|Home | About | Journals | Submit | Contact Us | Français|
Transient Receptor Potential Melastatin-like 7 (TRPM7) is a channel protein that also contains a regulatory serine-threonine kinase domain. Here, we find that Trpm7−/− T-cells are deficient in Fas-receptor induced apoptosis; TRPM7 channel activity participates in the apoptotic process and is regulated by caspase-dependent cleavage. This function of TRPM7 is dependent on its function as a channel, but not as a kinase. TRPM7 is cleaved by caspases at D1510, disassociating the carboxy-terminal kinase domain from the pore without disrupting the phosphotransferase activity of the released kinase, but substantially increasing TRPM7 ion channel activity. Furthermore, we show that TRPM7 regulates endocytic compartmentalization of the Fas receptor following receptor stimulation, an important process for apoptotic signaling through Fas receptors. These findings raise the possibility that other members of the TRP channel superfamily are also regulated by caspase-mediated cleavage with wide-ranging implications for cell death and differentiation.
TRPM7, a member of the TRPM subgroup of Transient Receptor Potential (TRP) channels, is a large (1862 a.a., 210 kDa) protein that contains both a cation-conducting pore and a serine-threonine kinase – a unique molecular configuration often referred to as a “chanzyme” for its channel-enzyme bifunctionality (Nadler et al., 2001; Runnels et al., 2001). TRPM7 is nonselective among cations but is permeant to and inhibitable by intracellular Mg2+ (Kozak and Cahalan, 2003; Nadler et al., 2001). Conducting only a few pA of inward current at physiological pH, it is potentiated at low extracellular pH (Jiang et al., 2005) and by phospholipase C-linked receptors in intact cells (Langeslag et al., 2007). Trpm7−/− embryos do not survive past day 7 of embryogenesis (Jin et al., 2008), indicating that TRPM7 has an essential and nonredundant role in mouse development. TRPM7 is expressed in all cell types examined (Kunert-Keil et al., 2006; Ramsey et al., 2006) and organ development is widely disrupted in tissue-specific Trpm7−/− deficient mice (Jin et al., 2011). Similarly, selective deletion of Trpm7 in the T-cell lineage disrupts thymocyte development and accelerates thymic involution (Jin et al., 2008).
TRPM7’s carboxyl-terminal kinase is homologous to a family of atypical serine-threonine kinases called α-kinases and is structurally similar to Protein Kinase A (PKA) (Yamaguchi et al., 2001). TRPM7 autophosphorylates at multiple sites and can phosphorylate Annexin A1 (Dorovkov and Ryazanov, 2004) and myosin IIA heavy chain (Clark et al., 2008), although its natural substrates are not known. The proximity of the kinase domain to the Mg2+-permeating but nonselective pore may be of significance to signal transduction, with TRPM7 mediating a localized cation flux (Ca2+, Na+, Mg2+ and trace ions such as Zn2+) either at the plasma membrane or via specialized intracellular vesicles. The kinase domain, in turn, has been proposed to be involved in mediating the inhibition of TRPM7 channel by serving as a binding site for Mg2+ and Mg2+-bound nucleotides (Demeuse et al., 2006). Alternatively, since the KD for Mg2+ inhibition of TRPM7 is ~0.6 mM (Nadler et al., 2001), near physiological levels of free Mg2+ (0.5 ± 0.2 mM) (Romani and Scarpa, 2000), and ATP is the primary regulator of free Mg2+, TRPM7 may be a metabolic sensor. Environmental acidic pH (Jiang et al., 2005), high intracellular PIP2 (Runnels et al., 2002) and high internal ATP levels (low free Mg2+) should maximize TRPM7 channel activity. A proposed requirement of TRPM7 in vertebrate Mg2+-homeostasis (Ryazanova et al., 2010; Schmitz et al., 2003) is incongruent with the evidence that TRPM7 currents are small at physiological membrane potentials (<10 pA at −60 mV), the recent identification of ubiquitous Mg2+ transporters (Li et al., 2011; Zhou and Clapham, 2009) and our finding that Trpm7−/− lymphocytes and other cell types are not deficient in acute Mg2+ uptake when exposed to extracellular 10 mM Mg2+, or at steady state in total Mg2+ levels (Jin et al., 2008; Jin et al., 2011). Importantly, the Mg2+-permeability of TRPM7 may be of greater significance to intracellular signals that respond to highly localized intracellular Mg2+ around the intracellular mouth of TRPM7 channels.
Using Trpm7fl/−(lck Cre) mice (also referred to as TM7−/− mice) with selectively targeted deletion of Trpm7 in the T-cell lineage, we showed that TRPM7 is required for normal T-cell development and that a large fraction of Trpm7−/− thymocytes arrest their thymic development at the DN3 stage, described by the cell surface immunophenotype CD4-CD8-CD44-CD25+(Jin et al., 2008). The thymocytes that escape this developmental block, presumably by delayed deletion of Trpm7, survive and populate the peripheral lymphoid organs, albeit with a lower number of T-cells. Interestingly, TRPM7 is deleted efficiently in these peripheral T-cells (Jin et al., 2008), providing an opportunity to study TRPM7 in a post-developmental system. We have now investigated these peripheral Trpm7−/− T-cells from TM7−/− mice in search of molecular mechanisms involved in the regulation and function of TRPM7. We show here that regulation of TRPM7 through caspase-mediated cleavage is important for Fas-induced apoptosis.
In the process of studying T-cell activation, we observed that in contrast to wt T-cells, Trpm7−/− T-cells do not show a substantial decrease in viability after T-cell receptor (TCR) restimulation. Based on these observations, we investigated the phenomenon of activation-induced cell death (AICD) – also referred to as restimulation-induced cell death (RICD), which results in the apoptosis of T-cells upon repeated TCR stimulation (Green et al., 2003; Krammer et al., 2007). When actively growing wt T-cells were re-stimulated with anti-CD3 antibodies, the percentage of Annexin A5 positive (ANXA5+) cells increased from 15% to 35% after 15h of re-stimulation (Figure 1A and S1A). In contrast, AICD in Trpm7−/− T-cells was significantly blunted (Figure 1A and S1A). The difference was also seen when AICD was measured after 30h of restimulation (Figure 1A and S1A). Accordingly, Trpm7−/− T-cells are also deficient in activation of caspases as measured by FAM-VAD-FMK, a fluorescent cell-permeable probe of caspase activity (Figure 1B). Since AICD is dependent on death-receptor signaling (Krammer et al., 2007), we tested the sensitivity to Fas-induced apoptosis in Trpm7−/− T-cells.
Actively growing T-cells were stimulated with anti-Fas agonist antibody and crosslinking agent, Protein G, and the resulting apoptosis measured by ANXA5-staining (Figure 1C). Flow cytometric histograms representing the increase in the percentage of ANXA5+ cells in response to increasing concentrations of anti-Fas agonist antibody are shown as overlays of wt and Trpm7−/− (KO) T-cells (Figure S1B). These results indicate that Trpm7−/− T-cells are significantly less sensitive to Fas-stimulated apoptosis throughout the effective range of the agonist antibody concentration. We also evaluated wt and mutant T cell sensitivity to plate-bound anti-Fas agonist antibody and confirmed the relative resistance of Trpm7−/− T-cells to Fas-induced apoptosis (Figure 1D). These findings are consistent with cell viability assays in response to Fas stimulation (Figure S1C). The cell surface levels of Fas receptor on activated Trpm7−/− T-cells were equal to wt T-cells (Figure 1E).
Next, we evaluated Fas-induced cell death in the human Jurkat T-cell leukemia cell line using an RNAi approach. To deplete TRPM7 levels in Jurkat cells, we generated a lentivirus that expresses a short hairpin RNA directed against the TRPM7 mRNA (LV-M7shRNA) and also marks the transduced cell through bicistronically expressed GFP. The resulting depletion of TRPM7 mRNA was evident by quantitative real-time PCR (qPCR) (Figure 2A). Patch-clamp recordings of GFP+ cells confirmed an ~10 fold reduction in the whole-cell TRPM7 current, ITRPM7 (Figures 2B, 2C and 2D). We measured apoptosis in Jurkat cells by immunoblotting the lysates for cleaved Poly-ADP ribose polymerase (PARP), a terminal substrate of executioner caspases. Jurkat cells transduced with LV-M7shRNA were resistant to Fas receptor dependent cell death (Figure 2E) but not to Etoposide-induced cell death (Figure 2F), which relies on a different signaling mechanism for initiating apoptosis (Lowe et al., 1993). The sensitivity of Fas-induced apoptosis in Jurkat T-cells to depletion of TRPM7 supports the view that the defective Fas receptor signaling in Trpm7−/− T-cells is not a result of a developmental defect. These findings strongly suggest that TRPM7 participates in Fas receptor signaling.
For mechanistic insight, we set out to distinguish the relative contributions of the TRPM7 pore from its kinase activity in Fas-receptor signaling. Since we failed to express TRPM7 variants with substantial efficiency in Jurkat cells, we screened for Fas-sensitive cell lines that enable the ectopic expression of TRPM7, a protein that is difficult to express transiently in most cell lines. LN18, an adherent human glioblastoma cell line, can express full-length variants of TRPM7 at an efficiency of 5–10%. In these cells, Fas-receptor signaling can be evaluated through immunofluorescent detection of the cleaved caspase-3 (active caspase-3), 3h after Fas-stimulation (Figure S2E). Additional characterization of LN18 cells showed that these cells belong to the type 2 category of Fas-sensitive cells – reliant on the mitochondrial amplification loop for the efficient completion of apoptosis. LN18 cells release cytochrome c in response to Fas-activation and ectopic expression of Bcl-2 reduces their sensitivity to Fas-induced apoptosis as detected by the percentage of cells with active caspase 3 (Figure S2A). Next, we confirmed that LN18 cell Fas-sensitivity is affected by TRPM7 function. Transfection of LN18 cells with siRNA duplexes directed against the TRPM7 mRNA (M7siR#3) results in a substantial depletion of TRPM7 mRNA as measured by qPCR (Figure S2B). Testing these cell populations for Fas-induced apoptosis showed that depletion of TRPM7 in LN18 cells leads to reduced sensitivity to Fas-induced apoptosis as detected by the percentage of cells with active caspase 3 (Figures S2C).
Next we evaluated the Fas-sensitivity of LN18 cells. TRPM7 variants were expressed through a previously described plasmid, Red-pTracer-CAG (Brauchi et al., 2008), that bicistronically encodes monomeric RFP in transfected cells (Figure S2D). We confirmed that RFP+ cells also expressed the encoded TRPM7 variant (Figure S2D). When compared to LN18 cells transfected with vector alone, expression of wild type TRPM7 (M7-wt) augments the activation of caspase-3 as measured 3h after Fas-receptor stimulation (Figures 2G and S2E). In contrast, the ectopic expression of a TRPM7 variant with a mutant (nonconducting) pore (M7-pm) (Krapivinsky et al., 2006), does not affect caspase-3 activation. A TRPM7 variant with the amino acid substitution, K1646A, renders the kinase inactive (Matsushita et al., 2005), but it augmented Fas-receptor signaling in a manner comparable to wt TRPM7 (Figure 2G). Similar results were obtained in 293T cells, which provide an avenue for high transfection efficiency. Since 293T cells express very low levels of Fas and are insensitive to Fas-stimulation, the assay was carried out by sequentially expressing the Fas-receptor and the TRPM7 variants. In this experiment, the Fas-induced apoptosis was measured by quantitative immunoblotting of cleaved PARP and α-tubulin (Figure 2H and S2G). The cleavage of PARP has been quantified by deriving the densitometric ratio of cleaved-PARP and αTubulin detected in the same immunoblot. These results show that ectopic expression of TRPM7-wt and TRPM7-KA, but not TRPM7-pm, potentiates Fas-induced cell death (Figure 2H). The expression of TRPM7 variants and Fas in these cells is shown in Figure S2H. Ectopic expression of M7-wt or M7-pm did not affect the expression or localization of coexpressed Fas (Figure S2H and data not shown). Interestingly, ectopic expression of TRPM7 in LN18 cells also potentiates TRAIL-induced apoptosis but does not have a substantial effect on apoptosis induced by exposure to UV light, etoposide and staurosporine (Figure S2F). Overall, these results indicate that the role of TRPM7 in extrinsic apoptosis is dependent on the channel activity of TRPM7 and does not depend on its kinase activity.
Although TRPM7 current (ITRPM7) is readily detectable in most cells, the TRPM7 protein is expressed at very low levels and is biochemically detectable by immunoblotting only after enrichment through immunoprecipitation. Compared to unstimulated LN18 cells, the Fas-stimulated LN18 cells contain a 40–42-kDa TRPM7 fragment that is detectable by immunoblotting with an antibody raised against the carboxyl-terminal kinase domain of TRPM7 (Figure 3A). The emergence of this fragment is evident 1h after Fas-ligation (Figures 3A and S3) and peaks 3h after Fas-ligation. To rule out artifacts due to antibody cross-reactivity, we confirmed the cleavage using a different TRPM7 antibody for immunoprecipitation of endogenous TRPM7 (Figure 3B). When LN18 cells were treated with a pan-caspase inhibitor, ZVAD-fmk, the cleavage was abolished, indicating a dependence on caspase activity (Figure 3C). The anti-TRPM7 antibodies used in these immunoblots are reactive to the extreme carboxyl-terminal amino acid sequence of TRPM7, indicating that the 42-kDa fragment of TRPM7 retains the carboxyl-terminus kinase domain of TRPM7. We also generated a stable cell line of LN18 expressing a TRPM7 variant that was tagged at the N-terminus with streptavidin-binding peptide (SBP). After Fas-induced TRPM7 cleavage, we pulled down the N-terminal fragment by using streptavidin beads and immunoblotted TRPM7 using a rabbit polyclonal antibody that was generated against the TRPM7 amino acid sequence 1277–1380 (N-terminal to the cleavage site D1510). This analysis shows the generation of an ~160 kDa N-terminal fragment of TRPM7 in Fas-stimulated cells (Figure 3D).
Since TRPM7 cleavage was detectable 1h after Fas-stimulation, we reasoned that the cleavage was mediated by caspase-8, the apical or initiator caspase during Fas-induced apoptosis. Inhibitors of caspase-8 however are not useful in illuminating this aspect because inhibition of the initiator caspase prevents the activation of all subsequent caspases. To gain insight, we took advantage a cell-free system that allows the activation of caspases in vitro through the addition of cytochrome c and dATP to cytosolic extracts (Li et al., 1997; Liu et al., 1996). Although this type of activation is initiated through the activation of apoptosome (Apaf-1-cytochrome c-caspase-9 complex), it leads to a comprehensive activation of caspases (including caspase-8) through a physiologically relevant, albeit in vitro, process that does not involve Fas signaling (McStay et al., 2008). We investigated the effect of immunodepletion of caspase-8 and caspase-3 on the in vitro cleavage of TRPM7 by such cytosolic preparations. When caspases in the cytosolic S100 preparations from Jurkat cells were activated by the addition of cytochrome c and dATP, the activated (cleaved forms) of caspase-8 and caspase-3 were readily detected (Figure 3E). In contrast, immunodepletion of caspase-3 and caspase-8 prior to activation by cytochrome c and dATP resulted in near complete loss of these caspases in the activated preparations (Figure 3E). We then used these preparations to cleave immunopurified TRPM7 in vitro. As shown in Figure 3E (lower panel), cleavage of TRPM7 is reduced by immunodepletion of caspase-8. However, we also find a similar reduction upon immunodepletion of caspase-3, indicating that caspase-8 contributes to TRPM7 cleavage but is not the exclusive mediator of TRPM7 cleavage. In the experiments discussed below, the in vitro cleavage of TRPM7 by caspase-8 is also demonstrated using purified caspase-8 (Figure 3F).
Based on the molecular size of the TRPM7 carboxy-terminal cleavage fragment (~40 kDa), we reasoned that the TRPM7 cleavage site was approximately 350 amino acids upstream of the terminal amino acid and identified putative sequences that could serve as caspase substrates. The residue D1510 (mouse TRPM7 sequence) is flanked by a sequence that is congruent with the experimentally etermined substrate selectivity of caspase-8 (Mahrus et al., 2008; Pop and Salvesen, 2009). In an in vitro assay of TRPM7 cleavage, using an 35S-labeled 58 kDa C-terminal fragment of TRPM7 as the substrate, we confirmed that the caspase 8-mediated cleavage generates a 40–42-kDa C-terminal fragment (Figure 3F). Substitution of D1510 to an alanine (D1510A) abrogated the cleavage, suggesting that the cleavage occurred at D1510 (Figure 3F). Since the adjoining residue, S1511, is known to be autophosphorylated by TRPM7 kinase (Matsushita et al., 2005), we evaluated the effect of a phosphomimetic substitution, S1511E, and its converse substitution, S1511A. The S1511E substitution abolished TRPM7 cleavage but S1511A was cleaved efficiently (Figure 3F). These results indicate that TRPM7 is cleaved early during Fas-receptor stimulation at D1510, resulting in the dissociation of TRPM7’s kinase domain from its channel domain. The results from the S1511 substitutions suggest a possible role for phosphorylation in the regulation of caspase-dependent TRPM7 cleavage – with the important caveat that caspase-8 strongly prefers small uncharged residues in this position (P1’) (Pop and Salvesen, 2009). The cleavage of TRPM7 at D1510 is also demonstrated through the analysis of ectopically expressed TRPM7 variants containing a carboxy-terminal T7 immunoepitope tag (Figure 3G). In contrast to wt TRPM7, the TRPM7 variant with a D1510A mutation is not cleaved in response to Fas-stimulation of LN18 cells (Figure 3G). These results indicate that Fas-receptor signaling results in the cleavage of TRPM7.
Taken together, these data lead us to conclude that in response to Fas-signaling, TRPM7 undergoes a caspase-mediated cleavage at D1510 to generate two distinct proteins, the TRPM7 channel and the TRPM7 kinase. Based on the observation that the cleavage is detectable in the early phase of apoptosis (1h after Fas stimulation), the immunodepletion experiments and in vitro cleavage of TRPM7 by caspase-8, we propose that the cleavage is initiated by caspase-8 but sustained by downstream caspases during the entire process of apoptosis. In contrast, caspase-8 substrates like Bid (Li et al., 1998) and RIPK3 (Kaiser et al., 2011; Oberst et al., 2011) are thought to be cleaved selectively by caspase-8 during extrinsic apoptosis. The caspase-dependent cleavage of TRPM7 is illustrated in Figure 3H.
The cleavage of TRPM7 at D1510 releases the 42–kDa polypeptide containing the kinase domain (M7kin42) from the membrane-resident channel. As discussed earlier, the participation of TRPM7 in Fas-induced apoptosis is dependent on the channel activity and not the kinase activity. In accord with this, we show that although the released 42-kDa fragment retains kinase activity, it does not modulate Fas-induced apoptosis. We tested whether the released kinase is capable of phosphotransferase activity by conducting in vitro kinase assays with ectopically expressed FLAG-tagged M7kin42. Autophosphorylation activity of M7kin42 in the presence of 33P-labelled ATP was evaluated by autoradiography and M7kin42 was found to be active (Figure 4A). In contrast, M7kin42GV, a variant of the M7 kinase domain with a targeted substitution (G1619V) in the ATP-binding pocket, showed minimal autophosphorylation, indicating that the phosphorylation was not a result of an endogenous kinase in the immunoprecipitate (Figure 4A). Both variants of M7kin42; wt and GV are expressed at comparable levels (Figure 4A and S4A). LN18 cells transfected with M7kin42-wt or M7kin42-GV variants showed no appreciable difference in the Fas-induced activation of caspase-3 (Figures 4B, S4B). Next, we tested whether the cleavage of TRPM7 is necessary for its participation in Fas-induced apoptosis.
To evaluate the role of cleavage in extrinsic apoptosis, we investigated the ability of TRPM7 to potentiate Fas-induced apoptosis when the critical residue D1510 is mutated (TRPM7-D1510A or simply M7DA). We also evaluated the ability of the cleaved TRPM7 channel domain in this assay. Since the cleaved TRPM7 channel lacks the kinase domain, the TRPM7 variant truncated at D1510 is referred to as TRPM7-ΔK or M7-ΔK. We expressed the empty vector, M7-wt, M7-DA (TRPM7-D1510A mutant), M7ΔK (TRPM7 truncated at D1510) and M7ΔK-pm (nonconducting pore mutation in truncated TRPM7) in LN18 cells. Caspase cleavage-resistant M7D1510A mutant did not potentiate Fas-induced apoptosis, indicating that cleavage was necessary for this effect (Figure 4C and S4C). The cleaved channel greatly potentiated caspase-3 activation when the pore was functional (M7-ΔK) but not when the pore was mutated (M7-ΔK-pm) (Figure 4C and S4C). Similar results were obtained in 293T cells coexpressing the M7 variants and Fas, as measured by PARP cleavage (Figure 4D). Expression of M7ΔK but not M7D1510A or M7ΔK-pm increased the Fas-sensitivity of 293T cells coexpressing the Fas-receptor. Representative Immunoblots of cleaved PARP and α-tubulin are shown in Figure S4D. The expression of TRPM7 variants and the Fas-receptor is shown in Figure S4E. These results show that the critical TRPM7 function in Fas-receptor signaling is dependent on cleavage-mediated modulation of TRPM7 channel activity, but is not due to cellular Ca2+ overload. Although TRPM7 is a cationic channel permeant to Ca2+, the inward conductance of TRPM7 at physiological membrane potential does not lead to a substantial influx of Ca2+ in cells expressing ectopic TRPM7 (Figure 4E). The precise mechanism by which TRPM7 modulates TRPM7 is not clear at this point but is likely to involve dynamic changes in subcellular microdomains of Ca2+, Mg2+ or Zn2+ during death receptor signaling.
The cleavage of TRPM7 at D1510 dissociates the 40-kDa kinase domain from the 172-kDa ion-conducting pore (M7-ΔK). To evaluate the effect of cleavage on TRPM7 channel conductance, we expressed TRPM7 variants in CHO cells. These cells have very low endogenous TRPM7 currents and allow electrophysiological resolution of ITRPM7 elicited by ectopically expressed TRPM7 variants without excessive overexpression. ITRPM7 characteristically develops over time after membrane rupture, a process commonly referred to as “break-in” during patch clamp recording. All functional TRPM7 variants exhibited similar current-voltage relationships and were expressed at comparable levels in CHO cells (Figure S5A). We analyzed the mean current density in transfected CHO cells at +100 mV (outward current, Figure 5A, B) and at -100 mV (inward current, Figure 5C, D). In cells expressing M7-wt, the current increases to a mean current density of 250 pA/pF after 50s and to 600 pA/pF after 500s, while M7-ΔK currents develop faster and exhibit mean current densities that are significantly higher (720 pA/pF after 50s and 925 pA/pF after 500s). In contrast, cells transfected with M7-pm lack substantial current after break-in (Figure 5A, B, blue). TRPM7-D1510A (M7-DA) current densities are somewhat lower, but not significantly different than wt. In agreement with a previous report (Schmitz et al., 2003), point mutations that inactivated the kinase did not affect TRPM7 channel activity (Figure S5B and S5C). Overall, these data indicate that after the cleavage of TRPM7 by caspase-8, higher currents are achieved when compared to full length, uncleaved TRPM7. We also confirmed that ITRPM7 is expressed in LN18 cells. A Mg2+-inhibitable current develops in a time-dependent manner (Figure S5D) and shows the signature current-voltage relationship of ITRPM7. Furthermore, we show that this current shows pharmacological sensitivity identical to that of TRPM7. It is inhibited by 0.5 mM 2APB but fully reactivated by 2 mM 2APB (Figure S5E). ITRPM7 in LN18 cells is also activated by 10 mM NH4Cl (data not shown).
Next, we evaluated the effect of reconstituting TRPM7 variants in Trpm7−/− T-cells. In comparison to wt T-cells, the vector-transfected Trpm7−/− T-cells were less than half as sensitive to FAS stimulation (Fig. 5E, 5F and Figure S5D). Surprisingly, the Trpm7−/− T-cells transfected with TRPM7-wt and TRPM7-ΔK exhibited spontaneous (without FAS-stimulation) apoptosis. This effect was not seen in cells expressing TRPM7-pm and TRPM7-DA, suggesting that Trpm7−/− T-cells are capable of apoptosis but are restrained due to a lack of TRPM7 channel activity. Overexpression of competent TRPM7 variants likely leads to apoptosis due to the presence of active caspases in proliferating T cells (Kennedy et al., 1999; Leverrier et al., 2010; Siegel, 2006). Since TRPM7-pm and TRPM7-DA do not result in apoptosis, the results support the conclusion that the cleavage of TRPM7 at D1510 and the consequent channel activity plays an important role in FAS-induced apoptosis.
For mechanistic insight and to pinpoint the process that is regulated by TRPM7 during Fas-induced apoptosis, we first determined whether TRPM7 acts upstream or downstream of cytochrome-c release from mitochondria in LN18 cells. We generated a stable cell line, LN18 (CytC-GFP), expressing cytochrome c-GFP in the LN18 background. In this cell line, cytochrome c-GFP localizes predictably to the mitochondria and is released from the mitochondria in response to Fas-stimulation (Supplemental movie 1). When these cells are transfected with siRNA targeting TRPM7, the release of cytochrome c-GFP from the mitochondria is abrogated (Figure 6A and 6B). After 180 minutes of Fas-stimulation, 37% of the cells transfected with control siRNA released cytochrome c. In contrast only 5% of the cells transfected with TRPM7 siRNA released cytochrome c-GFP (Figure 6B). A similar difference was seen at 240 minutes of Fas-stimulation (Figure 6B). These data indicates that TRPM7 acts upstream of cytochrome c release in type 2 cells. We also evaluated the effect of knocking down TRPM7 in SKW6.4 cells, a prototypical type 1 Fas-sensitive cell line that does not depend on the mitochondrial amplification loop to execute Fas-induced apoptosis (Schmitz et al., 1999). The reduction in TRPM7 levels upon transfection of TRPM7-targeting siRNA is shown through immunoprecipitation and immunoblotting of TRPM7 (Figure S6A). SKW6.4 cells with reduced TRPM7 levels show significantly reduced sensitivity to Fas-induced apoptosis (Figure S6B).
We analyzed the DISC complex in SKW6.4 cells with knocked down TRPM7 in consideration of the hypothesis that TRPM7 regulates the assembly of the DISC complex upon Fas stimulation. This was done by immunoprecipitating the Fas receptor from lysates generated at specific time points after Fas-stimulation and immunoblotting for the two key components of DISC – Procaspase-8 and FADD. These results, shown in Figure 6C, do not support the hypothesis that TRPM7 regulates the assembly of DISC. After 10 minutes of Fas stimulation, both cell populations (control siRNA and TRPM7 siRNA-treated) show recruitment of FADD and procaspase-8. The composition of DISC at 60 minutes shows modest differences in the levels of procaspase-8 but our results did not allow us to conclude that this difference was statistically significant. Interestingly, the activity of caspase-8 at the same time points shows modest but significant differences (Figure 6D). These data indicate that although TRPM7 does not regulate the assembly of DISC, it influences the processing of caspase-8 in the subsequent steps. Fas receptor transmits apoptotic as well as survival signals (Peter et al., 2007). The process by which the cellular context dictates whether Fas signaling results in apoptosis or non-apoptotic signals is not well understood, but internalization and compartmentalization of the DISC complex has been proposed as a critical event driving the transmission of apoptotic signal (Schutze et al., 2008). In SKW6.4 cells transfected with TRPM7 siRNA, we observed that decreased sensitivity to apoptosis (Figure 6E and S6B) was also accompanied by activation of the MAPK pathway – as detected by the presence of phosphorylated Erk1/2 in cell lysates (Figure 6E). We then considered the possibility that SKW6.4 cells with reduced TRPM7 were compromised in the process of Fas-receptor endocytosis after Fas-stimulation – favoring the transmission of non-apoptotic signals downstream of Fas.
Internalization of the Fas receptor in SKW6.4 cells following receptor stimulation has been described previously (Algeciras-Schimnich et al., 2002; Siegel et al., 2004). We used an anti-Fas agonist antibody conjugated to a fluorophore (Alexa 568) to stimulate the Fas receptor and observe its subsequent internalization. The staining of the Fas receptor (on the cell surface) with this antibody is detectable after 15 min of incubation in cell culture conditions. As shown in Figure S6C, internalization follows this receptor engagement and 75% of the SKW6.4 cells can be seen to have internalized the Fas receptor after 1h of stimulation (Figure 6F and 6G). In contrast, the SKW6.4 cells that are transfected with TRPM7 siRNA do not exhibit Fas internalization at a comparable level; only 30% of the cells show Fas internalization (Figure 6F and 6G). We carried out the same experiment with LN18 cells. In the case of LN18 cells, the Fas receptor levels are lower and their distribution in the plasma membrane was strikingly polarized. Nevertheless, internalization is readily detectable in LN18 cells transfected with control siRNA. Again, the cells transfected with TRPM7 siRNA shows considerably reduced incidence of Fas receptor internalization after 1h of stimulation (Figure 6H and 6I). A time-course of internalization in the LN18 cells is shown in Figure S6D as a montage of images acquired between 15min and 1h of Fas stimulation. The punctate pattern of internalized Fas is not due to redistribution of the Fas receptor within the plasma membrane – the punctae are indeed inside the cells. This is demonstrated through the acquisition of confocal images along the vertical axis (Z-stack of images) of cells transfected with control siRNA and TRPM7 siRNA. These images are shown as a combined movie (Supplemental movie 2). Overall, these results clearly show that TRPM7 participates in regulating the internalization of Fas receptor.
TM7−/− mice show a pathophysiological similarity to the mice in which the Fas receptor is selectively deleted in T-cells (TFas−/− mice). In contrast to the phenotype of global deletion of Fas, which results in lymphoproliferative, autoimmune pathology in young animals (Watanabe-Fukunaga et al., 1992), the selective deletion of Fas in T-cells leads to lymphopenia and a high incidence of pulmonary inflammation and lung-tissue damage upon aging (Hao et al., 2004). Similarly, the lungs of 13–15 month-old TM7−/− mice exhibit pulmonary inflammation and emphysema. Hematoxylin-Eosin (HE) staining of lung tissues revealed alveolar damage and significant enlargement of alveolar airspaces, similar to the histopathology of emphysema (Figure 7A). We quantified the enlargement of alveolar airspaces through morphometric analysis (Jacob et al., 2009) of HE-stained lung slices of wt and TM7−/− as briefly illustrated in Figure S7. A comparison of Mean Linear Intercept (MLI) and the statistical moments (D0, D1 and D2) of the equivalent diameter deq between the airspaces of wt and TM7−/− mice lung are shown as statistical box charts in Figure 7B. Antibody staining using anti-CD3, (Figure 7C, left panel), B220 (Figure 7C, right panel) and Ly6G (Figure 7D) document the increased infiltration of T-cells, B-cells and neutrophils into the lungs of aged TM7−/− mice (Figure 7E). We did not observe a significant difference between the lungs of young mice, suggesting that pulmonary inflammation and emphysema seen in the TM7−/− mice is age-dependent. This is remarkably similar to the phenotype of mice in which Fas was selectively deleted in T-cells. The selective deletion of Fas in T-cells leads to lymphopenia without obvious pathology in young mice, but a high incidence of pulmonary inflammation and lung-tissue damage upon aging (Hao et al., 2004). This phenotype is distinct from global deletion of Fas, which results in an autoimmune pathology that is evident in young animals (Watanabe-Fukunaga et al., 1992).
We demonstrated that TRPM7 is cleaved by caspase-8 at D1510, dissociating the kinase from the ion-conducting pore. The cleaved channel exhibits substantially higher ITRPM7 and potentiates Fas-receptor signaling. In contrast, neither the cleaved kinase domain nor the cleavage-resistant TRPM7 D1510A mutant potentiates this pathway. Cleavage-induced current enhancement is specific to D1510, as TRPM7 truncated at other sites within the kinase domain does not elicit an increase in currents (Matsushita et al., 2005) (Schmitz et al., 2003). The cleavage-induced augmentation of TRPM7 channel activity is not due to inactivation of its kinase. Cleavage elicits two distinct functional proteins (channel and kinase) that can now differentially localize. These data indicate that TRPM7 cleavage is an important regulatory step for TRPM7 function in Fas receptor signaling. In cells where TRPM7 levels have been reduced substantially through siRNA treatment, the Fas receptor is not internalized normally in response to receptor stimulation. In these cells, the reduction in their sensitivity to Fas-induced apoptosis correlates with an increase in activation of MAPK pathway, a nonapoptotic signaling route for Fas signaling. Lastly, in accord with the cellular phenotype, the TM7−/− mice show a distinct pathophysiological similarity to the mice in which Fas is selectively deleted in T-cells (TFas−/− mice).
Since many TRP channels have emerged as regulators of cell death (Aarts et al., 2003; Hara et al., 2002), inflammation (Barbet et al., 2008; Link et al., 2010; Sumoza-Toledo and Penner, 2011; Yamamoto et al., 2008) and pain (Basbaum et al., 2009), this study raises the possibility that in these roles, TRP channels are regulated through caspase-mediated cleavage events. The findings thus offer an avenue of research to guide pharmacological intervention of multiple diseases.
Reagents and detailed methods are in the Supplemental Experimental Procedures (SEP).
Mice were bred and maintained according to guidelines and procedures approved by the Institutional Animal Care and Use Committee (IUCAC).
HEK-293, LN18, Jurkat (clone E6-1), CHO and SKW6.4 cell lines were obtained from American Type Culture Collection (ATCC). Fas-receptor stimulation for indicated times was with anti-human Fas antibody (clone CH11). Primary mouse T-cells were activated using bead-bound anti-CD3 and anti-CD28 in the presence of 50U/ml recombinant mouse IL-2 and maintained at a concentration of 0.5–0.75 million cells/ml in RPMI growth medium supplemented with 10% FBS, 1 mM Glutamine and 0.001% 2-mercaptoethanol. Fas-receptors were activated by anti-mouse Fas (CD95) Jo2 clone in 0.5 µg/ml Protein G. Alternatively, apoptosis was induced by culturing T-cells with plate-bound (10 µg/ml) anti-Fas antibody (see Supplemental Experimental Procedures).
LN18 cells were lysed in ice-cold lysis buffer (20mM Tris-HCl, 150mM NaCl, 1.0% Triton-X-100, pH 8.0) containing protease and phosphatase inhibitor cocktail (Pierce). Immunoprecipitations were carried out using the C47 antibody (2µg/ml) or N74/25 antibody (50μl hybridoma supernatant/ml) at 4°C for 12–16h. TRPM7 was detected by immunoblotting with C-47 antibody (2 µg/ml). For in vitro cleavage, the 35S-labelled TRPM7 C-terminal polypeptide sequences (amino acids 1364-1863) were synthesized in vitro using the TNT-T7 (Promega). 2µl of reaction mixture containing the 35S-labelled polypeptide was incubated with 2U (2µl) of purified caspase-8 in 15 µL of AMC buffer (20 mM PIPES [pH 7.2], 100mM NaCl, 1mM EDTA, 0.1% CHAPS, 10% sucrose, 10mM DTT, 1mg/ml Pefablock SC) at 37°C for 2.5h prior to analysis by electrophoresis and autoradiography. See SEP for additional methods pertaining to biochemical analysis of TRPM7 cleavage.
SKW6.4 cells were stimulated with 2µg/ml of anti-Fas antibody (clone APO-1-1, Kamiya Biomedical) at 37°C and lysed at indicated time points. The lysis buffer (30 mM Tris-HCl, pH 7.4, 150mM NaCl, 1% Triton-X-100 and 10% glycerol) was supplemented with protease inhibitor cocktail (Roche) immediately before use. For the untreated or 0 min condition, antibody was added after lysis of unstimulated cells. The DISC (Fas immunoprecipitate) was isolated by incubating the lysates with protein A sepharose for 2h at 4°C and washed with lysis buffer prior to separation by SDS-PAGE electrophoresis and immunoblotting. For caspase-8 assays, the lysis buffer was supplemented with 0.5µM MG-132 protease inhibitor. Caspase-8 was immunoprecipitated from lysates and washed thoroughly in the lysis buffer prior to activity assays. The IP pellets were resuspended in caspase-Glo 8 reagent (Promega). The cleavage reactions were incubated for 30min at RT and relative luminescence measured. siRNA cocktail was composed of 5nM M7siR#2 and 5nM M7siR#3 (see SEP).
Cells plated in glass-bottomed dishes were cultured in synthetic serum-free medium (Opti-MEM, Invitrogen) for 6h prior to internalization assays. Cells were incubated with 2µg/ml Alexa568-conjugated anti-Fas antibody (CH11 clone, Millipore) for 15min in cell culture conditions (antibody binding phase). The cells were washed in Opti-MEM (37°C) and images acquired using an Olympus FV1000 confocal microscope (maintained at 37°C during imaging). siRNA cocktail was composed of 5nM M7siR#2 and 5nM M7siR#3 (see SEP).
In case of CHO cells, standard bath solution contained (in mM): 135 Namethanesulfonate (Na-MeSO3), 5 CsCl, 1 CaCl2, 10 HEPES; pH 7.4 with NaOH. The standard pipette solution was (in mM): 120 Cs-MeSO3, 5 CsCl, 10 BAPTA, 3.1 CaCl2, 3 ATP-Na, 0.6 GTP-Na, 10 HEPES; pH 7.2 with CsOH. Signals were sampled at 10Khz and low-pass filtered at 2 kHz. For Jurkat cells and LN18 cells, the modified methods are described in the Supplementary Experimental Procedures (SEP). In shRNA-treated Jurkat cells, the TRPM7-dependent component of outward current (IMIC) was defined by superfusion of 5mM extracellular Mg2+ to block IMIC. Currents were measured at +100mV following voltage ramps from -100mV; holding potential = 0mV.
Immediately after euthanizing 13–15 month-old mice, the lungs were inflated and fixed by intratracheal instillation of freshly prepared 5% buffered PFA, excised and immersed in the same fixative for 24h. The fixed tissue samples were embedded in paraffin, sectioned into 4–6 μm slices, and hematoxylin-eosin (HE) stained. IHC experiments were in formalin-fixed, paraffin-embedded tissue sections (See SEP). See Figure S7 and EEP for lung morphometric measurements.
Statistics pertaining to lung pathology are shown as Box (± SEM), vertical bars (± SD) and data overlap. Statistics pertaining to TRPM7 currents are shown as Box (± SEM), vertical bars (5–95 percentile) and data overlap. All individual data points are shown and the mean value denoted by an empty square and the median shown as a horizontal line. P-values were calculated using Student’s t-test and Analysis of Variance (Anova).
We thank Jie Jin for assistance regarding Trpm7-targeted mice, Andrew Scharenberg (University of Washington) for the 293T cell line expressing FLAG-tagged TRPM7, Douglas Green (St. Jude Children’s Research Hospital) for the cytochrome c-GFP plasmid, Grigoriy Losiyev (Brigham and Women’s Hospital; BWH) for flow cytometry, the BWH Histopathology core, and Clapham lab members for helpful discussions.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.