy.[6364] Inhibition of autophagy results in enhanced apoptosis following treatment with PI3K inhibitors, in particular, dual-targeted PI3K/mTOR inhibitors. In this sense, Type I PI3K inhibitors (lithium and carbamazepine), type III PI3K inhibitors (3-MA, LY294002 and wortmannin), AKT inhibitors (perifosine and API-2), and mTORinhibitors (rapamycin, RAD001 and CCI-779) currently undergoing clinical evaluation are all promising anticancer agents to improve treatment outcomes in pancreatic adenocarcinoma.[656667]

Esophageal cancer

Malignant cell clones resistance to chemotherapy is a major cause of treatment failure in esophageal squamous carcinoma cells. Several studies have revealed that induction of autophagy plays a significant role in the resistance and recovery of chemotherapeutic drug-treated esophageal cancer cells.[3068697071] In most studies, the inhibition of autophagy leads to increased esophageal cancer cell apoptosis, indicating that autophagy might be a prosurvival mechanism rather than a cell death mechanism. Efforts have been made to investigate the exact self-protective mechanism of autophagy, and it was found to be associated with PI3K/Akt/mTOR[7273] and Stat3/Bcl-2 pathway.[74] Recently, a typical protein kinase CI (PKCI) has been reported to regulate β-catenin in an autophagy-dependent manner in esophageal squamous cell carcinoma cells.[75] Moreover, PKCI may regulate autophagy via intracellular ROS, a known autophagy inducer that promotes autophagy by inactivating the mTOR pathway[76] or inhibiting ATG4,[77] indicating that PKCI could be used as an autophagy inducer in killing esophageal cancer cells.

Gallbladder carcinoma

So far, there are no adjuvant chemotherapeutic combinations widely accepted for the primary GBC due to their toxicity, drug resistance, and limited efficacy resulting in a low survival rate, and almost half of patients already have metastatic disease at the time of surgery.[7879] Currently, 5-FU has been used in phase II trial of combination chemotherapy for advanced cancers of the gallbladder; the toxicity was tolerable but substantial.[80]

We recently observed that combination treatment of CQ and 5-FU was more efficient in killing GBC cells, and pretreatment with CQ increased the 5-FU-induced apoptosis and the G0/G1 arrest in vitro.[81] It is possible that cell cycle influences autophagic degradation, and inhibition of autophagy may cause cells to be arrested to the G0/G1-phase.[26] Given that both apoptosis and autophagy are crucial mechanisms regulating cell survival and homeostasis, the relationship between them is quite complicated.[82] In some cases, they had no connection[8384] while, in some instances, it was demonstrated that autophagy might promote or even restrain apoptosis.[8586] The exact mechanism for the inhibition of autophagy through an increase in the cytotoxicity of 5-FU in GBC cells needs to be verified.

Gastric cancer

The cytoprotective role of autophagy in response to chemotherapy has been confirmed in the 5-FU treatment of gastric cancer cells.[3887] In agreement with this, Zhu et al. showed that PI3K inhibitor promotes the antitumor activity of 5-FU through autophagy.[36] Interestingly, a study that was conducted recently showed that 5-FU may suppress miR-30 to upregulate Beclin-1 and thus induce autophagic cell death and cell proliferation arrest in GC cells[37] indicating that 5-FU may have its inhibitory effect through induced autophagy and specifically autophagic cell death.

AUTOPHAGIC CELL DEATH CONTRIBUTES TO 5-FLUOROURACIL-BASED CHEMOTHERAPY

Autophagy is generally considered to be a survival mechanism. However, when the severity or the duration of the stress is too long, or in apoptotic-deficient cells, autophagy may participate in cell death. Therefore, it is called a nonapoptotic form of programmed cell death (PCD) as autophagic cell death or type II PCD (type I being apoptosis itself).

As mentioned above, autophagy is believed to have both pro- and anti-oncogenic effects on tumor cells.[88] Besides the protective mechanism to mediate the acquired resistance phenotype of certain cancer cells during chemotherapy, autophagy is also considered to play a pro-death role associated with autophagosome, potentially functioning as a tumor suppressor mechanism similar to apoptosis.[689] To date, autophagy inducers are widely used to kill cancer cells;[90919293] it has been reported that some drugs were used for cancer treatment due to their effect on cell autophagy. For example, aloe-emodin-induced rat C6 glioma autophagic death;[94] Resveratrol-induced ovarian cancer cell death through autophagy;[95] 6-shogaol-induced A549 autophagy by suppressing the AKT/mTOR pathway.[96]

In gastrointestinal cancer types, many studies demonstrated that autophagy may mediate cell death in certain cancer cells where apoptosis is defective or difficult to induce. For instance, triptolide, the precursor of tripchlorolide, inhibits the growth of hamster cholangiocarcinoma,[97] and human tumors transplanted into nude mice.[98] It also suppresses the growth of pancreatic cancer[99] and induces cell death through apoptosis and autophagy.[100] Furthermore, at the molecular level, autophagic cell death could be induced in PUMA- or Bax- deficient human colon cancer cells after treatment with 5-FU, resulting in significantly reduced cell proliferation.[101] Thus, inducing autophagy when apoptosis is inhibited or directly triggering autophagic signaling such as PI3K-Akt-mTOR pathway,[102] and inhibitors are possible strategies that can be applied to cancer therapy. These strategies complicate the use of autophagy inhibitor or inducer in cancer chemotherapy and the specific role that autophagy plays at different stages in cancer progression and determination of its cell type and genetic context-dependency needs to be clarified.

CONCLUSIONS AND PERSPECTIVES

Although research on autophagy in chemotherapy has expanded dramatically, it is still controversial whether autophagy activation leads to cell survival or cell death in cancer chemotherapy since autophagy plays a dual role in tumor promotion and tumor suppression. Understanding the novel function of autophagy may allow us to develop a promising therapeutic strategy to enhance the effects of chemotherapy and improve clinical outcomes in the treatment of cancer patients.

Prior to the clinical applications, a mechanistic understanding of the biology of autophagy is urgently needed. There are several questions to be addressed in future studies. First, although 5-FU induces autophagy in many gastrointestinal cancer cells, it is still difficult to explain whether the autophagy accompanies or induces cell death, or only functions as a protective mechanism activated in response to stress-induced by the treatment of 5-FU or is a cell death pathway activated when apoptosis is disabled, or whether all the effects arise in different contexts. In fact, it is very likely that the outcome of autophagy activation is highly dependent on the tumor types.[103104] Second, more new and reliable methods for measuring autophagy in 5-FU treated samples are needed to be developed to maximize the potential of autophagy in the stringent clinical study. Third, among the autophagy inhibitors, only CQ and HCQ are approved by the FDA, but the toxicities and minimal single-agent anticancer efficacy of CQ or HCQ have restricted their clinical application. New and exciting autophagy inhibitors are worthy of further investigation in the future. Overall, our efforts in these areas would increase the understanding of the functional relevance of autophagy within the tumor microenvironment and ongoing dialogue between emerging laboratory and clinical research about targeting autophagy and provide a promising therapeutic strategy to circumvent resistance and enhance the effects of anticancer therapies for cancer patients.

Financial support and sponsorship

This study was supported by a grant of Zhejiang Provincial Natural Science Foundation of China (No. LY13H180001).

Conflicts of interest

There are no conflicts of interest.