1. Introduction

CQ and HCQ are quinoline derivatives that exhibit a variety of activities and are able to cross cell membranes by passive diffusion. These drugs accumulate and are subsequently protonated in acidic vesicles such as lysosomes. This accumulation leads to a weakened acidic environment, thereby disrupting the endolysosomal system [144]. The fusion of lysosomes and autophagosomes is a critical step for the completion of autophagy, sometimes referred to as autophagic flux. CQ interferes with the fusion process and is therefore considered to be an effective inhibitor of late-stage autophagy [145]. The combination of CQ with cisplatin has been reported to not only increase the chemotherapeutic efficacy of cisplatin-sensitive cancer cells [145] but also to effectively improve the chemosensitivity of cisplatin-resistant cancer cells (Table 1). For example, as listed in Table 1 below, CQ increased cisplatin sensitivity by inhibiting autophagy in cisplatin-resistant A549 (A549/cisplatin) [146], endometrial cancer [147], urothelial carcinoma [148], epithelial ovarian cancer [149], esophageal cancers [29], and neuroblastoma [150]. Apart from the generally recognized autophagy inhibitory effect, recent studies have also reported autophagy-independent activities of CQ against breast cancer [100], as well as increased cisplatin sensitivity to laryngeal tumor cells by promoting repolarizing tumor-associated macrophages from M2 to M1 in vivo and in vitro [151].

The above studies appear to suggest that the combination of CQ and cisplatin is indeed a potential approach to overcoming cisplatin resistance, where this sensitizing effect is largely relevant to specific tumor types. For instance, in patients with cisplatin-resistant oral squamous cell carcinoma, inhibition of autophagy does not seem to be an ideal treatment. A recent study showed that although CQ increased the apoptosis rate of cisplatin-treated SSC-4 cells, it had a very limited effect on the apoptosis of cisplatin-resistant SSC-4 cells (only about a 5–7% increase) [152].

For cells that are inherently sensitive to cisplatin, the outcomes for CQ in combination with cisplatin seem to be more relevant to the tumor types. For example, in cisplatin-sensitive glioblastoma, pediatric medulloblastoma cell lines, atypical teratoma/rhabdomyosarcoma cell lines [153], low ARHI-expressing ovarian cancer SKOV3 cells [77], p53 wild-type lung cancer H460 cells [50] and p53-null mouse breast cancer 67NR and 4T1 cells [100], CQ had no significant effect on the activity of cisplatin (Table 1).

In addition to CQ, Bafilomycin-A1 and PI3K inhibitors (3-MA and wortmannin) are also common inhibitors of autophagy and have been reported to increase the chemosensitivity of cisplatin in different tumor cells in a large number of studies, which have been summarized in many reviews [28]. Nevertheless, there are also exceptions. An example is that the 3-MA showed no effect on the cell proliferation rate in cisplatin-treated nasopharyngeal carcinoma CNE1 cells [82]. More interesting examples are cisplatin-resistant gastric cancer KATO-III cells, for which studies have shown that the resistance is independent of MRP1 and MDR1 and rather linked to Aldoketoreductase1 C1 and C3 (AKR1C1 and AKR1C3). When AKR1C1 and AKR1C3 were inhibited, the combination of 3-MA paradoxically decreased the cell death rate of KATO-III induced by cisplatin [154].

Based on the data presented above using combinations of autophagy inhibitors with cisplatin, although most of the reported cisplatin-resistant cells demonstrated sensitization from the combination treatments, inefficient or even ineffective outcomes were also evident. Moreover, most of the experimental data were generated solely in in vitro cellular models. Furthermore, the paradigm that “autophagy is a drug resistance mechanism” may have occasionally resulted in a less than objective interpretations of the data.

In addition to these more classical autophagy inhibitors, there are some compounds and nanomaterials that have also been found to overcome cisplatin resistance by disrupting autophagy. U0126, a MAPK inhibitor, enhanced cisplatin-induced apoptosis by inhibiting autophagy in cisplatin-resistant ovarian cancer cells [155] and NSCLC cells [156]. Some natural products have also been found to promote the activity of cisplatin by regulating autophagy (Table 2). For example, astragaloside IV (AS-IV) derived from Astragalus membranaceus sensitizes cisplatin-resistant NSCLC cells to cisplatin by inhibiting ER stress and autophagy [157]. However, the role of autophagy induced by natural products in combination with cisplatin is not consistent. For instance, Gardenia jasminoides (GJ), a medicinal herb abundant with flavonoids, in combination with cisplatin paradoxically activated cytotoxic autophagy in glioblastoma multiforme [158]. In addition, the combination of autophagy-inhibiting nanoparticles or materials with cisplatin enhanced the chemosensitivity or reduced the resistance to cisplatin. For instance, a nanoparticle-based co-delivery system (siBec1@PPN) is centered on the efficient co-delivery of Beclin1 siRNA (Beclin1 is an autophagy initiation factor) and cisplatin to enhance the inhibitory effect on cisplatin-resistant A549 cells by inhibiting autophagy in vivo and in vitro [159]. Compared with the poly lactic acid (PLA) + cisplatin nanoparticles (CDDP-PLA NPs), the PLA + cisplatin-CQ nanoparticles (CDDP/CQ-PLA NPs) reduced autophagy and enhanced the ROS and apoptosis of Cal-27 cells [160].

One final issue that we suggest is worthy of additional attention is the “switch” between the roles played by autophagy in different tumors. As mentioned earlier, although in many if not most cases where cisplatin-induced autophagy has been detected the autophagy was functionally cytoprotective, particularly in cisplatin-resistant cells, the non-protective form of autophagy (i.e., where autophagy inhibition failed to influence cisplatin sensitivity) has also been observed, particularly in cisplatin-sensitive cells [50,77,100]. However, interestingly, re-expression of ARHI in SKOV3 cells allowed for CQ to suppress cisplatin sensitivity [77], whereas knockdown of p53 in H460 cells resulted in enhanced cisplatin sensitivity [50], which is what we refer to as the “autophagic switch”. This again argues that studies to improve the efficacy of chemotherapy by altering autophagy function should first distinguish the role of autophagy in specific models. In the absence of such a strategy, experimental conclusions in preclinical models may prove to be flawed, while translation of the work could be compromised.

The above extensive literature strongly suggests that inhibition of autophagy could, in fact, prove to be an effective strategy for combating cisplatin resistance in the clinic. However, this approach does not fully consider, for example, the troublesome issue of cisplatin toxicity to the kidneys. Cisplatin nephrotoxicity is related to the excretion of cisplatin, which occurs through the kidney tubular epithelial cells. Asymptomatic elevation of serum creatinine levels or even acute tubular injury requiring dialysis therapy occurs with cisplatin chemotherapy in the clinic. Patients who present with this condition often need to have their medication doses reduced to avoid further kidney damage, resulting in under-treatment of the disease [167]. Renal injury is often accompanied by other complications such as water and nitrogenous waste retention and is associated with a poorer patient prognosis [168]. Recent studies have shown that proximal tubule-specific autophagy-deficient mice are more susceptible to kidney injury after cisplatin treatment than wild-type mice [169]. This may be related to autophagy protecting the proximal tubular cells from mitochondrial oxidative stress and protecting the proximal tubular cells from DNA damage. Furthermore, autophagy also protects the proximal tubular cells from ischemic injury [170]. Zhang et al. reported that the mechanism of cisplatin nephrotoxicity may be related to the inhibition of autophagy by the activation of protein kinase C δ [171]. Li et al. found that 3-dehydroxyceanothetric acid 2-methyl ester (3DC2ME) isolated from the roots of jujube (Ziziphus jujuba, Rhamnaceae) protected against cisplatin-induced renal epithelial LLC-PK1 cell injury via autophagy modulation [172]. Retinoic acid, a major derivative of vitamin A, attenuates cisplatin-induced acute kidney injury by activating autophagy [173]. Numerous reports have demonstrated a protective effect of autophagy against cisplatin-induced renal cell injury [174]. Thus, prolonged coadministration of high doses of CQ or other autophagy inhibitors may exacerbate the nephrotoxicity of cisplatin. An example of this outcome can be taken from a study of amniotic fluid stem cells (AFSC) by Minocha et al., who found that AFSC reduced cisplatin-induced renal apoptosis in rats and served to protect against acute kidney injury, but CQ counteracted the renal protective effect of the AFSC [175]. The protective effect of autophagy was also demonstrated in cisplatin-induced damage to the cochlear cells [176]. In addition, the current study also demonstrates the importance of autophagy in enhancing the therapeutic potential of stem cell therapy in attenuating cisplatin-associated liver injury [177]. The two-sided nature of autophagy inhibition during cisplatin treatment suggests the necessity of elucidating the pattern of autophagy in therapy and finding ways to target the delivery of autophagy inhibitors to lesions while mitigating nephrotoxicity as well as other normal tissue injury associated with the administration of cisplatin in cancer therapy.