Solving the secretory acid sphingomyelinase puzzle: Insights from lysosome-mediated parasite invasion and plasma membrane repair

Summary

Acid Sphingomyelinase (ASM) is a lysosomal enzyme that cleaves the phosphorylcholine head group of sphingomyelin, generating ceramide. Recessive mutations in SMPD1, the gene encoding ASM, cause Niemann-Pick Disease Types A and B. These disorders are attributed not only to lipid accumulation inside lysosomes, but also to changes on the outer leaflet of the plasma membrane, highlighting an extracellular role for ASM. Secretion of ASM occurs under physiological conditions, and earlier studies proposed two forms of the enzyme, one resident in lysosomes and another form that would be diverted to the secretory pathway. Such differential intracellular trafficking has been difficult to explain because there is only one SMPD1 transcript that generates an active enzyme, found primarily inside lysosomes. Unexpectedly, studies of cell invasion by the protozoan parasite Trypanosoma cruzi revealed that conventional lysosomes can fuse with the plasma membrane in response to elevations in intracellular Ca²⁺, releasing their contents extracellularly. ASM exocytosed from lysosomes remodels the outer leaflet of the plasma membrane, promoting parasite invasion and wound repair. Here we discuss the possibility that ASM release during lysosomal exocytosis, in response to various forms of stress, may represent a major source of the secretory form of this enzyme.

‘Lysosomal’ and ‘Secretory’ forms of ASM: A historical perspective

ASM, encoded by the single-copy gene SMPD1, is a mammalian phosphodiesterase that hydrolyzes sphingomyelin to ceramide and phosphocholine. It is a soluble glycoprotein with six potential N-glycosylation sites, of which at least five are utilized (Ferlinz et al., 1997; Qiu et al., 2003). Seven splice variants were identified in various tissues, but only one full-length transcript has been demonstrated to encode a catalytically active enzyme. Inactive splice variants were proposed to have regulatory and/or dominant-negative effects, but their physiological roles remain largely unknown (Rhein et al., 2012). Consistent with its primary localization inside lysosomes (Jones et al., 2008; Lee et al., 2007) ASM has a pH optimum of 4.5–25.0 (Schuchman, 2010). ASM also shows a pattern of maturation that is often seen with lysosomal enzymes: C-terminal processing of a 75 kDa proenzyme generates a 65 kDa active form that co-fractionates with lysosomes (Jenkins et al., 2011). A free C-terminal thiol group was proposed to maintain the ASM proenzyme in a low activity form prior to arrival in the lysosome, in agreement with reports that modification or loss of the C-terminal cysteine enhances enzymatic activity (Jenkins et al., 2011; Qiu et al., 2003).

Defective C-terminal processing and loss of ASM activity is associated with several mutations found in human patients suffering from Niemann-Pick Diseases Types A and B (NPDA/B) (Jenkins et al., 2011). NPDA patients have very low to undetectable levels of ASM activity, and suffer from a rapidly progressing form of neurodegeneration that often leads to death at about three years of age. NPDB is a later-onset disease form, characterized by a heterogeneous set of clinical manifestations that are associated with variable levels of residual ASM activity (McGovern et al., 2013; Schuchman, 2009). After the catalytic domain, the C-terminus is the second most common region in ASM where NPDA/B-associated mutations are found – a finding consistent with the need for proper C-terminal processing within lysosomes for normal ASM activity (Jenkins et al., 2011; Schuchman, 2010).

Cells from NPDA/B patients show abnormal levels of lipid accumulation inside lysosomes, a process thought to play a role in the development of pathology. Interestingly, in the last decade it has become apparent that the clinical findings of NPDA/B may also be due to ASM-mediated reorganization of lipid microdomains on the outer leaflet of the plasma membrane (Schuchman, 2010). Such a role is supported by several studies showing that ASM is secreted extracellularly, and can hydrolyze sphingomyelin-containing substrates at neutral pH (Schissel et al., 1998a; Schissel et al., 1998b; Schissel et al., 1996; Spence et al., 1989). The ASM activity detected in cell supernatants was enhanced in the presence of zinc, while ASM isolated from cell lysates was zinc-insensitive. These were key observations that gave rise to the concept that still prevails: Mammalian cells would contain two ASM forms, one ‘lysosomal’ and refractory to zinc activation, and another ‘secretory’ and zinc-activated.

However, subsequent studies showed that both lysosomal and secretory forms of ASM could be inactivated by a zinc-specific chelator, indicating that ASM within lysosomes was probably already tightly associated with zinc, displaying maximal activity (Schissel et al., 1998b). In agreement with this view, it is known that lysosomes accumulate zinc and can function as ‘zinc sinks’ in cells (Hwang et al., 2008). Furthermore, sequence comparisons with known metalloproteins identified several potential zinc-binding sites in SMPD1 (Schissel et al., 1998b). Given that expression of a single SMPD1 cDNA in mammalian cells generates both lysosomal and secretory ASM forms, and that SMPD1 mutations in NPDA/B patients affect both forms of the enzyme, the observed differences in zinc susceptibility were attributed to differential exposure to this metal during intracellular traffic (Schissel et al., 1998b). Glycosidase sensitivity assays found different glycosylation patterns between ASM purified from cell lysates (considered to be largely the lysosomal form) or from cell supernatants (considered to be largely the secretory form). This heterogeneity in glycosylation gave rise to a model suggesting that ASM containing high mannose-type oligosaccharides would traffic from the trans-Golgi to zinc-rich lysosomes through the classical M-6P receptor-mediated pathway, while ASM displaying complex-type N-linked oligosaccharides would be directly packaged into secretory vesicles, and thus kept under low zinc exposure conditions (Schissel et al., 1998b). In some cell types, however, secreted ASM appears to be resistant to zinc activation (Marathe et al., 1998), a property more consistent with the lysosomal form.

Importantly, several laboratories found that various forms of cellular stress function as a trigger for ASM release from cells. Irradiation, heat shock or exposure to UV light induce the rapid translocation of active ASM from intracellular stores onto the extracellular leaflet of the plasma membrane, generating ceramide-enriched platforms on the cell surface (Gulbins, 2003). Notably, in some instances the ASM secreted in response to stress stimuli had properties of the lysosomal form, such as insensitivity to added zinc (Charruyer et al., 2005) or post-translational modifications associated with a lysosomal localization (Zeidan and Hannun, 2007; Zeidan et al., 2008). Below we discuss how studying host cell invasion by parasites and the mechanism by which mammalian cells repair plasma membrane wounds unexpectedly provided us with useful insights on the origin of ‘secretory’ ASM, and on the role of cellular stress in promoting ASM secretion.

A surprising discovery: Conventional lysosomes behave as Ca²⁺-regulated secretory vesicles

While investigating the mechanism used the intracellular protozoan parasite Trypanosoma cruzi to invade host cells, our group made an unexpected observation. Host cell lysosomes clustered intracellularly at sites of trypomastigote attachment, and gradually fused with the plasma membrane as the parasites entered the cells (Tardieux et al., 1992) (Figure 1). This unusual invasion process was independent of the host cell actin cytoskeleton and resulted in a tight vacuole containing lysosomal membrane markers, where the parasites resided prior to disrupting the vacuole membrane and replicating in the cytosol (Andrews, 1995). An important next step was the demonstration that T. cruzi trypomastigotes trigger intracellular free Ca²⁺ transients in host cells, a process required for lysosome-mediated host cell invasion (Tardieux et al., 1994). Further investigations showed that intracellular free Ca²⁺ concentration increases above 1 μM were sufficient to trigger exocytosis of conventional lysosomes in several cell types (Rodriguez et al., 1997). At that time it was thought that only modified lysosomes (also referred to as ‘lysosome-related organelles’) present in specialized secretory cells such as cytotoxic lymphocytes had the required molecular machinery to undergo Ca²⁺−regulated exocytosis (Griffiths, 1996). Subsequent studies revealed that Ca²⁺−dependent fusion of lysosomes with the plasma membrane is a widespread process (Kukic et al., 2014; Li et al., 2008; Luzio et al., 2007; Miao et al., 2015; Naegeli et al., 2017; Samie et al., 2013; Settembre et al., 2013; Wang and Yamada, 2017) regulated by specific membrane-associated proteins such as SNAREs, the exocyst complex (Naegeli et al., 2017), the ubiquitously expressed synaptotagmin VII Ca²⁺ sensor, and the lysosomal Ca²⁺ channel mucolipin-1 (TRPML1) (Martinez et al., 2000; Rao et al., 2004; Samie et al., 2013). Interestingly, studies using Total Internal Fluorescence (TIRF) microscopy in Chinese Hamster Ovary cells and in mouse embryonic fibroblasts showed that the large majority of lysosomes undergoing exocytosis in response to Ca²⁺ belongs to a pre-existing peripheral population of lysosomes (Jaiswal et al., 2002). Until recently, lysosomes were mostly thought to accumulate at the perinuclear area as a result of dynein-mediated retrograde movement along microtubules. However, recent studies identified molecular complexes that can promote lysosome movement towards the cell periphery, by mediating association with the anterograde molecular motor kinesin (Pu et al., 2016). In the future it will very interesting to determine if peripheral lysosomes have unique characteristics that favor the recruitment of factors that promote kinesin-mediated anterograde transport, and also their docking with the plasma membrane prior to exocytosis.

Figure 1. Release of lysosomal ASM mediates pathogen invasion, endocytosis and plasma membrane repair.

Ca²⁺ influx induced by various external stimuli, including pathogens, agents that permeabilize the plasma membrane or other forms of stress trigger exocytosis of lysosomes and secretion of acid sphingomyelinase (ASM). ASM-mediated hydrolysis of sphingomyelin on the outer leaflet of the plasma membrane generates ceramide-enriched domains, which facilitate membrane remodeling/deformation, pathogen entry and endocytosis/plasma membrane repair.

The availability of these molecular tools and the generation of specific knockout mice allowed an important question to be addressed: What could be the physiological role of Ca²⁺− regulated exocytosis of lysosomes, a process present in most cell types? These studies again resulted in a surprising discovery: Lysosomal exocytosis, triggered by Ca²⁺ entry through wounds in the cellular membrane, initiates a repair process that rapidly restores plasma membrane integrity (Reddy et al., 2001). Earlier studies identified exocytosis as an important step in plasma membrane repair, but the identity of the vesicles that responded to Ca²⁺ influx by fusing with the plasma membrane was unknown (McNeil and Steinhardt, 1997). The detection of lysosomal membrane proteins on the cell surface and of lysosomal hydrolases in the supernatant of wounded cells, combined with biochemical and genetic inhibition approaches, established that Ca²⁺−triggered exocytosis of lysosomes plays a central role in the mechanism of plasma membrane repair (Chakrabarti et al., 2003; Cheng et al., 2014; Reddy et al., 2001).

ASM released from lysosomes remodels the cell surface promoting wound repair and pathogen invasion

Strikingly, a close examination of cells permeabilized with pore-forming toxins revealed that Ca²⁺-triggered exocytosis of lysosomes was followed by a vigorous wave of endocytosis, which promoted rapid internalization and subsequent trafficking of the pores into lysosomes for degradation (Corrotte et al., 2012; Idone et al., 2008; Thiery et al., 2011). The endocytic vesicles formed through this process lacked clathrin coats, and a few minutes after cell injury displayed an overall morphology similar to plasma membrane-derived vesicles previously observed in cells treated with bacterial sphingomyelinase (Zha et al., 1998). This finding led to an important advance in our mechanistic understanding of plasma membrane repair because it linked ASM, a specific lysosome-resident mammalian sphingomyelinase, to the process. Cells isolated from NPDA patients or subjected to RNAi-mediated ASM silencing were impaired in plasma membrane resealing, not only after injury with pore-forming toxins but also after mechanical injury. Importantly, extracellular addition of purified ASM or bacterial sphingomyelinase restored membrane repair, directly implicating ASM in the mechanism by which lesions are removed from the plasma membrane (Tam et al., 2010). As discussed above, ASM hydrolyses sphingomyelin, an abundant lipid on the outer leaflet of the plasma membrane, generating ceramide-enriched domains. In addition to signaling roles (Charruyer et al., 2005) such ceramide-enriched platforms have been shown to promote membrane invagination (Grassme et al., 2007; Gulbins and Kolesnick, 2003; Holopainen et al., 2000; van Blitterswijk et al., 2003), suggesting a mechanism by which injury-induced ASM release from lysosomes can result in vigorous endocytosis and wound repair (Andrews et al., 2014). Impaired endocytosis after mechanical wounding and a loss in contractile force after eccentric contraction-induced injury was also observed in isolated skeletal muscle fibers from a mouse model of NPDA/B (Michailowsky et al., 2019) (Figure 2).

Figure 2. ASM deficiency inhibits the massive endocytosis response to plasma membrane wounding in skeletal muscle fibers.

Flexor digitorum brevis fibers isolated from wild type (WT) or ASM^−/− littermate mice were left intact (no wounding) or mechanically injured by passage through a 30-gauge needle (wounding) and fixed and processed for transmission electron microscopy after 1 minute at 37°C. The arrowheads point to open caveolae on the sarcolemma, and the arrows point to intracellular vesicles that were triggered by wounding in WT but not in ASM^−/− fibers (Michailowsky et al., 2019).

Interestingly, extracellularly-added ASM also promotes T. cruzi entry in ASM-depleted cells, indicating that these parasites subvert the ASM-mediated plasma membrane repair process as a strategy for host cell invasion (Fernandes et al., 2011). This invasion strategy appears to be more widespread than originally thought, having been also demonstrated for fibroblast invasion by the T. cruzi-related protozoan Leishmania (Cavalcante-Costa et al., 2019) and for adenovirus infections (Luisoni et al., 2015). Earlier studies also found a role for ASM in host cell entry by the bacteria Neisseria and Pseudomonas and Sindbis viruses, although the origin of the extracellular ASM generating ceramide-enriched plasma membrane domains and pathogen internalization was unclear (Grassme et al., 1997; Jan et al., 2000; Simonis et al., 2014). It is becoming increasingly clear that many pathogens that produce membranedamaging agents also trigger a repair mechanism involving exocytosis of lysosomal ASM, generating ceramide-enriched cell surface domains that facilitate cell invasion.

Is ‘secretory’ ASM a result of stress-induced lysosomal exocytosis?

As discussed above, extensive evidence indicates that plasma membrane injury inflicted by various processes, including interaction with pathogens, is followed by lysosomal exocytosis and extracellular release of ASM. These unexpected findings offered a new perspective on the possible origin of the ASM form historically referred to as ‘secretory’. Given that ASM secretion has often been associated with events that cause cellular stress and possibly plasma membrane wounding, it is tempting to suggest that exocytosis of conventional lysosomes triggered by Ca²⁺ influx might represent a major source of the extracellularly secreted enzyme (Figure 1), without the need for postulating differential intracellular trafficking.

How can this hypothesis be reconciled with the different properties that have been associated with the ‘secretory’ and ‘lysosomal’ forms of ASM? First, ASM activity found in cellular supernatants is often increased by zinc, while the lysosomal form appears to be refractory to zinc addition, presumably due to previous exposure to high concentrations of this metal inside lysosomes (Schissel et al., 1998b). Second, some studies have reported a larger molecular size for the extracellular, secretory ASM, when compared to the form isolated from cell lysates and assumed to correspond to the mature lysosomal form (Schissel et al., 1998b). Third, differences in posttranslational glycosylation have been reported between the two forms of the enzyme, suggesting an explanation for how a common ASM precursor trafficking through the Golgi complex could give rise to two enzyme populations, one targeted to lysosomes and another one packaged in secretory vesicles (Hurwitz et al., 1994; Schissel et al., 1998b; Takahashi et al., 2005). It is important to note, however, that different extraction/purification procedures have been used for ASM associated with cell lysates or supernatants (Jenkins et al., 2011; Schissel et al., 1998a; Schissel et al., 1998b; Schissel et al., 1996), and it is conceivable that stress/injury-induced lysosomal exocytosis could result in rapid displacement of zinc from lysosomal ASM, restoring it to a zinc-sensitive condition when reaching the extracellular medium. Cellular factors may also influence how tightly-associated zinc might be released from lysosomal ASM extracellularly, a scenario consistent with reports that ASM secreted by endothelial cells is only minimally responsive to zinc (Marathe et al., 1999). However, the larger size reported for secreted ASM and, in particular, the different glycosylation patterns remain consistent with the existence of two enzyme forms and should be revisited experimentally in a larger number of cell types before conventional lysosomes can be further considered as a major source of ‘secretory’ ASM. In this respect, examining the trafficking consequences of targeted genetic mutations on the putative ASM glycosylation sites should be informative. In the meantime, this novel perspective on the cellular origin of secretory ASM certainly provides one more example of how the study of intracellular pathogens can suddenly shed light on obscure, slowly progressing areas of cell biology.