Lysosomes are acidic organelles essential for degradation, signalling and cell homoeostasis. In addition, they play a key role in cell death. Permeabilization of the lysosomal membrane and release of hydrolytic enzymes to the cytosol accompanies apoptosis signalling in several systems. The regulatory mechanism of lysosomal stability is, however, poorly understood. Lipophilic or amphiphilic compounds with a basic moiety will become protonated and trapped within lysosomes, and such lysosomotropic behaviour is also found in many pharmacological drugs. The natural sphingolipid sphingosine exhibits lysosomotropic detergent ability and is an endogenous candidate for controlling lysosomal membrane permeabilization. The lysosomotropic properties of certain detergents might be of use in lysosome-targeting anticancer drugs and drug delivery system in the future. The present review summarizes the current knowledge on the targeting and permeabilizing properties of lysosomotropic detergents from a cellular and physicochemical perspective.
Lysosomes were originally described in the 1950s by Christian de Duve and are the key degradative compartments of the cell. They are advanced organelles involved in many cellular processes and are considered crucial regulators of cell homoeostasis (reviewed in ). The interior is acidic with a pH of <5 and contains over 50 hydrolases, able to degrade all constituents of the cell. Lysosomes are limited by a single 7–10-nm phospholipid bilayer that functions as an interface to regulate communication between the lysosomal lumen and the cytosol and is therefore transversed by several membrane proteins . In addition to the limiting lysosomal membrane, lysosomes and late endosomes contain intralysosomal membranes, which represent the main site of degradation within this organelle . Lysosomal membrane proteins are generally heavily glycosylated at their luminal domain and form a glycocalyx, which is suggested to protect the membrane from the action of the hydrolytic interior . The lysosomal membrane also facilitates interaction and fusion with other cellular compartments, including endosomes, autophagosomes and the plasma membrane .
Lysosomal membrane permeabilization in cell death
Owing to their high hydrolase content, leakage of lysosomal content to the cytosol is potentially harmful to the cell. Partial permeabilization of the membrane induces apoptosis, whereas massive lysosomal rupture induces necrosis [6,7]. Lysosomal membrane permeabilization (LMP) and release of the lysosomal proteases, the cathepsins, to the cytosol have been shown to be the critical step in lysosome-mediated apoptosis [8–12]. Although the proteolytic activity of the cathepsins is dependent on an acidic pH, their mechanism of action in the cytosol has also been found to be by proteolytic processing of cytosolic substrates. This might be explained by stabilization of the cathepsins by substrate binding or cytosolic acidification by protons released from the lysosome [13–15]. Bid processing into its pro-apoptotic truncated form (tBid) is the most studied cathepsin substrate [11,16,17]. Other substrates have also been identified and have been reviewed elsewhere . After LMP, downstream signalling often involves activation of the caspase cascade, via the intrinsic [18,19] and extrinsic  pathways. The mechanism of LMP by which the lysosomal content escapes the lysosomes remains elusive, although several compounds have been identified as both stimulating and preventing the loss of lysosomal integrity (reviewed in ). The possibility of inducing cell death through LMP appears interesting in tumour treatment because the apoptosis mechanism is frequently impaired in cancer cells, whereas lysosomes could be considered an Achilles heel.
The acidic interior of lysosomes makes them susceptible to the accumulation of weak bases able to diffuse across the lysosomal membrane. As they reach this compartment, these substances become protonated and, as charged molecules, their diffusion back into the cytosol is severely hindered . Such substances are referred to as lysosomotropic agents, and, via this mechanism, trapped substances may accumulate to concentrations hundredfold that of the cytosolic concentration  (Figure 1). Several well-known drugs used in the clinic have documented lysosomotropic abilities including the antimalarial drug chloroquine, several antipsychotic drugs (chlorpromazine, thioridazine and aripiprazole) and antidepressants (desipramine, imipramine and clomipramine). A recent screening of lysosomotropism  found that drugs with a ClogP > 2 (partition coefficient of the neutral species of a compound between octanol and water), and a pKa between 6.5 and 11, caused lysosomal accumulation. The lysosomotropism in itself could probably contribute to cytotoxicity , and several of these compounds have been shown to cause LMP. The mechanistic explanation for this differs depending on their chemical structures. Lysosomal lysis due to accumulation of the non-permeable charged substances such as free amino acids after metabolism of L-leucyl-L-leucine methyl ester (LeuLeuOMe) or Gly-Phe-β-naphthylamide could build up an osmotic pressure across the lysosomal membrane, which results in the inflow of water [12,25]. Other toxic effects of lysosomotropic drugs are attributed to inhibition of acid sphingomyelinase (aSMase), a lysosomal enzyme that catalyses the degradation of sphingomyelin to ceramide . In most cases, however, a mechanistic explanation for LMP is lacking.
A lysosomotropic agent accumulates in lysosomes due to ion trapping
An interesting group of substances are the lysosomotropic detergents (LDs), which, according to their amphiphilic nature, can partition in the phospholipid bilayer and translocate across membranes as uncharged molecules. Lysosomotropic amines contain a moderately basic amino group with a pKa of 5–9 and are able to passively diffuse across cell membranes. If the compound also contains a long hydrophobic chain, it will act by burying its hydrophobic tail in the membrane with the hydrophilic protonated head group facing the aqueous interior of the lysosomes . The amine will accumulate inside the lysosomes until the concentration is sufficient to solubilize the lysosomal membrane . In contrast with other detergents that kill cells by acting at the plasma membrane, LDs primarily act from within the lysosomes. The lines of evidence leading to this conclusion include the absence of cytotoxicity of LDs towards cells lacking lysosomes (e.g. erythrocytes), the dependence of cytotoxicity on the capacity of LDs of being protonated and on their hydrocarbon chain length, the intracellular vacuolization preceding cell destruction and the protection against cytotoxicity exerted by raising lysosomal pH [27,29,30].
Different LDs have been extensively studied using cell models and can be used to study the cellular effect of targeted lysosomal permeabilization. O-methyl-serine dodecylamine hydrochloride (MSDH), an amine with a 12-carbon hydrophobic chain and a pKa of 5.9 , induces apoptotic cell death in several cellular systems including macrophage-like cells and human fibroblasts [6,17]. The apoptosis-inducing mechanism relies on the activity of lysosomal proteases released to the cytosol followed by release of cytochrome c from mitochondria and initiation of the caspase cascade. Noteworthy, MSDH-induced lysosomal leakage is detected without complete disintegration or lysis of the organelle. Immunofluorescence staining of lysosomal associated membrane protein-2 (LAMP-2) shows that the lysosomal spherical shape is clearly visible even after considerable amounts of proteinases are found in the cytosol . The LD N-dodecylimidazole also shows detergent properties, and the cell-killing mechanism is dependent on acidic conditions and functioning lysosomes . The dependence of lysosomal enzymes for cell killing was also elegantly shown by the use of fibroblasts from patients with mucolipidosis II (I-cell disease), which lack the ability to add mannose 6-phosphate to pro-cathepsins and target them to lysosomes. Thus N-dodecylimidazole toxicity was reduced in I-cell fibroblasts as compared with normal fibroblasts . The amphiphilic amine siramesine, a σ-2 receptor ligand and potent anticancer agent, has been found to cause rise in the lysosomal pH and cause cell death in tumour cells in a detergent-like manner. After treatment, the lysosomes were significantly more prone to disintegrate under isotonic conditions in vitro than the lysosomes from untreated cells. This effect was emphasized further when the isolated lysosomes were exposed to hypotonic stress . A recent study was, however, unable to verify the lysosomotropic effect of siramesine in HaCaT cells .
Sphingosine: the endogenous LD?
Sphingosine is a bioactive lipid of cellular membranes that is involved in numerous physiological functions. It belongs to the sphingolipids and is implicated in LMP and apoptosis [7,34,35]. Sphingosine is proposed to accumulate inside lysosomes and to permeabilize the membrane in a detergent-like fashion. Sphingosine is generated in the lysosome by the sequential action of aSMase and acid ceramidase, which convert sphingomyelin into ceramide and, subsequently, sphingosine (Figure 2). In contrast with sphingosine, sphingomyelin has been shown to protect against lysosomal destabilization . The addition of sphingosine to isolated lysosomes induces membrane destabilization, which is dependent on the presence of cathepsin B . In apoptotic cells, the levels of sphingosine may increase due to the enhanced activity of aSMase and cathepsin B-mediated degradation of sphingosine kinase-1 , which normally converts pro-apoptotic sphingosine into anti-apoptotic sphingosine 1-phosphate. This effect may explain, at least in part, the dependence of sphingosine on cathepsin B for efficient LMP induction. On the other hand, it was demonstrated that apoptosis could be induced by sphingosine acting as a typical LD , i.e. by being accumulated in the lysosomal acidic compartment and selectively permeabilizing the lysosomal membrane. Sphingosine was shown to produce early lysosomal leakage as assessed by the relocation of fluorescent dyes and lysosomal enzymes, and this lysosomal membrane disruption was shown to occur before any sign of apoptosis could be detected. The authors also showed that raising lysosomal pH protected the cells from sphingosine-induced apoptosis, which points to the importance of pH-dependent accumulation of sphingosine inside lysosomes.
Conversion of sphingomyelin into sphingosine
Addition of exogenous sphingosine was shown to induce cell death [37,38], with concomitant alteration of components of the apoptotic signalling pathway. However, the interconversion of sphingosine into ceramide and the difficulties in inhibiting this metabolic step [39,40] complicate a final conclusion about the effect of sphingosine on classic apoptotic cascades independent of that well established for ceramide .
Model membrane permeabilization by sphingosine
As mentioned above, numerous cell studies have been carried out using several different LDs. In contrast, the physicochemical mechanism of permeabilization using model membranes has only been examined for the natural LD sphingosine. Its chemical structure consists of a long hydrocarbon chain, usually C18 and C20, which contains an amine group that can become protonated owing to its pKa being ~6.6 [42,43]. Its apparent critical aggregation concentration, i.e. the concentration above which sphingosine monomers start forming aggregates, depends upon pH , and it was found to be ~2.4-fold higher for the protonated form. The structure of sphingosine aggregates is also dependent on pH. Increasing uncharged sphingosine concentration leads to an unrestrained growth in aggregates size, whereas increasing protonated sphingosine concentration leads to an increase in aggregate concentration. These results were explained in terms of a shift in the hydrogen-bonding network from intra- to inter-molecular when sphingosine becomes protonated. The change in the aggregate structure can also be understood in terms of packing parameters, where the increased headgroup repulsion between charged sphingosines changes the effective headgroup area in the aggregate, inducing a greater curvature on the aggregate surface. Sphingosine partitions into and stabilizes cholesterol–sphingomyelin-ordered domains , and it also rigidifies membranes composed of phosphatidylcholine (PC), phosphatidylserine, phosphatidic acid or their mixtures . The ordering effect of sphingosine on fluid PC membranes or on PC–sphingomyelin–cholesterol raft-mimicking membranes was shown to depend on sphingosine concentration and on pH. Gel domains enriched in sphingosine were more easily formed at neutral pH, where most of the sphingosine molecules are uncharged . Model membranes were shown to be permeabilized by sphingosine , and the extent of permeabilization was shown to depend on the lipid composition. Interestingly, the permeabilization exerted by sphingosine was not a consequence of a detergent-like effect, i.e. membranes were not solubilized by sphingosine, or by the transient formation of non-lamellar inverted structures. Rather, the increased permeability induced by sphingosine was a consequence of its selective partition and resulting rigidification of less fluid lipid domains, which creates defects at the interfaces between these domains and more fluid ones that coexist. Sphingosine was also shown to form channels in biological membranes of different lipidic composition such as mitochondria and plasma membrane . These channels were characterized as having short open lifetimes and diameters not large enough (~2 nm) to allow for the leakage of apoptotic proteins from mitochondria. The above-mentioned physicochemical studies point to the importance of the effect of pH on the permeabilization properties of LDs.
The ability to target the lysosomal compartment and specifically to permeabilize the lysosomal membrane was identified early as being potentially useful to trigger cell death. In this sense, LDs appear to be particularly interesting molecules because they not only share this ability with lysosomotropic agents in general, but also, being amphiphilic molecules, spontaneously organize in aggregates that could serve as carrier systems. In order to exploit the targeting and permeabilizing properties of LDs, a detailed understanding of their physicochemical behaviour is required. Thus it becomes necessary to carry out studies using model membranes. These studies will in turn provide complementary information on the physiological regulation of LMP.
Membrane Morphology and Function: A Biochemical Society Focused Meeting held at Hotel del Camerlengo, Fara San Martino, Abruzzo, Italy, 5–8 May 2014. Organized and Edited by Banafshé Larijani [IKERBASQUE, Basque Foundation for Science and Unidad de Biofísica (CSIC-UPV/EHU), University of the Basque Country, Spain] and Marco Falasca (Barts and The London School of Medicine and Dentistry, U.K.)
Our work is supported by the Swedish Research Council [grant number 3214] and the Konung Gustaf V:s och Drottning Victorias Frimurarestiftelse.