The P2X4 receptor has a widespread distribution in the central nervous system and the periphery, and plays an important role in the function of immune cells and the vascular system. Its upregulation in microglia contributes to neuropathic pain following nerve injury. The mechanisms involved in its regulation are not well understood, although we have previously shown that it is constitutively retrieved from the plasma membrane and resides predominantly within intracellular compartments. Here, we show that the endogenous P2X4 receptors in cultured rat microglia, vascular endothelial cells and freshly isolated peritoneal macrophages are localized predominantly to lysosomes. Lysosomal targeting was mediated through a dileucine-type motif within the N-terminus, together with a previously characterized tyrosine-based endocytic motif within the C-terminus. P2X4 receptors remained stable within the proteolytic environment of the lysosome and resisted degradation by virtue of their N-linked glycans. Stimulation of phagocytosis triggered the accumulation of P2X4 receptors at the phagosome membrane. Stimulating lysosome exocytosis, either by incubating with the Ca2+ ionophore ionomycin, for normal rat kidney (NRK) cells and cultured rat microglia, or the weak base methylamine, for peritoneal macrophages, caused an upregulation of both P2X4 receptors and the lysosomal protein LAMP-1 at the cell surface. Lysosome exocytosis in macrophages potentiated ATP-evoked P2X4 receptor currents across the plasma membrane. Taken together, our data suggest that the P2X4 receptor retains its function within the degradative environment of the lysosome and can subsequently traffic out of lysosomes to upregulate its exposure at the cell surface and phagosome.
Extracellular ATP acts as an important signalling molecule regulating numerous biological processes throughout the nervous, immune, endocrine, respiratory and circulatory systems (Khakh and North, 2006). The effects of ATP are mediated through activation of plasma membrane purinergic P2 receptors, which are divided into the G-protein-coupled P2Y receptors and the ionotropic P2X receptors (Abbracchio and Burnstock, 1994).
The P2X receptors are a family of non-selective cation channels consisting of seven subtypes, P2X1 through P2X7, which differ in their cellular and subcellular distributions. The P2X4 receptor has a particularly widespread distribution and is found not only in the central nervous system but also in the periphery, including epithelia, endothelia and immune cells (Bowler et al., 2003; Guo et al., 2004; Naemsch et al., 1999; Yamamoto et al., 2000a; Yeung et al., 2004). It is the predominant subtype expressed in vascular endothelial cells, where it mediates a vasodilator response to shear stress and regulates vascular remodelling (Yamamoto et al., 2006). In macrophages and microglia, its role is less well understood, although it is also one of the predominant subtypes expressed, together with P2X7. It is upregulated in spinal cord microglia as a result of peripheral nerve injury, and its activation is required for the expression of neuropathic pain (Tsuda et al., 2003). A recent study by Coull et al. suggests that, after nerve injury, ATP-stimulated microglia release brain-derived neurotrophic factor (BDNF), which acts on neighbouring dorsal horn neurones to promote hyperexcitability (Coull et al., 2005).
The mechanisms involved in the upregulation of P2X4 receptors are unknown. In naïve animals, expression of P2X4 protein in the spinal cord is low, but it increases following nerve injury as a result of an increase in the number of activated microglia showing high levels of P2X4 and OX42 expression (Tsuda et al., 2003). This suggests that the gene encoding P2X4 might be one of several genes whose transcription is enhanced in microglial cells during the process of activation. Previously, we showed that, when P2X4 receptors are expressed heterologously in neurones, they undergo rapid, constitutive endocytosis from the plasma membrane (Bobanovic et al., 2002; Royle et al., 2002; Royle et al., 2005). This is a dynamin-dependent process mediated by the interaction of an unconventional tyrosine-based endocytic motif and adaptor protein 2 (AP-2). Less is known about the mechanisms of regulating P2X4 receptor trafficking in other cell types and what role this plays in determining the function of P2X4.
In this study, we examined the targeting and trafficking of endogenous P2X4 receptors in microglia, macrophages and endothelial cells. We find that the receptors are located predominantly within lysosomal compartments and are targeted there by N- and C-terminal motifs. They resist degradation by virtue of N-linked glycans and can subsequently traffic to phagosomes and to the plasma membrane. Stimulation of lysosome exocytosis in freshly isolated macrophages increased P2X4 receptor currents evoked by extracellular ATP, suggesting a novel mechanism of regulating an ionotropic receptor.
P2X4 receptors in primary macrophages, microglia and vascular endothelial cells are contained predominantly within lysosomes
P2X4 receptors are known to be expressed in macrophages, microglia and vascular endothelial cells and to play a role in the pathogenesis of pain and control of blood pressure (Tsuda et al., 2003; Yamamoto et al., 2006). We examined the subcellular distribution of the endogenous receptors in freshly isolated mouse peritoneal macrophages, cultured rat microglia and brain endothelial cells. Cells were permeabilized before labelling with an antibody against P2X4 and imaged by confocal immunofluorescence microscopy. All three cell types showed intense P2X4 immunofluorescence concentrated within intracellular puncta (Fig. 1A). There was extensive overlap between the distribution of P2X4 and the lysosome-associated membrane protein 1 (LAMP-1) but not with a marker for early endosomes (EEA-1). This differs from our prior observations of the distribution of heterologously expressed P2X4 in neurones, where there was considerable overlap between P2X4 and EEA-1 along the dendrites (Bobanovic et al., 2002). A single band of the appropriate size for the glycosylated form of the P2X4 receptor (64 kDa; Fig. 1B) was detected by western blot using cell lysates from immune and brain endothelial cells. A band of similar size plus a band of lower molecular mass corresponding to the non-glycosylated form of P2X4 was detected in normal rat kidney (NRK) cells transfected with P2X4 but not in nontransfected cells.
In the J774 macrophage cell line, P2X4 receptor expression was downregulated in comparison with that of the primary cells. The distribution of heterologously expressed GFP-tagged P2X4 receptors in these cells, however, was very similar to that of the endogenous receptors in primary cells. Receptors were localized to lysosomes, as shown by incubation with Lysotracker Red (Fig. 1D).
Identification of motifs that target P2X4 receptors to lysosomes
Next, we wanted to identify how P2X4 receptors traffic to lysosomes, and to do this we heterologously expressed in NRK cells vectors encoding either P2X4 with EGFP fused to the C-terminus (P2X4-GFP) or P2X4 with an AU5 epitope inserted into the extracellular loop (P2X4-AU5). Previously, these constructs were shown to traffic and function in a manner similar to that of the wild-type receptor (Bobanovic et al., 2002). The receptors were targeted to lysosomes, as shown by colabelling with either antibody against LAMP-1 or Lysotracker (supplementary material Fig. S1), but they could not be labelled en route by incubating live cells with extracellular antibody against AU5 at 37°C (Fig. 2A), except in ∼1% of transfected cells. This suggests that either the receptors traffic directly from the trans-Golgi network (TGN) to late endosomes or are retrieved from the plasma membrane too rapidly for the antibody to bind. Inhibiting clathrin-dependent endocytosis by coexpression of dominant-negative mutants of either AP180 (AP180-C) or dynamin I (K44A) (Henley et al., 1998; Marsh and McMahon, 1999) increased surface labelling of P2X4, receptors, suggesting that at least some of the receptors traffic via the plasma membrane (supplementary material Fig. S2).
P2X4 receptors contain two tyrosine-based motifs within the C-terminus (Y372xxV and Y378xxGL), and we have shown that the non-canonical motif at position 378 preferentially interacts with AP-2 to mediate endocytosis (Royle et al., 2002; Royle et al., 2005). Consistent with this, we show here that mutation Y378F substantially increased surface labelling of receptors expressed in NRK cells, whereas the Y372F mutation did not (supplementary material Fig. S3A). A previously described fusion protein incorporating the transmembrane region of CD8 and the C-terminus of P2X4 undergoes rapid endocytosis (Royle et al., 2005) but did not colocalize as extensively with LAMP-1 as the intact P2X4 receptor (supplementary material Fig. S3B). This suggests that motifs outside of the C-terminus might also be involved in the lysosomal targeting of P2X4. Within the N-terminus of P2X4 is a dileucine-type motif (L22I23) similar to the motif shown to mediate the direct trafficking of the CLN3 protein from the TGN to late endosomes and lysosomes (Kyttala et al., 2004). We compared the effects of mutating both L22 and I23 to alanine and Y372 and Y378 to phenylalanine on the labelling of receptors at the surface and their subsequent retrieval. Both double mutations increased surface labelling so that most transfected cells were stained with antibody against AU5 (Fig. 2A). To quantify the effects, we compared surface versus total P2X4 staining post permeabilization, selecting only those wild-type-expressing cells that showed some surface labelling. The effects of mutating the dileucine and tyrosine motifs were similar; both produced a significant increase in the ratio of surface to total labelling (Fig. 2B) and they also increased the fraction of surface-labelled receptors that remained at the surface rather than being internalized (Fig. 2C). When the N- and C-terminal mutations were combined, there was a considerable further increase in receptor expression at the plasma membrane, as shown by antibody labelling and surface biotinylation (Fig. 2A-D). Similar results were obtained with the GFP-tagged receptors when the extent of receptor colocalization with LAMP-1 was compared for wild-type cells and mutants. Both double mutations reduced the degree of colocalization with LAMP-1 (Fig. 2E), but there was a much greater reduction when the N- and C-terminal mutations were combined. The simplest interpretation of these results is that the N-terminal dileucine and C-terminal tyrosine-based motifs independently contribute to both endocytosis and lysosomal targeting of P2X4 receptors.
P2X4 receptors in lysosomes resist proteolysis by virtue of N-linked glycans
Our finding that P2X4 receptors are concentrated within lysosomes suggests that they might resist degradation within this proteolytic environment. To investigate this further, we compared how the levels of heterologously expressed P2X4 receptors and those of another transmembrane receptor, the M4 muscarinic receptor, decreased over time following the incubation of NRK cells with cycloheximide (CHX). There was a substantial drop in the level of the M4 receptor within 12 hours, whereas there was little decline in P2X4 receptor levels over 24 hours (Fig. 3A). Similarly, there was very little decline in the level of endogenous P2X4 in the Thp-1 monocytic cell line after 24 hours in CHX. From this, we conclude that P2X4 receptors are able to resist rapid degradation within lysosomes.
Other lysosome-targeted proteins, such as LAMP-1, have been shown to resist proteolysis by virtue of N-linked glycans decorating the lumenal side of the protein (Kundra and Kornfeld, 1999). To investigate whether glycans play a role in protecting the P2X4 receptor, we first identified how many sites normally were glycosylated. The P2X4 receptor has seven potential N-linked glycosylation sites (Hu et al., 2002) and shows partial resistance to treatment with endoglycosidase H (Endo H), indicating that it undergoes complex glycosylation in the Golgi (Fig. 3B). When NRK cells were cultured in deoxymannojirimycin (DMJ; an inhibitor of Golgi α-mannosidase I) post-transfection, incubation of the solubilized cell lysate with Endo H completely removed all glycans from P2X4, giving rise to a single band of the same size as observed following treatment with N-glycosidase F. This indicates that DMJ prevents complex glycosylation of P2X4. It did not, however, alter the trafficking of the receptor to lysosomes (supplementary material Fig. S4). In a similar experiment, incubation of the cell lysate with a very low concentration of Endo H produced six bands migrating at slightly different molecular masses, corresponding to the different glycosylated states of the receptor (Fig. 3C). To test how removal of N-glycans in vivo by Endo H affected the rate of degradation of P2X4, NRK cells were incubated in the presence of DMJ post transfection, and then with Endo H for different time periods. In cells treated with Endo H, there was a decrease in the size of the P2X4 band, corresponding to de-glycosylation, and a reduction in band intensity over a period of 12-24 hours, whereas, in the absence of Endo H, P2X4 receptor levels were maintained (Fig. 3D). A similar result was obtained for the endogenous P2X4 receptor in Thp-1 cells. Finally, we compared the effect of de-glycosylation on the rate of degradation of LAMP-1 and P2X4 in RAW264.7 cells and observed a rapid loss of both proteins from cells incubated in Endo H (Fig. 3E). From this, we conclude that lysosome-resident P2X4 receptors are stable and resist degradation by virtue of the N-linked glycans on the extracellular loop, similar to LAMP-1.
Trafficking of P2X4 receptors from lysosomes to phagosomes
For proteins that resist degradation, delivery to lysosomes is not necessarily an endpoint in their trafficking pathway. Lysosome exocytosis and fusion of lysosomes with the developing phagosome delivers membrane proteins as well as soluble components to these other compartments. In macrophages, phagocytosis can be triggered by means of activation of different signalling pathways, and this subsequently has an effect on downstream events (Stuart and Ezekowitz, 2005). To investigate the trafficking of P2X4 from lysosomes to phagosomes, we compared the localization of P2X4 in peritoneal macrophages in response to a variety of stimuli, including zymosan, latex beads and live Escherichia coli (DH5α) expressing GFP. In all cases, P2X4 receptors were concentrated in the phagosome surrounding the ingested particle, colocalized with LAMP-1 (Fig. 4A). Similar results were obtained with P2X4-GFP expressed in RAW267.4 cells, after incubation with latex beads (Fig. 4B). The accumulation of P2X4 receptors at the phagosome membrane was also demonstrated by the isolation of phagosomes containing latex beads from bone-marrow-derived macrophages by flotation on a sucrose gradient. Phagosome membranes were subject to SDS PAGE followed by western blotting with antibodies against either P2X4 or LAMP-1. The proportion of total P2X4 within the phagosome fraction was similar to that of LAMP-1 (Fig. 4C).
Lysosome exocytosis delivers P2X4 receptors to the plasma membrane
In addition to the secretion of lysosome-related organelles by specialized cells such as cytotoxic T cells, exocytosis of conventional lysosomes has been shown to occur in nonspecialized cells and to be triggered by a variety of stimuli, including wounding, a rise in cytosolic Ca2+ and infection by Trypanosoma cruzi (Blott and Griffiths, 2002; Reddy et al., 2001; Rodriguez et al., 1997; Tardieux et al., 1992). This process is a potential mechanism for upregulation of P2X4 receptors at the plasma membrane. To investigate this further, we tested the effect of the Ca2+ ionophore ionomycin on lysosome secretion and surface exposure of P2X4 receptors and LAMP-1 in transfected NRK cells. Similar to previous reports, incubation with ionomycin increased the release of the lysosomal enzyme beta hexosaminidase (β-Hex; Fig. 5A) (Rodriguez et al., 1997). P2X4 and LAMP-1 were detected at the cell surface by using antibodies to lumenal/extracellular domains and there was a large increase in surface immunolabelling following treatment with ionomycin (Fig. 5A). Similar results were obtained in cultured rat microglia; the surface expression of endogenous P2X4 receptors, as measured by biotinylation followed by western blot analysis, and of LAMP-1 was increased following exposure to ionomycin (Fig. 5B).
In primary macrophages, lysosome secretion is triggered by agents that increase lysosomal pH, including the weak base methylamine (MA) and inhibitors of the H+-ATPase (Tapper, 1996; Tapper and Sundler, 1990; Tapper and Sundler, 1995). Pre-incubation of peritoneal macrophages with MA for different time periods stimulated the release of β-Hex, with the maximal rate of release occurring between 10 and 20 minutes after exposure (Fig. 5C). This was cell-type dependent because, in human embryonic kidney 293 (HEK 293) cells, MA failed to stimulate the release of β-Hex. Incubation of cultured primary peritoneal macrophages for 15 minutes with MA caused a substantial increase in the surface biotinylation of both P2X4 and LAMP-1 (Fig. 5D), again indicating that the endogenous P2X4 receptors traffic from lysosomes to the surface following exocytosis. Another reported trigger of lysosomal secretion in human monocytes is ATP acting at P2X7 receptors (Andrei et al., 2004). We compared the effects of 1 mM ATP on β-Hex release from peritoneal macrophages with and without pretreatment with lipopolysaccharide (LPS). The combination of LPS and ATP caused an approximately fourfold increase in β-Hex release, but it also caused a large increase in the release of the cytoplasmic enzyme lactate dehydrogenase (LDH), indicating disruption of the integrity of the cell (supplementary material Fig. S5). By contrast, incubation in MA for up to 60 minutes had no effect on the release of LDH.
Lysosome exocytosis potentiates P2X4 receptor currents
To test whether the recruitment of P2X4 receptors from lysosomes to the cell surface results in an increase in responsiveness of cells to ATP, whole-cell currents were recorded from peritoneal macrophages using the perforated patch-clamp technique, before and after treatment with MA. Inward currents were measured in response to rapid application of 30 μM Mg2+-ATP, which should maximally activate P2X4 receptors (North, 2002) but have little effect on P2X7. In the control group of cells, the currents were small (∼40 pA at –60 mV) but were potentiated approximately threefold by the P2X4-selective allosteric modulator, ivermectin (Fig. 6A) (Khakh et al., 1999; Priel and Silberberg, 2004). In some cells, the inward current was followed by a delayed outward current, presumably mediated by a P2Y receptor-mediated Ca2+-dependent conductance, and in ∼10% of cells there was a rapidly activating and very transient inward current, characteristic of P2X1 receptors. This component of the response was easy to separate from the P2X4-mediated currents because of its rapid kinetics, and it was not included in the analysis. The preferred agonist at P2X7 receptors, 2′, 3-O-(4-benzoyl) benzoyl-ATP (BzATP), evoked currents of a magnitude similar to that evoked by ATP when applied at 30 μM (Fig. 6A), indicating that surface expression of P2X7 as well as P2X4 receptors was low.
Following incubation with 50 mM MA at 37°C, macrophages were transferred to the patching chamber and perfused with extracellular solution. The peak current amplitude in response to 30 μM ATP was increased fivefold following MA treatment (P<0.001; Fig. 6B). The currents were still ivermectin sensitive and were potentiated a further threefold in the presence of this modulator. The co-application of MA with ATP had no effect on peak current amplitudes, suggesting that there was no direct effect of MA on the channel. In addition, brefeldin A, a fungal metabolite that inhibits protein trafficking between the Golgi apparatus and the cell surface, had no effect on the MA-induced increase in ATP-evoked currents, suggesting that it was not dependent upon the delivery of newly synthesized receptors through the secretory pathway to the plasma membrane (Fig. 6B). The amplitude of the response evoked by 30 μM BzATP was now only a third of the response to ATP, indicating that the MA-induced upregulation was selective for the P2X4 subtype (Fig. 6C). Finally, in HEK 293 cells expressing P2X4 receptors, incubation with MA for 30 minutes produced no increase in the amplitude of inward currents evoked by 30 μM ATP (Fig. 6D), which is consistent with its inability to promote lysosome exocytosis in these cells. Taken together, our results show that there is a selective upregulation of P2X4 receptor currents in macrophages following lysosome exocytosis that is independent of the delivery of newly synthesized receptors to the cell surface. This suggests that lysosomal P2X4 receptors retain their functional integrity and provide a reserve pool that can be mobilized to other compartments.
Our results reveal a novel mechanism for the regulation of a ligand-gated ion channel. The endogenous P2X4 receptors in microglia, endothelial cells and macrophages are localized primarily in lysosomes, and this targeting is mediated by N-terminal dileucine and C-terminal tyrosine motifs. The constitutive retrieval of P2X4 from the plasma membrane and targeting to lysosomes limits the number of receptors expressed at the surface. For most ligand-gated ion channels, delivery to lysosomes represents an end point in both their trafficking and function within the cell. P2X4 receptors, however, are surprisingly stable within this degradative environment, which suggests they could have subsequent roles either within intracellular compartments or at the plasma membrane. After stimulating lysosome exocytosis, P2X4 receptor-mediated currents were enhanced, indicating that lysosome-resident receptors retain their function and provide a pool that can be mobilized to upregulate the responsiveness of the cells to extracellular ATP.
P2X4 receptors are one of the predominant subtypes expressed in vascular endothelial cells, microglia, macrophages and monocytes, and their regulation at the plasma membrane will therefore be an important determinant of the responsiveness of these cells to extracellular ATP (Bowler et al., 2003; Coull et al., 2005; Guo et al., 2004; Guo et al., 2005; Tsuda et al., 2003; Wang et al., 2004; Wang et al., 2002; Yamamoto et al., 2000b; Zhang et al., 2006). In vascular endothelial cells, the receptors play a central role in the response to changes in blood flow, and their expression is also regulated by blood flow (Korenaga et al., 2001; Yamamoto et al., 2000a). They mediate the Ca2+ response induced by shear stress and the subsequent production of nitric oxide, and P2X4-knockout mice are not only hypertensive but show impaired flow-dependent vascular remodeling (Yamamoto et al., 2006). In microglia/macrophages, P2X4 receptor upregulation and activation plays a central role in neuropathic pain caused by injury to peripheral nerves (Tsuda et al., 2003). Here, the receptors are involved in the release of inflammatory mediators such as BDNF and cytokines (Coull et al., 2005). Upregulation of microglial P2X4 receptors is also caused by glucose/oxygen deprivation (Cavaliere et al., 2005), and, in the periphery, upregulation of P2X4 is similarly associated with pathology, for example, dystrophic muscle shows an increase in P2X4 expression associated with infiltrating macrophages (Yeung et al., 2004).
Lysosome exocytosis provides a mechanism for rapidly upregulating P2X4 receptors at the plasma membrane, and it can be triggered by a rise in intracellular Ca2+ acting by means of the Ca2+-binding protein synaptotagmin VII (Arantes and Andrews, 2006). Given the very high permeability to Ca2+ of P2X receptors, it is possible that there exists a positive-feedback loop whereby activation of these receptors facilitates the recruitment of additional P2X4 receptors to the cell surface. Macrophages, microglia, monocytes and vascular endothelial cells express P2X7 receptors in addition to P2X4 (Cavaliere et al., 2005; Di Virgilio et al., 2001; Rassendren et al., 1997; Wilson et al., 2004), and activation of P2X7 in human monocytes promotes lysosome exocytosis (Andrei et al., 1999; Andrei et al., 2004). We measured an increase in the release of lysosomal β-Hex from macrophages following exposure to millimolar ATP, although the accompanying release of LDH suggests that cell lysis also occurred. Recruitment of P2X4 receptors by means of the activation of the much lower-affinity P2X7 receptor would provide a mechanism for sensitizing cells to ATP following an initial exposure to a high concentration of this ligand. In macrophages, lysosomal secretion is triggered by agents that cause a rise in lysosomal pH (Sundler, 1997), and, although experimentally this has been demonstrated using agents such as MA, of more physiological relevance, the alkalinization and secretion of lysosomes is also triggered by the interaction of mononuclear phagocytes with sterile inflammatory agents such as zymosan particles (Schorlemmer et al., 1977). Lysosome exocytosis also occurs in response to wounding and is thought to be important in mediating plasma membrane repair in fibroblasts (Reddy et al., 2001). A role for P2X4 receptors in this process has not been established, although the ATP released during wounding is reported to facilitate membrane repair (Yin et al., 2007). Damage to endothelial cells by shear stress could provide a mechanism for delivering lysosomes to the plasma membrane, leading to increased P2X4 responsiveness.
As well as considering the regulation of P2X4 receptors at the plasma membrane, our results suggest the possibility of a functional role for these receptors within lysosomes and phagosomes, although there is as yet no direct evidence for this. A recent study by Fountain and colleagues (Fountain et al., 2007) identified a protein in Dictyostelium discoideum that is weakly related to vertebrate P2X receptors and that appears to act intracellularly. This receptor, when expressed in HEK 293 cells, formed channels in the plasma membrane activated by extracellular ATP. In Dictyostelium, however, it is localized to the intracellular contractile vacuoles rather than the plasma membrane and it plays a role in the regulation of cell volume and the emptying of these vacuoles. How intracellular P2X receptors might be activated is unclear. The ATP-binding site is predicted to face the lumen, and the protective role of the N-linked glycans that decorate the extracellular loop is consistent with this. Within the phagosome, P2X4 receptors could be activated by ATP released from phagocytosed material. The initial pH of the phagosome has been reported to rise to pH 8, and this is important for microbicidal activity (Reeves et al., 2002; Segal et al., 1981). P2X4 currents are potentiated by a pH above 7.4 and so could initially be functional in the phagosome (Clarke et al., 2000). As the phagosomes acidifies, P2X4 function is likely to be inhibited by the acidic pH (Clarke et al., 2000; Stoop et al., 1997; Wildman et al., 1999). A recent study by Zhang and colleagues (Zhang et al., 2007) showed that lysosomes in astrocytes contain abundant ATP, and this is released in a stimulus-dependent manner upon lysosome exocytosis. If this is a general feature of lysosomes in other cell types, then the ATP levels within phagosomes will be elevated upon fusion of lysosomes. Alternatively, there could be additional mechanisms for activation of intracellular P2X receptors such as the NAD-dependent ribosylation previously described for the P2X7 receptor (Seman et al., 2003) or even mechanical force. P2X4 receptor currents at the plasma membrane of vascular endothelial cells are mechanosensitive, and this might involve the direct recognition of membrane stretch by the receptor (Yamamoto et al., 2000a). The activation of P2X4 receptors at the phagosome membrane would provide a route for Ca2+ flux from the vacuole to the cytoplasm and would thus contribute to a localized Ca2+ signal. Other ion channels, for example, the BK channel in neutrophils and the cystic fibrosis transmembrane conductance regulator (CFTR) in alveolar macrophages, accumulate at the phagosome membrane and are reported to be involved in phagosome function and the microbicidal activity of these cells (Ahluwalia et al., 2004; Di et al., 2006). In lysosomes, it is tempting to speculate that P2X4 receptors could form a Ca2+-release channel such as that gated by nicotinic acid adenine dinucleotide phosphate (NAADP) (Churchill et al., 2002) or be involved in the pH regulation of lysosomal Ca2+ in macrophages (Christensen et al., 2002). A recent study by McGuiness and colleagues suggests that, in hippocampal neurones (McGuinness et al., 2007), Ca2+ release from axonal lysosomes contributes to action-potential-induced Ca2+ transients, and a role for P2X4 receptors in synaptic potentiation was recently demonstrated using neurones cultured from P2X4-knockout mice (Sim et al., 2006).
In conclusion, we have identified a novel means of regulating the functional expression of an ionotropic receptor, which involves its initial targeting to lysosomes and subsequent exposure at the cell surface and the phagosome.
Materials and Methods
Antibodies and reagents
The following primary antibodies were used: anti-P2X4 (1:100, Alomone Labs, Jerusalem, Israel), anti-LAMP-1 (1:250, 1D4B, Santa-Cruz Biotechnologies, CA; 1:250, LY1C6, Abcam, UK), anti-GFP (1:500, Abcam, UK), anti-lgp120 (1:250; a kind gift from J. P. Luzio, Cambridge, UK). Fluorescein isothiocyanate (FITC)- and indocarbocyanine (Cy3)-conjugated goat anti-mouse, anti-rabbit and anti-rat secondary antibodies (1:250, Jackson ImmunoResearch, West Grove, PA). HRP-conjugated secondary antibodies: anti-rabbit (Amersham Biosciences, UK), anti-mouse (Pierce, UK) and anti-rat (Abcam, UK). All other reagents were from Sigma Aldrich or Invitrogen (UK). E. coli DH5 alpha expressing gfp+ were kindly provided by M. Niederweiss (University of Birmingham, AB) (Scholz et al., 2000).
The construction and characterization of P2X4 receptors with enhanced green fluorescent protein fused to the C-terminus (P2X4-GFP) has been described previously (Bobanovic et al., 2002). Briefly, to generate cDNA encoding P2X4 with GFP fused to the C-terminus, the rat cDNA (a kind gift from P. P. A. Humphrey) was amplified by PCR using oligonucleotide primers to introduce a Kozak initiation sequence (Kozak, 1987), to remove the stop codon and to introduce NheI and SacII sites at the 5′ and 3′ ends, respectively. Amplification products were then cloned into the pEGFP-N1 vector (Clontech, CA). The construction and characterization of P2X4 receptors containing the AU5 epitope inserted into the extracellular loop has also been described previously (Bobanovic et al., 2002). Briefly, the AU5 epitope (TDFYLK) was substituted into P2X4 at position 76 (TSQLGF) by two-step PCR. Point mutations were made by either two-step PCR or the Quick Change II site-directed mutagenesis kit (Stratagene). For expression of either wild-type or AU5-tagged P2X receptors, the coding sequence of P2X4 was amplified to reintroduce the stop codon and an XbaI site. The subcloning of these fragments into the pEGFP-N1 vector excised the coding sequence of GFP. The sequences of all amplified regions were verified using automated DNA sequencing (Dept of Genetics, University of Cambridge, UK). CD8 containing the C-terminus of P2X4 was generated as described previously (Royle et al., 2005).
Dominant-negative AP180 (AP180-C)-RFP and dynamin (K44A)-RFP were kindly provided by S. J Royle. The M4 muscarinic GFP-tagged receptor was kindly provided by J. M. Edwardson.
NRK and human embryonic kidney 293 (HEK 293) cells were maintained in DMEM containing 10% foetal bovine serum and 100 U/ml penicillin-streptomycin at 37°C and 5% CO2. J774, RAW264.7 and Thp-1 cells were maintained in RPMI containing 10% foetal bovine serum and 100 U/ml penicillin-streptomycin at 37°C and 5% CO2. For immunocytochemistry and electrophysiology, cells were plated on poly-l-lysine-coated coverslips. For lysosome exocytosis assays, cells were plated in 24-well plates. Cells were plated 12 hours before they were assayed or transfected.
Mouse peritoneal macrophages were isolated from adult male CD-1 mice (Charles River, UK; 4-7 weeks) as described previously (Le Feuvre et al., 2002). Briefly, mice were sacrificed and the peritoneal cavity lavaged using 8 ml RPMI containing 10% foetal bovine serum and 100 U/ml penicillin-streptomycin. Cells were collected by centrifugation at 800 g, resuspended in fresh medium and plated at a density of 0.5×106 cells/well and on poly-l-lysine-coated coverslips for electrophysiology and immunocytochemistry or on 24-well plates for β-Hex and LDH release assays. Macrophages were allowed to adhere for 1 hour before washing with fresh medium to remove non-adherent cells and then left for a further 2 hours at 37°C, 5% CO2. Cells were primed for 2 hours with 1 μg/ml LPS (Escherichia coli 026:B6; Sigma-Aldrich, UK).
Bone-marrow-derived macrophages (BMDMs) were derived from male CD-1 mice that were 5-6 weeks old. Mice were sacrificed, the femur was excised, and the epiphyses removed before flushing out the bone marrow. Cells were washed and resuspended in RPMI 1640 supplemented with 20% foetal calf serum (FCS), 100 U/ml of penicillin/streptomycin and 40 ng/ml of recombinant macrophage colony stimulating factor (M-CSF; Promokine). Cells were cultured for 7 days before use in phagosome isolation experiments.
Rat brain microglia were prepared as described previously (Yu et al., 2005), and cells collected by shaking and plated onto uncoated coverslips or six-well plates. Rat brain endothelial cells (RBECs) were kindly provided by J. Lim (University of Cambridge, UK). Protein synthesis was blocked by incubating cells in normal culture medium containing 5 μg/ml CHX.
NRK and J774 cells were transfected using lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. HEK 293 cells were transfected using a modified calcium phosphate method as described previously (Royle et al., 2002). RAW264.7 cells were transfected using the AMAXA system according to the manufacturer's instructions. Cells were assayed the day after transfection.
Cells were fixed in 3% paraformaldehyde (PFA) and 4% sucrose in PBS for 15 minutes at room temperature (RT). If required, permeabilization of NRK cells was carried out using 0.1% Triton X-100 in PBS for 10 minutes at 4°C. For permeabilization of macrophages, 0.025% saponin was included in the blocking buffer at all steps. Nonspecific sites were blocked using PBS containing 4% normal goat serum and 3% bovine serum albumin (blocking buffer). Antibodies were diluted to their final concentration in blocking buffer and applied for 2 hours at RT. Cells were then rinsed five times with PBS, and secondary antibodies applied for 2 hours at RT. Finally, cells were washed five times with PBS and mounted onto slides with Vectashield (Vector Laboratories, CA) as a mounting medium.
For surface labelling of P2X4 AU5 constructs, transfected cells were incubated in anti-AU5 in serum-free culture medium for 30 minutes at 37°C. Cells were then washed five times and fixed using PFA. Cells were permeabilized using 0.1% Triton X-100, incubated in blocking buffer and then with an anti-mouse Cy3-conjugated secondary antibody for 2 hours at RT. Cells were then washed three times in PBS, and total receptors stained using anti-P2X4 followed by anti-rabbit FITC-conjugated secondary antibody, as described previously (Bobanovic et al., 2002).
For detection of surface and internalized receptors, cells were incubated with anti-AU5 and fixed as described above. Fixed, nonpermeabilized cells were then incubated in blocking buffer for 1 hour and then incubated with anti-mouse Cy3-conjugated secondary antibody for 2 hours, permeabilized with 0.1% Triton X-100 and incubated with anti-mouse FITC-conjugated secondary antibody for 2 hours at RT.
Cells were loaded with Lysotracker using 50 nM Lysotracker Red DND-99 (Molecular Probes) in cell culture media by incubation at 37°C for between 30 minutes and 1 hour. Cells were then washed once in normal extracellular solution (NES): 140 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM D-glucose, 10 mM HEPES (pH 7.3), placed in a live-cell imaging chamber in extracellular solution and visualized by confocal microscopy.
For immunostaining phagosomes, macrophages were incubated with zymosan or latex beads at a ratio of 1:30 for 30 minutes in normal culture medium at 37°C. Cells were then washed five times and left in culture medium for 1 hour at 37°C. For phagocytosis of gfp+ E. coli, macrophages were incubated with E. coli at a multiplicity of infection of 1:50 for 90 minutes.
Surface labelling of LAMP-1 was carried out by incubation of live cells with ionomycin (5 μM for 10 minutes) at 37°C in NES. Cells were then washed three times in NES at 12°C and incubated with anti-LAMP-1 (LY1C6) in NES at 12°C for 30 minutes. Cells were then washed five times with NES at 12°C, fixed with 3% PFA, and labelled LAMP-1 was then detected with anti-mouse Cy3-conjugated antibody.
Image collection and analysis
Fluorescence was visualized using a Zeiss Axiovert LSM510 confocal microscope, using 63× or 100× oil-immersion objectives. Identical acquisition parameters were used for image capture of individual experiments. Images were imported into ImageJ (Wayne Rasband, NIH), cells outlined and mean pixel values obtained.
All experiments were repeated at least three times, analysing at least 20 cells each time. When one cell is shown, this cell is representative of the general cell population on the coverslip. Colocalization analysis was performed using LSM 510 software. A threshold was set corresponding to average background fluorescence (determined by an area without cells). Cells were then outlined and the Pearson's correlation coefficient obtained.
Cells were washed twice with phosphate-buffered saline (PBS: 1.5 mM NaH2PO4, 8 mM Na2HPO4 and 145 mM NaCl, pH 7.3) and collected directly into lysis buffer (50 mM Tris-Cl, pH 8.0, 150 mM NaCl, 0.1% SDS, 1% Nonidet P-40 and 0.5% sodium deoxycholate). The cell lysate was sonicated and left on ice for 30 minutes. Samples were cleared by centrifugation at 21,000 g for 15 minutes at 4°C and incubated with SDS polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer. Proteins were separated by 10% SDS-PAGE (Laemmli, 1970) and transferred to nitrocellulose membranes. The membrane was probed with anti-P2X4 (1:500), anti-LAMP-1 (1:500) or anti-GFP (1:500) antibody, followed by horseradish peroxidase (HRP)-conjugated secondary antibody and detected with the enhanced chemiluminescent substrate (ECL; Pierce).
Cells were washed once with ice-cold PBS and incubated with 1 ml sulfo-NHS-SS-biotin (Pierce) solution (freshly prepared; 1 mg/ml in PBS) for 20 minutes at 12°C. Excess biotin was quenched by washing the cells once with PBS containing 50 mM glycine and twice with PBS. Cells were solubilized with lysis buffer, incubated on ice for 30 minutes, after which time they were sonicated and cleared by centrifugation. A portion of the supernant was incubated with immobilized streptavidin-biotin-binding protein beads (Pierce) on a rotating rack for 2 hours at 4°C in order to precipitate biotinylated proteins. The rest of the supernatant was kept in order to assess total protein for each sample. After incubation, beads containing precipitated biotinylated proteins were spun for 1 minute at 15,000 g at a temperature of 4°C. The supernatant was removed, and the beads were washed with lysis buffer. This was repeated three times, and the protein was eluted from the beads by incubation in 20 μl Laemmli buffer. Proteins were separated by SDS-PAGE and detected by immunoblotting, as described above.
Removal of N-glycans by treatment with Endo H
To prevent complex glycosylation of proteins, NRK cells were incubated with 1 mM DMJ (Alexis) post-transfection and for at least 24 hours before addition of Endo H. Thp-1 and RAW264.7 cells were treated with 1 mM DMJ for at least 3 days before use and the DMJ-containing medium was replaced daily. In vitro deglycosylation of proteins with either Endo H or N-glycosidase F was carried out as described previously (Ormond et al., 2006). The method used to analyse the role of N-glycans in P2X4 receptor stability was essentially as described previously (Kundra and Kornfeld, 1999). Endo H (50 mU/ml) was added for the indicated number of hours at 37°C. Cells were subsequently washed, scraped in PBS and recovered by centrifugation. The cell pellet was solubilized with lysis buffer, subjected to SDS-PAGE and immunoblotted with anti-P2X4 or anti-LAMP-1 (1D4B clone).
Isolation of latex bead phagosomes
Phagosomes were isolated essentially as described previously (Desjardins et al., 1994). Phagosomes were formed by incubating 7-day-old BMDMs with latex beads (0.8 μm diameter, 10% suspension, blue dyed; Sigma) diluted at 1:200 in RPMI and incubated at 37°C for 2 hours and chased for 1 hour to allow internalization. The cells were washed, scraped in ice-cold PBS and pelleted by centrifugation at 100 g for 7 minutes. Cells were resuspended in homogenization buffer (HB: 250 mM sucrose, 3 mM imidazole, pH 7.4) at 4°C, washed once by centrifugation and resuspension in HB at 4°C. Cells were then homogenized by passage through a 22-gauge syringe needle. Unbroken cells were pelleted at 400 g for 7 minutes at 4°C and the supernatant, containing the phagosomes, was recovered and mixed with a 62% sucrose solution to bring it to a final concentration of 42% sucrose. The phagosomes were then isolated on a discontinuous sucrose gradient consisting of 62% sucrose, phagosome sample at 42% sucrose, 35% sucrose, 25% sucrose and 10% sucrose. The gradients were centrifuged in a swinging-bucket rotor at 100,000 g for 1 hour at 4°C. Following centrifugation, the phagosomes were isolated as a single band at the 10-25% sucrose interface. The phagosomes were resuspended in cold PBS and centrifuged at 40,000 g for 30 minutes at 4°C. The pellet was then resuspended in lysis buffer and analysed by western blotting.
Lysosome exocytosis and cell death assays
Cells were plated in a 24-well plate at 0.5×106 cells/well. Following stimulation, lysosome exocytosis was assayed by release of beta-hexosaminidase. 100 μl of supernatant was incubated with 200 μl 1 mM 4-nitrophenyl N-acetyl-β-D-glucosaminide in 0.05 M citrate buffer (0.025 M citric acid, 0.025 M trisodium citrate, pH 4.5) for 1 hour at 37°C. Reactions were stopped by addition of 500 μl of 0.05 M sodium carbonate buffer (pH 10). Absorbance was then measured at 405 nm. Cell death was assayed by the CytoTox 96 non-radioactive cytotoxicity assay (Promega, WI) according to the manufacturer's instructions. Total release was measured by lysing the cells with 1% Triton X-100.
Perforated patch recordings were performed at RT using an Axopatch 200A amplifier (Axon Instruments). Patch pipettes (3-8 MΩ) were pulled from thick-walled borosilicate glass (GC150F-10, Harvard Apparatus) and filled with intracellular solution (IS) containing 70 mM K2SO4, 10 mM KCl, 1 mM MgCl2, 10 mM HEPES and 200 μg/ml amphotericin B (pH 7.3). The normal extracellular solution (NES) comprised: 140 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM D-glucose, 10 mM HEPES (pH 7.3). ATP-evoked responses were measured at –60 mV.
Whole-cell currents were low-pass filtered at 2 kHz and digitized at 10 kHz. ATP was applied locally using a Picospritzer II (Parker Instrumentation, NJ). To ensure delivery of drug, 0.05% (w/v) Fast Green was used (local applications of 1% Fast Green evoked no response). To visualize HEK 293 cells expressing P2X receptors, cells were co-transfected with vector encoding GFP (0.5 μg of pEGFP-N1 vector included in precipitate) and were observed under a microscope with an epifluorescence attachment (Nikon). Nontransfected cells and cells expressing GFP alone were found to have no inward current in response to application of ATP. Acquisition was performed using HEKA Pulse 8.30, and data were subsequently analyzed using IgorPRO 3.16.
Methylamine (MA, 50 mM) was added to cells in normal culture medium at 37°C for 30 minutes. Cells were then transferred to the patching chamber where they were perfused with NES. ATP (magnesium salt) and BzATP were applied at a concentration of 30 μM in NES (containing normal concentrations of divalent ions). To test the effects of ivermectin, cells were perfused with 3 μM ivermectin for 3 minutes before application of ATP. To block transport through the secretory pathway, cells were incubated with brefeldin A (10 μg/ml) for 1 hour in normal culture medium at 37°C before, and during incubation with, MA.
Statistics and data analysis
Plots were carried out in Igor Pro (Wavemetrics, OR) or Excel (Microsoft, WA). Statistical analyses were performed with a Welch unpaired t test using InStat software (version 2.01; GraphPad Software, CA). All results are given as mean+s.e.m. or s.d.
This work was supported by the Biotechnology and Biological Sciences Research Council. We thank D. Brough for general advice and help with peritoneal macrophage isolation and culture. We thank S. J. Royle, M. Masin, S. J. Ormond and L. K. Bobanovic for experimental assistance.