Lysosomes are acidic organelles that contain hydrolytic enzymes that mediate the intracellular degradation of macromolecules. Damage of these organelles often results in lysosomal membrane permeabilization (LMP) and the release into the cytoplasm of the soluble lysosomal contents, which include proteolytic enzymes of the cathepsin family. This, in turn, activates several intracellular cascades that promote a type of regulated cell death, called lysosome-dependent cell death (LDCD). LDCD can be inhibited by pharmacological or genetic blockade of cathepsin activity, or by protecting the lysosomal membrane, thereby stabilizing the organelle. Lysosomal alterations are common in cancer cells and may increase the sensitivity of these cells to agents that promote LMP. In this review, we summarize recent findings supporting the use of LDCD as a means of killing cancer cells.
The lysosome is an acidic organelle that acts as the ‘stomach’ of the cell . Lysosomes contain over 60 hydrolytic enzymes that mediate the degradation of extracellular components that are delivered via endocytosis and phagocytosis or intracellular material that is delivered to the lysosome via autophagy . After degradation, the resulting products are recycled back to the cytosol via transport channels or released into the extracellular space via exocytosis. While initially considered solely as a cellular garbage disposal system, recent findings have uncovered additional roles of lysosomes as essential cellular signaling platforms .
Lysosomal enzymes digest cellular compounds in an acidic environment (the pH of the lysosomal lumen pH is 4.5–5), but some, such as cathepsin B, D, and L, are also active at higher pH, although their stability and activity vary among them .
Proper localization of lysosomal enzymes within the lysosome is critical to maintain a correct cellular homeostasis, as their leakage into the cytoplasm can cause cell death by degrading essential cellular components and/or activating apoptotic pathways. This leakage is the consequence of lysosomal membrane permeabilization (LMP), a multi-pathway process that results in lysosome-dependent cell death (LDCD), which, in most cases, can be prevented by pharmacological or genetic blockade of protease activity [5,6].
LMP and LDCD
Cell death as a result of lysosomal rupture was first described by the discoverer of lysosomes, Christian de Duve, who defined them as ‘suicide bags’, owing to their content of powerful hydrolytic enzymes . Although the concept was defined almost 40 years ago, this regulated cell death pathway remained poorly understood due to limitations in the methodologies used to study it. However, the recent development of new techniques has rekindled interest in this field [8,9].
LMP is a perturbation of the lysosomal membrane that allows translocation of the lysosomal contents, including lysosomal enzymes, to the cytoplasm of the cell [5,10]. Massive lysosomal rupture induces the release of the entire contents of the lysosome, triggering a cascade of hydrolysis of the cytoplasmic contents, and resulting in generalized cytoplasmic acidification with lethal consequences for the cell . In contrast, partial and selective LMP triggers a cascade of regulated cell death mediated by the release of specific lysosomal enzymes into the cytosol [11–14] (Figure 1). Among them, Cathepsin B and D are the main active proteases after LMP, but other lysosomal proteases such as chymotrypsin B and proteinase 3 have also been implicated in LDCD [15,16].
Inducers of LMP.
Pore formation in the lysosomal membrane is the simplest mean of triggering cathepsin release into the cytoplasm. Several venoms and bacterial toxins have this effect and induce apoptotic cell death [17,18]. Recent findings indicate that in addition to bacterial and viral proteins, endogenous toxin-like molecules, such as aerolysin-like proteins, are also capable of inducing LMP . These molecules can translocate to lysosomes, forming high molecular-mass detergent-stable oligomers that promote lysosomal destabilization and cathepsin B release.
Lysosomotropic detergents are compounds that can cross membranes and remain trapped within the lysosome after protonation, from where they induce LMP . Examples include l-leucyl-l-leucine methyl ester [21,22], the cationic amphiphilic drug (CAD) siramesine, which it has been described to detach acid sphingomyelinase (ASM) from the lysosomal membrane, promoting lysosomal destabilization [23,24] and O-methyl-serine dodecylamide hydrochloride (MSDH) [25,26], all of which induce LDCD.
LMP can also be induced in several other circumstances. For example, an increase in free-radical levels can provoke the lysosomal membrane destabilization. Lysosomes contain high concentrations of iron as a consequence of the enzymatic digestion of iron-containing protein complexes. Hydrogen peroxide, which readily diffuses through membranes, can react with intralysosomal iron in the Fenton reaction, resulting in the formation of highly reactive hydroxyl radicals that cause membrane damage, inducing LMP. In contrast, deferoxamine (DFO)-induced iron chelation can protect lysosomal membranes and reduce cell death in many conditions [27–30].
The lipid composition of lysosome membranes is a key factor regulating lysosome membrane stability. Heat shock protein 70 (Hsp70) promotes the binding between the lipid bis(monoacylglycero)phosphate (BMP) and ASM. This interaction increases ASM activity and lysosomal stability . Interestingly, a rise in cholesterol levels leads to the sequestration of BMP [31,32]. Interestingly, ASM mediates the degradation of sphingomyelinase to ceramide and ASM activity, or decreases in sphingomyelin levels reduce lysosomal membrane stability .
The activation of calpains, a group of cytosolic calcium-activated cysteine proteases, has often been associated with LMP triggering cell death under various conditions, for example in neuronal tissue after ischemia [34,35], in mouse models of retinal dystrophies  and also in physiological conditions as during mammary gland involution . The molecular mechanisms have been associated with the cleavage of several lysosomal proteins such as Hsp70, LAMP2 that result in lysosomal membrane destabilization [34,38]. Together, these observations suggest the existence of a calcium–calpain cascade responsible for LMP and cell death .
LMP is not an in vitro artifact induced solely by altering lysosomal homeostasis in laboratory conditions . Cell death following LMP has been observed in several physiological settings, including during mammary gland involution [37,40], during cell death associated with neutrophil regulation , and in bacteria-infected macrophages to prevent the spread of infection . LMP has also been implicated in certain pathologies. For example, cathepsin B is released from lysosomes in the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) mouse model of Parkinson's disease [42,43], and calcium accumulation into photoreceptors promotes LMP and lysosomal cell death in the rd10 mouse model of retinitis pigmentosa . Lysosome destabilization was also observed in a mouse model of Niemann Pick type A disease, in which neuronal cells exhibit LMP , although other report using fibroblasts from NPA patients did not show signs of LMP . Whether LMP contributes to NPA pathology is still unknown. Given the implication of LMP in several diseases, researchers have sought to identify ways to stabilize the lysosomal membrane. For example, Hsp70 stabilizes in vivo the lysosomal membrane in Niemann Pick disease types A and B . Promoting membrane stability may therefore represent a new therapeutic strategy for certain neurodegenerative conditions.
LMP can induce cell death through several pathways (Figure 2), including caspase-dependent cell death or apoptosis, as in some cases, cathepsins can activate caspases directly or in a caspase-independent manner. [5,14,46,47]. Furthermore, LMP can activate the inflammasome, inducing cytokine release and activating macrophages via a process known as pyroptosis . A key characteristic of LMP-induced cell death is that in most cases can be blocked by cathepsin inhibition by using, for example, CA-074Me and E64-d to inhibit cathepsin B and pepstatin A to inhibit cathepsin D-dependent cell death, respectively [14,40].
LDCD and apoptosis are triggered after LMP.
Over the years, research has highlighted the difficulty in classifying specific cell-death subroutines based on the features of dying cells, as distinct initiation cascades can result in similar cell death morphotypes. In the case of LDCD, this issue is further complicated by the fact that lysosomal permeabilization does not produce concrete morphological alterations that can be identified by electron microscopy . Some of the most powerful tools used to study LMP have only recently been developed and will probably facilitate the discovery in the near future of new conditions associated with LDCD . These tools include the galectin-3 assay, which enables the identification of permeabilized lysosomes using specific antibodies, such as anti-galectin-3. Galectins are glycoprotein-binding proteins that gain access to the luminal side of the heavily glycosylated lysosomal membrane proteins after lysosomal permeabilization .
Lysosomal alterations in cancer
Rapidly dividing cancer cells are highly dependent on proper lysosomal function. Accordingly, transformation and cancer progression involve dramatic changes in the volume, composition, and cellular distribution of lysosomes . Many of these changes provide transformed cells with selective advantages, promoting invasive growth, angiogenesis, and drug resistance. The same changes can, however, markedly sensitize cells to LMP and therefore to the effects of anticancer drugs that target the lysosome . Among the changes observed in cancer cells is an increase in the levels of lysosomal enzymes. For example, the expression of heparanase, an endo-β-d-glucuronidase that participates in extracellular matrix remodeling, is increased in several cancers, including gastric carcinoma and lung and breast cancer, in which it facilitates remodeling of the extracellular matrix to promote metastasis [52,53].
Cathepsins are also frequently overexpressed in cancer cells. Increased expression of cathepsin B has been proposed as a marker of colorectal cancer in some populations . The expression of cathepsin D is also increased in several tumor types, including melanoma, glioma, and lung cancer, and contributes to tumor progression in each of these scenarios [55,56].
In cancer cells, lysosomes exhibit alterations in morphology and intracellular distribution, and are more vulnerable to LMP owing to their larger size [51,57,58]. Accordingly, tumor cells may be more sensitive that non-transformed cells to the toxic effects of lysosomotropic agents , which are being investigated as potential anticancer agents. The pharmacological or genetic inhibition of endogenous stabilizers of the lysosomal membrane (e.g. Hsp70) is another potential means of killing cancer cells. Supporting this view, the response to LMP-based therapies is correlated with Hsp70 levels in several cancer cells lines . Interestingly, expression of SMPD1, the gene encoding ASM, is decreased in gastrointestinal, hepatocellular, salivary gland, renal, and head and neck carcinomas, further sensitizing lysosomes to the effects of destabilizing agents .
Strategies to induce lysosomal cell death in cancer cells
As discussed above, lysosomes may constitute a powerful weapon with which to kill cancer cells from the inside. Selective induction of LMP in cancer cells could prove an efficient means of inducing cancer cell death. A better understanding of the role of lysosomal proteins and lipid composition in cancer cells could facilitate the development of more powerful drugs to kill tumors through the induction of LMP and LDCD. Below, we describe several strategies to kill cancer cells that are currently being studied.
Cationic amphiphilic drugs
Several research groups have focused their efforts on developing CADs to induce LMP and cathepsin release from cancer cells. These molecules exhibit more effects in cancer cells due to lipidomic alterations during the tumorigenic process [23,51]. CADs such as antidepressants (siramesine), anti-malarials, and antihistamines can enter lysosomes and become trapped after their protonation, leading to the accumulation in this compartment. CADs are capable of altering the lipid profile by inducing the detachment of ASM from the lysosomal membrane . The increase in ASM levels within the lysosome results in cathepsin release and lysosomal cell death. The efficacy of these drugs has been already demonstrated in non-small cell lung cancer, in which antihistamine CADs such as loratadine and astemizole reduce patient mortality . Terfenadine, another antihistamine CAD, induces cell death in prostate cancer cells , while astemizole reduces cell proliferation in breast tumor-derived cells and in leukemia cancer cells both in vitro and in vivo . It is worth mentioning that CADs might be beneficial in the treatment of multi-drug-resistant cancers [62,64,65]. Taken together, these results suggest that CADs may hold significant promise as anticancer therapies.
Microtubules facilitate the movement of the lysosome through the cell and constitute another target through which lysosomal stability and function can be altered. Disturbances in microtubule dynamics result in lysosomal instability. For example, vincristine, a microtubule-destabilizing antimitotic drug, induces lysosomal destabilization in both HeLa and MCF7 cell lines . Vincristine treatment induces cell death accompanied by cell cycle arrest and cathepsin release. Interestingly, activation of Bax and caspase and cytochrome c release, three hallmarks of apoptosis, were found in dying cells . Another microtubule-destabilizer, paclitaxel, inhibits lysosomal trafficking and induces the lysosome expansion, leading to cathepsin release and LDCD in several cancer cell lines . A study to identify microtubule proteins implicated in LDCD demonstrated that the depletion or pharmacological inhibition of the molecular motor KIF11/Eg5 induced LMP . These proteins have now been proposed as potential drug targets, but further research needs to be done to understand the process of LMP after microtubule destabilization and whether they could be applied for cancer therapy.
Autophagy induction and cathepsin maturation
The findings of a recent study describe inhibition of cathepsin maturation as a novel means of treating cancer cells. Liu et al.  studied the mechanism of cell death induced by IMB-6G, an anti-inflammatory natural derivative of sophoridine, in two pancreatic cancer cell lines. Their data show that IMB-6G induced apoptosis via LMP. Interestingly, silencing of Atg5 expression attenuated the lysosomal release of cathepsin B into the cytosol and the apoptosis induced by IMB-6 treatment, suggesting that increased autophagy activity was detrimental for cell survival . However, the molecular link between cathepsin maturation and lysosomal stability remains elusive and further studies are needed to clarify this point.
Endoplasmic reticulum stress-inducing compounds
Several lines of evidence indicate a link between endoplasmatic reticulum (ER) stress and lysosomal destabilization. For example, in human glioma cells, tetrahydrocannabinol (THC), the main active compound of cannabis, promotes autophagy activation, sphingolipid synthesis, and ER stress . The treatment with THC induced de novo ceramide synthesis and ER stress that induces increased phosphorylation of eIF2α. Mutant of eIF2α attenuated this response. Interestingly, eIF2α increased the mRNA expression of ER stress-response genes such as ATF4, CHOP, and TRB3 . Later studies demonstrated that THC treatment induced alterations in sphingolipid composition in the ER, which resulted in lysosomal destabilization, cathepsin release, and LMP [68,69]. Additionally, in their study of the relationship between ER stress and cell death in pancreatic cancer, Dauer et al.  found that inhibition of the transcription factor-specific protein 1, Sp1 (a key component of the unfolded protein response), induced LMP in the MiaPaCa-2 pancreatic cell line. These findings indicate that ER stress promotes lysosomal destabilization, although the underlying molecular mechanisms remain to be unraveled.
Histone deacetylases and sirtuins
The sirtuin family is a group of NAD(+)-dependent deacetylases . In 2014, Lahusen and Deng  studied the effects of SRT1720, a SIRT1 activator, in the MDA-MB-231 breast cancer cell line, and found that SRT1720 treatment induced cell death independently of Sirt1. While SRT1720-induced cell death was preceded by autophagy activation as a protective mechanism, the authors reported that cell death was inhibited with lysosomal acidity inhibitors, suggesting the implication of lysosomes during cell death under these conditions.
Calcium-dependent LMP pathways
Lysosomes act as essential calcium stores and play an important role in calcium signaling [73,74], modulating calcium release through multiple calcium channels in the lysosomal membrane. As previously mentioned above, calpains, which are calcium-activated proteases, are capable of cleaving Hsp70, inducing lysosomal destabilization . Additionally, in their study in a gastric cancer cell line, Lv et al.  found that mushroom-derived hispidin induced cell death via caspase 9 cleavage. This cell death could not be rescued with the general caspase inhibitor z-VAD. Further analysis of the underlying mechanism revealed that hispidin induced LMP. While cathepsins were not implicated in this cell death process, treatment with the calcium chelator BAPTA restored cell viability after hispidin treatment, indicating a role of lysosomal calcium. This result suggests that lysosomal calcium may trigger cell death after the induction of LMP. Whether this effect is associated with calcium-induced damage to other organelles such as mitochondria and ER, or mediated by calpain activation, remains to be elucidated [34,37].
Mitochondrial membrane permeabilization
In many instances, LMP occurs upstream of mitochondrial membrane permeabilization (MMP), and cells deficient in Bax and Bak are completely resistant to LDCD induced by several antibiotics and lysosomotropic agents [47,76]. Interestingly, LMP still occurred in Bax and Bak double deficient cells, demonstrating that LMP is upstream of MMP [47,76]. In the HaCaT skin cancer cell line, siramesine at high concentrations induces MMP and cell death in an LMP-independent manner . Conversely, and in agreement with siramesine acting as a CAD, low doses of siramesine result in a lower rate of cell death that occurs in a lysosome-dependent manner [77,78]. In other circumstances, MMP appears to occur upstream of lysosomal damage. For example, following combined treatment with tumor necrosis factor alpha (TNF-α) and cycloheximide, activation of caspase 8 and caspase 3 is observed 1 h after treatment, while cathepsin B and cathepsin D are detected 6 h after treatment . While these effects could reflect differences in the sensitivity in the detection methods for caspases and cathepsin activity, these findings suggest that LMP is induced by ROS (reactive oxygen species) generated by MMP and this can represent an amplifying loop in the cell death cascade. It is also worth mentioning that TNF-α effects are highly cell-type dependent. In future studies, it will be essential to identify the pathways and processes that lead to MMP or LMP as the first step in the induction of cell death and whether the MMP–LMP axis could be exploited for therapeutic purposes.
The lethal effects of photodynamic therapy on cancer cells, including breast cancer cells [80,81], are well described [80,81]. Photodamage induces both LMP and cell death [82,83]. Interestingly, lysosomes become more resistant to photodamage with increasing intraluminal pH, suggesting that the toxic effects of photodamage are dependent on lysosomal pH .
LDCD can be induced by targeting several processes, including autophagy, cathepsin maturation, endoplasmic stress, MMP, and lysosomal calcium release. While a large variety of stimuli have been implicated in LMP, further research will be necessary to fully understand the mechanisms that underlie this process. Moreover, a greater knowledge of the pathways involved in lysosomal membrane stabilization, including the interaction between Hsp70 and BMP, could allow researchers to target this process to kill cancer cells. Autophagy inhibition, another consequence of lysosomal disturbance, constitutes another novel therapeutic approach for several types of malignances. Indeed, the anticancer effects of the autophagy inhibitor hydroxychloroquine combined with other chemotherapy agents have been studied in several recent clinical trials , for example, in renal carcinoma, pancreatic adenocarcinoma, or melanoma [86,87]. In conclusion, the targeting of lysosomes can be a novel therapeutic approach to kill cancer cells [88–90].
cationic amphiphilic drug
Heat shock protein 70
lysosome-dependent cell death
lysosomal membrane permeabilization
mitochondrial membrane permeabilization
non-small cell lung cancer
reactive oxygen species
tumor necrosis factor alpha
A.S.-P. and P.B. wrote the manuscript.
Research in the P.B.'s laboratory is supported by grants BFU2015-65623 and BFU2015-71869-REDT from Spain's Ministerio de Economia, Industria y Competitividad (MINECO), COST Action Transautophagy (CA15138), and H2020-MSCA-ITN-2017 MSCA-ITN-ETN 765912. A.S.-P. received a FPI grant from MINECO.
The Authors declare that there are no competing interests associated with the manuscript.