TRPML: TRansPorters of Metals in Lysosomes essential for cell survival?

Abstract

Key aspects of lysosomal function are affected by the ionic content of the lysosomal lumen and, therefore, by the ion permeability in the lysosomal membrane. Such functions include regulation of lysosomal acidification, a critical process in delivery and activation of the lysosomal enzymes, release of metals from lysosomes into the cytoplasm and the Ca²⁺-dependent component of membrane fusion events in the endocytic pathway. While the basic mechanisms of lysosomal acidification have been largely defined, the lysosomal metal transport system is not well understood. TRPML1 is a lysosomal ion channel whose malfunction is implicated in the lysosomal storage disease Mucolipidosis Type IV. Recent evidence suggests that TRPML1 is involved in Fe²⁺, Ca²⁺ and Zn²⁺ transport across the lysosomal membrane, ascribing novel physiological roles to this ion channel, and perhaps to its relatives TRPML2 and TRPML3 and illuminating poorly understood aspects of lysosomal function. Further, alterations in metal transport by the TRPMLs due to mutations or environmental factors may contribute to their role in the disease phenotype and cell death.

Introduction

Lysosomes are components of the endocytic pathway responsible for storage and processing of digestive enzymes and for terminal degradation as well as absorption of the endocytosed material [1–3]. Lysosomes also digest the cellular material delivered to autophagosomes during cell renewal and cell death [4–6]. In addition to digestive enzymes, lysosomes contain a system of transporters that play several important roles. These include: i) establishing controlled acidic pH in the lysosomes, ii) absorption of the products of digestion, iii) the release of Ca²⁺ from the lysosomal lumen that drives the fusion of lysosomes with late endosomes, and iv) transport of the metals bound to endocytosed proteins across the lysosomal membrane into the cytoplasm [1–3]. This list of lysosomal ion transport activity is incomplete and it is limited by the available information on the lysosomal transport pathways and by the scope of this review. TRPML channels appear to be prominent lysosomal metal transporters. Several lines of evidence associate the lysosomal ion channel TRPML1 and its relative TRPML3 with these functions and their possible role in cell death. This review aims to summarize this evidence and delineate the remaining questions.

TRPML1 was identified in a search for genetic determinants of the lysosomal storage disease (LSD) Mocolipidosis type IV (MLIV) [7, 8]. TRPML1 is a member of the TRP superfamily of ion channels, and a founding member of the mucolipin subfamily. The main structural features of the mucolipins are in line with other TRP channels with the exception of a large extracytoplasmic loop (luminal for TRPML1 localized in lysosomes and for TRPML2 and TRPML3 when in intracellular organelles) connecting the 1st and 2nd transmembrane domains and the presence of intracellular targeting signals in their C- and N-terminals responsible for localization of these channels in the vesicles comprising the endocytic pathway [9–12]. There are 3 mucolipins in mammals, coded by 3 genes: MCOLN1, MCOLN2 and MCOLN3. The presence of splice variants has been reported in mammals [13, 14], but their function and regulation is unknown at present. Drosophila and C. elegans appear to contain a single gene coding for a mucolipin channel [15–17]. The genetics and the pathogenesis of MLIV as well as expression and regulation of the mucolipins have been the subject of several recent reviews [9, 18, 19] and will not be discussed here further. Instead, we will focus on the recent evidence suggesting that TRPML1 is a lysosomal divalent metal transporter and the possible role of this activity in lysosomal function.

The localization of TRPML1 in the lysosomes and the clear LSD phenotype associated with TRPML1 downregulation suggest a prominent role of this ion channel in lysosomal digestive activity, biogenesis, and/or access to endocytosed material. This general division of lysosomal functions lists the three main topics for which experimental data were previously obtained as being altered in MLIV and thus assumed to be TRPML1 functions. Some insight pertaining to TRPML1 function can also be gleaned from the comparative analysis of localization and activity of TRPML1 and its close relative TRPML3. Whether and how TRPML1 affects lysosomal digestive activity, biogenesis, and/or access to endocytosed material to digestive enzymes is still scarcely delineated. Recent results suggest exciting new possibilities and developments in the mucolipins’ role in the endocytic pathway.

TRPML1 as a lysosomal Ca²⁺ channel in membrane trafficking

The evidence that TRPML1 functions as a lysosomal Ca²⁺ release channel relies on the findings that the TRPML channels are permeable to Ca²⁺ with higher selectivity for Ca²⁺ than for monovalent cations [20] and that TRPML1 is a lysosomal protein [11, 12, 21]. Membrane trafficking deficits [16, 22], as well as impaired fusion of lysosomes with autophagosomes in TRPML1 deficient cells [23] support the role of TRPML1-mediated Ca²⁺ release in the Ca²⁺ dependent membrane fusion along the endocytic pathway. Limitations of the chronic lysosomal storage disease models as experimental systems for studying membrane traffic have been discussed before [24, 25], including the fact that the same delays in membrane traffic were seen in LSDs whose pathogenesis is entirely enzymatic and has little direct relevance to membrane fusion or fission [26, 27]. The post-lysosomal lipid translocation defect observed in the acute TRPML1 knockdown model [24] can be interpreted as a post-lysosomal trafficking defect, or a degradation deficit. Nonetheless, the TRPML1 relative TRPML3 was shown to conduct Ca²⁺ [28–33] and TRPML3 downregulation by siRNA induced trafficking deficits in two model systems [28, 34]. It is important to note that TRPML3 function is inhibited, while TRPML1 function is potentiated [30, 31, 35] by the low pH typical of lysosomes. TRPML3 localizes in the upper (earlier), relative to TRPML1, endocytic pathway compartments, whose acidity is lower than the lysosomal acidity. Taken together, the Ca²⁺ permeability through TRPML1 and TRPML3 and their pH-dependence, matching their localization, support the idea that these channels regulate membrane fusion in specific compartments of the endocytic pathway.

Some of the requirements for TRPML1 to function as a Ca²⁺ release channel that triggers the membrane fusion in the endocytic pathway include: a) an activation signal that is relevant to endosomal function, b) the dependence of lysosomal Ca²⁺ content on TRPML1 status and c) perhaps changes in cellular Ca²⁺ signaling as a function of TRPML1 status since lysosomes have been shown to contribute to the receptor dependent Ca²⁺ signaling in several tissues [36–40]. Ca²⁺ signaling in human [41] and mouse [42] cells lacking TRPML1 appears normal with regard to Ca²⁺ release from internal stores and Ca²⁺ influx across the plasma membrane, with the exception of aberrant mitochondrial function [43]. However, aberrant mitochondrial function in TRPML1 deficient cells is likely secondary to the inhibited lysosomal function as the same Ca²⁺ phenotype have been reported in other models of lysosomal storage diseases [36, 37]. The reports on lysosomal Ca²⁺ content in TRPML1 deficient cells are sparse. Our data on MLIV fibroblasts showed no changes in the Ca²⁺ pool that is released into the cytoplasm by bursting the lysosomes with osmotic shock [43]. It is important to note that lysosomal Ca²⁺ changes have been reported in Niemann-Pick [36, 37], a disease not directly related to ion channel activity. Thus, such changes can be attributed to the effects of storage accumulation, which may bind Ca²⁺ and thus interpretations of such results should be done with caution.

In order to test lysosomal Ca²⁺ homeostasis in a system that is free of artifacts due to clonal variations and effects of storage material, we directly measured lysosomal Ca²⁺ while manipulating TRPML1 function. Fig. 1 shows an assay of lysosomal Ca²⁺ content based on Ca²⁺ released by glycyl-L-phenylalanine 2-naphthylamide (GPN) [44, 45] into the cytoplasm of acutely TRPML1 deficient HeLa cells. The cells were treated with TRPML1 siRNA for 48 hours. In a previous study we show that such timeframe is associated with minimal buildup of storage bodies [24]. The results show a statistically significant increase in a GPN-releasable Ca²⁺ pool in TRPML1 deficient HeLa cells. Recently, it was reported that the endosomal TRPML3 [28] may participate in Ca²⁺ homeostasis of intracellular organelles [34]. Interestingly, Fig. 1 shows that cells deficient in TRPML3 have increased lysosomal Ca²⁺ levels comparable to TRPML1 deficient cells. However, when considered relative to the size of total endoplasmic reticulum Ca²⁺ pool in the given cell type, only TRPML1 deficient cells show an increase in the lysosomal Ca²⁺ pool. This would suggest that depletion of TRPML3 results in expansion of the entire intracellular Ca²⁺ pool, including the ER pool. It is unclear, at present, why the ER Ca²⁺ pool is larger in TRPML3 deficient cells considering that TRPML3 is not constitutively expressed in the ER [28, 34].

Figure 1. Lysosomal Ca²⁺ content is increased in TRPML1 deficient HeLa cells.

HeLa cells were transfected with TRPML1 siRNA as described before [79] and cytoplasmic Ca²⁺ was measured 48 hours later using Fura 2AM. A) Ca²⁺ traces (cytoplasmic Ca²⁺ is proportional to the ratio of Fura 2AM fluorescence ratio measured at 340 and 380 nm excitation light). Lysosomes were burst using exracellular application of glycyl-L-phenylalanine 2-naphthylamide (GPN, 100 µM) [44]. Thapsigargin (Tg, 1 µM) was applied before GPN in order to deplete Ca²⁺ in the endoplasmic reticulum and remove its contribution to the GPN-induced Ca²⁺ release. Data represent 3–10 experiments and are expressed as mean ± S.E.M. * denotes p<0.05. B) Statistical analysis of Ca²⁺ measurements. Top: amplitudes of GPN-induced Ca²⁺ release expressed as a percentage of Ca²⁺ release in control cells. Middle: Tg-induced Ca²⁺ release expressed as a percentage of Ca²⁺ release in control cells. Bottom: the ratios of GPN-induced Ca²⁺ release to Tg-induced Ca²⁺ release in control, TRPML1 and TRPML3 deficient cells.

Recent work clearly identified the two-pore-channels (TPCs) as organellar Ca²⁺ channels, where TPC2 functions as a lysosomal Ca²⁺ channel that is activated by the second messenger NAADP [38, 46–49]. This topic is covered by several reviews in the recent Special Issue of Cell Calcium and the reader is referred to these reviews. It is important to note that the TRP activation by NAADP suggests a clear paradigm of lysosomal Ca²⁺ signaling, where oscillations in the cytoplasmic NAADP content facilitate TPC activation and, therefore, Ca²⁺ release followed by fusion. Up until recently, such a paradigm was not demonstrated for TRPML1 or other TRPML channels. Hence, although manipulation of TRPML1 (Fig. 1) and TRPML3 [34] affect lysosomal storage of Ca²⁺, active Ca²⁺ release from the lysosomes due to activation of any TRPML channel was not demonstrated, as was shown for the TPCs and their activation by NAADP. Nevertheless, if the TRPMLs function as organellar Ca²⁺ channels in the endocytic pathway, they may mediate distinct steps than are different than those regulated by the TPCs. It is also possible that the TRPMLs functions mainly in the endocytic pathway while the TPCs mediate fusion events in the exocytic pathway. It is also possible that the TRPMLs may have additional functions that are not related to Ca²⁺ homeostasis.

The recent finding that TRPML1 is activated by PI(3,5)P2 [50] suggests exciting new possibilities to explain the TRPML1 function in the endocytic pathway. Adapting the whole cell patch clamp technique to patching lysosomes enlarged by vacuolin, revealed that recombinant as well as native TRPML1 channels are activated by PI(3,5)P2 added to the cytoplamic side of the lysosomes. PI(3,5)P2 appears to effect TRPML1 activation through the TRPML1 C-terminal, since removal of the C-terminus abolishes the TRPML1 activation. A crucial point is that PI(3,5)P2 is localized in the distal portion of the endocytic pathway, the site of TRPML1 localization. Activation of TRPML1 by PI(3,5)P2 lends support to the proposed function of TRPML1 in membrane fusion in the distal portion of the endocytic pathway. It poses questions regarding the mechanics of PI(3,5)P2 effect on TRPML1 within the context of membrane interaction within the endocytic pathway. It is tempting to suggest that TRPML1 binding to PI(3,5)P2 in the opposing membrane of the interacting organelle triggers Ca²⁺ release, that actuates the SNARE-mediated fusion machinery. In this sense, the C-terminus of TRPML1 may work as a proximity sensor for the fusing organelles (Fig. 2). It is also possible, and has been recently proposed [51], that TRPML1 is activated by PI(3,5)P2 present within the same membrane. In this case, PI(3,5)P2 may work as a potentiating agent, which assures TRPML1 activation once it reaches the proper (PI(3,5)P2 enriched) membrane compartment.

Figure 2. Models of the role of the activation of TRPML1 by PI(3,5)P2 in the endocytic pathway.

A) TRPML1 as an organellar proximity sensor: TRPML1 activation by PI(3,5)P2 in the opposing membrane promotes organellar fusion. The lack of Ca²⁺ release through TRPML1 in the absence of PI(3,5)P2 binding precludes conformational change in SNARE that promotes fusion (step 1). The proximity of TRPML1-containing and target membranes promotes TRPML1 interaction with PI(3,5)P2 (step 2), Ca²⁺ release through TRPML1 (step 3) and membrane fusion (step 4). B) TRPML1 activation by delivery to PI(3,5)P2-rich compartments. TRPML1 is inactive in the delivery vesicles, but its activation is promoted by its delivery to PI (3,5)P2-rich organelles (steps 5 and 6). This is followed by Ca²⁺ release through TRPML1 (step 7) and membrane fusion (step 8).

The fact that TRPML1 down-regulation does not induce gross inactivation of membrane fusion in the endocytic pathway indicates that the proposed Ca²⁺ release through TRPML1 is responsible for only a subset of the endocytic fusion events. It is, therefore, possible that mucolipins, TPC, and, perhaps, yet unidentified channels cooperate to ensure the membrane flow through the endocytic pathway. Additionally, the recent set of data implicating TRPML channels in transition metal (TM) transport raises the possibility that TRPML channels combine functions of membrane fusion and TM uptake. We will argue here that a potentially central function of the TRPML channels is transport of TM that are byproducts of endocytic capture and degradation, for their transport out of the lysosomes and clearance by the cells.

Transition metal absorption through mucolipins

When cells engulf external medium during endocytosis, they capture extracellular macromolecular complexes that are almost inevitably bound to metal ions, including transition metals (TM), a group that contains chemically active elements such as Cu²⁺, Fe²⁺, Fe³⁺, Zn²⁺ and Co²⁺. Under normal conditions, the absolute majority of TMs are bound to plasma proteins such as albumin, transferrin and ceruloplasmin [52–54]. Degradation of these proteins in the endocytic pathway releases the metals into the lysosomal lumen. The active chemical environment of the lysosomal lumen promotes Fenton-like chemical reactions resulting in generation of free radicals, which are damaging to the lysosomal membrane and result in the production of the inert chemical compound called lipofuscin [55, 56]. The acute exposure to high concentrations of TMs has been linked to lysosomal storage phenotypes [57–59] and to lysosomal permeabilization [60, 61]. The long-term exposure to transition metals, e.g. in postmitotic cells, has been linked to buildup of lipofuscin, which inhibits lysosomal function and autophagy [62, 63]. Autophagy deficits have been shown in several models of LSDs [17, 23, 43, 64–70] and in TMs overload [71, 72]. According to the lysosomal-mitochondria axis models of aging and cell death in lysosomal storage diseases, the suppression of autophagy results in buildup of dysfunctional organelles, specifically mitochondria [62, 63]. We suggested that the loss of mitochondrial Ca²⁺ buffering capacity and of the Ca²⁺-driven ATP production positive feedback loop makes cells vulnerable to the pro-apoptotic effects of stimulation with Ca²⁺ mobilizing agonists (neurotransmitters, hormones, growth factors) (Fig. 3) [43, 69].

Figure 3. A model depicting mitochondrial deterioration and cell death in storage diseases.

In the normal cells (left) Ca²⁺ fluxes induced by hormones and neurotransmitters (step 1) are buffered by the energized mitochondria. The Ca²⁺ dependent components of oxidative phosphorylation chain (step 2) respond by producing more ATP. The resulting spike in ATP promotes Ca²⁺ extrusion and prevents pro-apoptotic effects of Ca²⁺. In cells with lysosomal storage diseases (right), the general mitochondrial function is impaired due to buildup of dysfunctional mitochondria (step 4), resulting in the loss of Ca²⁺/ATP-driven feedback loop, which leads to pro-apoptotic effects of Ca²⁺ and cell death.

The detrimental effects of transition metals in the lysosomes, combined with the cellular requirements for transition metals necessitate extraction of the metals from the lysosomal lumen into the cytoplasm, where their effects can be mitigated by chelating proteins or by extrusion out of the cell [52, 54]. Existence of such mechanisms has been shown before. Specifically, SLC11A1 (NRAMP1) and SLC11A2 (NRAMP2, DMT1) are ion transporters responsible for divalent cation uptake by cells [53, 54]. The intracellular localization of SLC11A2 and its apparent role in export of divalent cations from the lumen of lysosomes led to the suggestion that SLC11A2 is, or is part of, the endocytic divalent cation absorption pathway [53, 54]. As discussed above, it is possible that a range of ion channels may be involved in the endocytic function. The recently published data suggest that NRAMP may not be the only TMs absorption mechanism in the endocytic pathway [73, 74].

At the time of publication of the original report, the TRPML1 activation pathway has not been known and therefore activating mutations have been introduced into TRPML1 in order to detect current carried by this channel. The mutations mimic the mutation causing the varitint-waddler phenotype in TRPML3 [29–33, 35, 75], which is a spontaneously arisen gain-of-function mutation in mice leading to pigmentaiton defects and hearing loss [76]. Using this approach, TRPML1 was found to be permeable to Fe²⁺ [74]. TRPML1 was proposed to function as a lysosomal Fe²⁺ leak channel that exports Fe²⁺ from the lysosomal lumen into the cytopasm to prevent Fenton reactions and the buildup of lipofuscin catalysed in the lysosomes by Fe²⁺. The previously demonstrated buildup of auto-fluorescent material in MLIV patient skin fibroblasts was interpreted as lipofuscin and byproduct of abnormally high lysosomal Fe²⁺ levels [74]. MLIV patient fibroblasts were shown to contain higher levels of Fe²⁺ than matched controls, that is presumably present in the lysosomes [74].

The central paradigm of the Fe²⁺-transport model of TRPML1 function and MLIV pathogenesis is that TRPML1 is a lysosomal Fe²⁺ leak pathway, in the absence of which lysosomes become overloaded with Fe²⁺, leading to the loss of lysosomal function. One would expect then that the lysosomal phenotype of TRPML1 loss should be the same as the lysosomal phenotype of Fe²⁺ overload. Fig. 4 shows electron micrographs of fibroblasts obtained from MLIV patient and of fibroblasts obtained from a heterozygous relative that were treated with 100 µM of Fe²⁺ for 48 hours. Close analysis of storage bodies in the two samples shows that morphology of the inclusions present in Fe²⁺ treated cells is different from those in present MLIV knockdown cells. Therefore, the effects of TRPML1 loss and lysosomal Fe overload do not completely overlap.

Figure 4. Analysis of inclusion bodies in MLIV fibroblasts and fibroblasts treated with Fe²⁺.

Left: electron micrograph of an MLIV fibroblast (clone WG0909 from the McGill University cell line collection). Right: electron micrograph of a control fibroblast (clone WG0987 from the McGill University cell line collection, which is a heterozygous relative of WG0909) treated with 100 µM Fe²⁺ for 48 hours. Note the difference in inclusion morphology: the presence of large numbers of inclusions containing enfolded membranes in MLIV cells (designated by *), which are rare or absent in Fe²⁺-treated cells.

The second line of evidence for TRPML1 acting as a lysosomal heavy metal transporter comes from recent demonstration of Zn²⁺ accumulation in MLIV fibroblasts [73]. Using a combination of electrophysiology and chemical analysis, Zn²⁺ permeability through TRPML1 was observed, alongside with an increase in total cellular and lysosomal Zn²⁺ content. In these studies, the mutant, “activated”, form of TRPML1 was used as well. It is unclear exactly how well the TRPML1 “activated” by introducing the varitint-waddler mutations approximates the function of the wild type channel. In fact at least some of the aspects of TRPML physiology, such as regulation by pH, seem to dramatically change in the “activated” forms [35]. In TRPML3, the activation mutation changes the pore properties of the channel and its selectivity to divalent ions [77]. Demonstrating Fe²⁺ and Zn²⁺ permeability in wild type channel activated by PI(3,5)P2 would help resolved this question. Another question that remains to be resolved is whether the net increase in the metals associated with storage bodies can be explained by their binding to the undigested material. It is also important to note that the cellular response to TRPML1 loss may involve changes in expression of proteins that handle TMs. Microarray analyses of MLIV fibroblasts revealed a large number of genes whose activity changed, ostensibly as a result of TRPML1 loss [78]. It is possible and, indeed, likely, that some of these changes may affect transport of TMs across the endocytic membranes, or their retention by the storage bodies.

Despite of these questions, the findings discussed above open an interesting possibility for existence of a novel metal export pathway from the lysosomes that their aberrant function in MLIV and perhaps other LSDs play a critical role in the disease phenotype. Further studies, including comparative analysis of SLC11A1 and TRPML1 localization together with analysis of Fe²⁺ buildup in different tissues will be important for solidifying this exciting model.

Summary

Since its identification in 2000, TRPML1 provided an exciting opportunity to learn about lysosomal function as well as the function of the entire cell. Further advances in understanding TRPML1 transport and activation properties are likely to facilitate our understanding of such central cellular processes as metal handling and the fusion of organelles in the endocytic pathway. If TRPML1 functions not as a Ca²⁺ channel but as a heavy metals channel that have important role in fusion events associated with membrane trafficking, it will be of paramount interest to find out how aberrant heavy metals transport by the lysosomes leads to defective endocytosis. The central paradigms learned in this process are likely to extend beyond TRPML1 and MLIV pathogenesis, and to other biological processes and molecules. Furthermore, since a number of system-wide processes such as growth factor signaling and antigen handling depend on lysosomal function, TRPML1 activation and transport mechanisms are likely to affect research beyond the storage disease area.

Among other important questions raised by the TRPML1 ion transport and activation mechanism described above are the cell wide responses to TRPML1 loss and TRPML1 interaction with other transport molecules and regulatory proteins. As discussed above, the relative contribution of SLC11A2 and TRPML1 in divalent metals export from the lysosomes is unclear and comparative analysis of their localization would be extremely useful in delineating their roles. Similarly, recent identification of the TPCs as lysosomal Ca²⁺ channels raises the question of the role sharing between these two types of channels. Answering these and other questions pertaining to the activity of the lysosomal ion transport pathways will unquestionably improve our understanding of the lysosomal function and the cell- and organism-level activities that depend on it.