Introduction

Activation of ion channels is an essential step during regulated cell death [1]. Ion channels participate in execution of apoptosis and have also been assumed to be critical for necroptosis and other forms of caspase-independent cell death. Ion channels may shrink cells, thereby moving them towards apoptosis [2]. To induce apoptotic cell shrinkage, K+ and Cl− ions or osmotically active organic molecules are released from the cell [2]. Numerous K+ channels are known to contribute to this process, and recent studies demonstrate the molecular identity of the underlying anion permeable channels as members of the LRRC8 family of proteins [3–5].

In contrast to apoptotic cell shrinkage, necroptosis is characterized by cell swelling and disintegration of the cell membrane [1, 6]. Membrane permeabilization is thought to be caused by nonselective ion channels that allow influx of NaCl and water leading to cell swelling. Apart from other theories, a plasma membrane-localized MLKL complex has been proposed to act as a cation influx channel or, alternatively, to activate other proteins to increase Na+ influx and osmotic pressure, and to induce membrane rupture [7]. Because this concept is controversial [1], and knowledge about Ca2+ signaling and membrane currents in necroptosis is limited, we decided to analyze these parameters in more detail.

TMEM16F/anoctamin 6/ANO6 is a membrane-localized phospholipid scramblase and ion channel that has been implicated in apoptotic cell death induced by a large and steady rise in intracellular Ca2+ [8]. Activation of ANO6 was shown to generate membrane Cl− currents that can induce apoptotic cell shrinkage (AVD; apoptotic volume decrease). However, cell swelling and cellular disintegration are observed upon pronounced increase in intracellular Ca2+, which causes non-selectivity of ANO6 currents [8–10]. As the molecular nature of the necroptotic membrane channel remains obscure, we asked whether membrane permeabilization is due to the activation of ANO6. Although ANO6-typical membrane currents were activated during necroptosis, our present data indicate that the activation of ANO6 is not crucial for necroptotic cell death.

Results

Necroptosis was induced in NIH3T3 cells using TNFα, smac mimetic birinapant, and the pan-caspase inhibitor Z-VAD (TSZ) [11] (Fig. 1a, b). About 90 % of the cells were necroptotic after 4-h treatment with TSZ. Necroptotic cell death was completely inhibited by simultaneous treatment with necrostatin-1 (Nec-1) (Fig. 1b, c). Moreover, cell death was also induced by TSZ in L929 (murine skin fibroblasts), HT29 (human colonic epithelial cells), 16HBE (human bronchial epithelial cells), and H441 (human airway club cells) (Supplementary Fig. 1). Necroptosis was completely suppressed by 100 µM Necrostatin-1 (Nec-1) in all cell lines. In human epithelial cells, necroptotic cell death was also partially suppressed by 5 µM necrosulfonamide (NSA).

Fig. 1
figure 1

Induction of necroptosis in NIH3T3 cells. Necroptosis in NIH3T3 cells was induced by the necroptotic cocktail TSZ containing TNF (10 ng/ml), smac mimetic birinapant (5 µM), and the pan-caspase inhibitor Z-VAD (25 µM). a, b Flow cytometry indicating time-dependent membrane permeabilization (7-AAD positivity) and phosphatidylserine exposure (annexin V positivity). c Inhibition of necroptosis (AnnV/7-AAD positivity) by necrostatin 1 (Nec-1; 100 µM). Mean ± SEM (number of cells). #Significant increase in AnnV/7-AAD positivity (p = 0.002, 2 h; 0.004, 3 h; 0.00001, 4 h) and inhibition by Nec-1 (p = 0.00001) (ANOVA)

Full size image

It is commonly accepted that cells swell during necroptosis, due to influx of NaCl and water. This ultimately leads to membrane disintegration and cell death [6]. As proper cell volume measurements are still missing, we decided to analyze cell volume and cell morphology using quantitative phase microscopy (QPM). QPM is using interference holographic images of unstained cells allowing for accurate and seamless assessment of the cell volume with very little time delay [12–14]. 3D images taken over four hours of TSZ treatment indicate that cells initially swell within the first hour, but then start to shrink (Fig. 2a, b). Cell shrinkage was inhibited by Nec-1 (Fig. 2b). Morphological and volume changes were also assessed using differential interference contrast (DIC) (Fig. 2c). The cell volumes were determined from individual cells using DIC images and cell height determined from bottom/top focal planes, using the software Axiovision (Zeiss, Munich). These data also indicate an initial cell swelling within the first hour of TZS application and subsequent cell shrinkage (Fig. 2d). To further confirm these volume changes, we used flow cytometry. 30,000 cells were exposed to TSZ (or kept in control solution) and analyzed in a CASY flow cytometer. The assays were repeated 3–9 times. The data again indicate an initial swelling within the first 2 h which then changes into cell shrinkage (Fig. 2e). Finally, we used online membrane capacitance measurements in patch clamp experiments as a fourth independent technique. The capacitances were recorded continuously and read directly from the instrument. Surprisingly, we found a large and significant increase in cell membrane capacitance, which is known to correlate to an increase in plasma membrane surface area (Fig. 2f). These results suggest that changes in cell morphology occur mainly due to unfolding of membrane invaginations [15] or membrane exocytosis, which may contribute to membrane disintegration. The data do not support the concept of a persistent cell swelling and do not suggest activation of large nonselective ion currents that would otherwise lead to massive cell swelling [16].

Fig. 2
figure 2

Morphology of necroptotic cells. a 3D-analysis of non-stained NIH3T3 cells using holographic images (HoloMonitorTM, I&L Biosystems; Germany). Induction of necroptosis in NIH3T3 cells by necroptotic cocktail TSZ was paralleled by a time-dependent change of cell morphology. TSZ induced rounding up of the cells and increased cell height, leading to an initial cell swelling followed by an overall morphological cell shrinkage. Vertical scale bar indicates cell height ranging from 0 to 15.71 µm. b Holographic analysis of cell volume confirms initial cell swelling followed by subsequent cell shrinkage. c DIC images of NIH 3T3 cells treated with TSZ. d Analyzed volumes from individual cells indicating initial cell swelling followed by cell shrinkage. e Analysis of TSZ-induced cell volume changes by flow cytometry. Initial cell swelling is followed by cell shrinkage. f Measurement of cell membrane capacitance in patch clamp experiments suggesting an increase in cell membrane surface area during necroptosis. Bar indicates 10 µm. Mean ± SEM (number of cells). *Significant change in cell volume by TSZ (b, p = 0.00001–0.0001; d, p = 0.01–0.03) or capacitance (f, p = 001) (paired t test). #Significant change in cell volume by TSZ (e, p = 0.0001–0.001) (ANOVA)

Full size image

Cell volume was also measured in HT29 cells using quantitative phase microscopy and flow cytometry. HT29 epithelial cells are less susceptible towards necroptotic stimulation and were exposed to TSZ for 16 h to induce necroptosis (Supplementary Fig. 1). The data indicate that also HT29 cells swell within the first 4–5 h of TSZ exposure, but then start to shrink (Supplementary Fig. 2). Similar to NIH3T3 cells, the membrane capacitance is enhanced from 22.9 ± 3.1 to 34.8 ± 3.9 pF (n = 9). Similar swelling and subsequent shrinkage were observed for other cell lines (data not shown). Cell shrinkage was observed at later stages of necroptosis, i.e., several hours after exposure to TSZ. Cell shrinkage is actually a hallmark of apoptotic cell death [17]. To rule out a role of apoptosis, we induced necroptosis in the three different cell lines, NIH3T3 (mouse embryonic fibroblasts), L929 (murine skin fibroblast), and HT29 (human colonic carcinoma epithelial) cells using 100 ng/ml TNFalpha + 25 µM Z-VAD and 100 ng/ml TNF alpha + 25 µM + 1 µM Smac mimetics, respectively, and examined possible activation of apoptosis by Western blotting of Caspase-3. As a positive control, Jurkat T-lymphocytes were treated with 100 ng/ml alpha-Fas (clone 7C11; Immunotech) to induce apoptosis. Capase-3 Westerns demonstrated apoptosis in Jurkat cells, but complete absence of cleaved caspase-3 in necroptotic NIH3T3, L929, and HT29 cells. Apoptosis (cleaved caspase-3) in Jurkat cells was completely blocked by 25 µM Z-VAD (Supplementary Fig. 3). These data suggest that TSZ treatment indeed induced necroptosis.

Activation of Ca2+ permeable membrane channels and Ca2+ influx has been proposed as crucial events in necroptosis, and it has been suggested that Ca2+ influx occurs through TRPM7 channels [16]. We analyzed changes in intracellular Ca2+ occurring upon induction of necroptosis and found an only small increase in basal Ca2+ levels along with induction of Ca2+ oscillations (Fig. 3a, b). Increase in basal Ca2+ and Ca2+ oscillations was completely suppressed by dantrolene, an inhibitor of ryanodine Ca2+ release channels expressed in the endoplasmic reticulum (ER) membrane, while inhibitors of IP3 receptor/Ca2+ release channels (xestospongin C), TRPC membrane Ca2+ influx channels (SK&F96365), or TRPM7 channels (YM58483) were without any impact on intracellular Ca2+ changes (Fig. 3c). None of the chemical compounds used here induced cell death on their own when cells were incubated at the given concentrations for up to 24 h. Thus, our data do not confirm large intracellular Ca2+ increases during necroptosis, a significant Ca2+ influx or a role of TRPM7 channels as reported earlier [16].

Fig. 3
figure 3

Ca2+ signals in necroptotic cells. Analysis of intracellular Ca2+ signals during induction of necroptosis. a Ca2+ oscillations induced by TSZ. b Rise in intracellular Ca2+ levels by TSZ. c Rise in intracellular Ca2+ levels by TSZ was not affected by the (1) inhibitor of IP3 receptors, xestospongin C (1 µM), (2) the inhibitor of ORAI channels, YM58483 (5 µM), or the inhibitor of TRPC channels, SK&F96365 (20 µM), but was blocked by dantrolene (10 µM), an inhibitor of ryanodine receptors. Red bars indicate the presence of TSZ. Mean ± SEM (number of cells). #Significant increase of intracellular Ca2+ levels by TSZ (b, p = 0.0037, 2 h; p = 0.001, 3 h; p = 0.0015, 4 h), and inhibition of Ca2+ increase by dantrolene (c, p = 0.04) (ANOVA)

Full size image

To further exclude an essential role of Ca2+ influx for necroptotic cell death, we performed the experiments in the complete absence of extracellular free Ca2+ using EGTA. Even in the presence of 5 mM EGTA, TSZ-induced necroptosis in NIH3T3 cells was not suppressed (Supplement 4). Moreover, transient oscillatory Ca2+ rises observed in necroptotic cells were not abolished in the presence of Ca2+ free extracellular buffer (data not shown). Additional experiments were performed in TSZ-treated NIH3T3 cells, by chelating intracellular Ca2+ using the membrane permeable Ca2+ chelator BAPTA-AM. Chelating intracellular Ca2+ basically abolished TSZ-induced necroptosis, suggesting an essential role of intracellular Ca2+ for necroptosis (Supplement 4). Further experiments were performed in HT29 colonic epithelial cells. Removal of extracellular Ca2+ did not inhibit, but surprisingly even augmented necroptotic cell death. Moreover, chelating intracellular Ca2+ by BAPTA-AM did not inhibit but increased necroptotic cell death (Supplement 5). In addition, in the other cell lines used in the present study, extracellular Ca2+ removal had no significant effects on TSZ-induced necroptotic cell death (data not shown). Notably, TSZ induced a comparably small Ca2+ increase and activated a tannic acid-sensitive whole cell current in HT29 cells (Supplement 6). Taken together, the present data do not confirm the general importance of Ca2+ influx for necroptotic cell death.

We further examined whether whole cell currents are activated during necroptotic cell death. Indeed, we observed large outwardly rectifying and time-dependent whole cell currents 4 h after incubation with TSZ (Fig. 4a, b). As replacement of extracellular Cl− by gluconate (5Cl−) strongly inhibited these currents, the activated currents show features remarkably similar to TMEM16/anoctamin Cl− channels [18]. Moreover, TSZ-activated currents were inhibited by the ANO6 blocker tannic acid [8, 19] (TA; Fig. 4c). Concentrations of 5–20 µM were used in the present study. These concentrations are not toxic, as they come close to those achieved by consumption of red wine or green tea [20]. Moreover, the activation of the whole cell currents was no longer detected upon siRNA knockdown of ANO6 (Fig. 4d).

Fig. 4
figure 4

Whole cell currents in necroptotic cells. a Time-dependent whole cell currents activated by TZS, shown at clamp voltages ranging from −100 to +100 mV in steps of 20 mV. b Corresponding current voltage relationships indicate the activation of outwardly rectifying whole cell Cl− currents that are inhibited by replacement of extracellular Cl− with gluconate (5Cl−). c Inhibition of TSZ-activated currents by the ANO6 blocker tannic acid (TA, 5–20 µM). d Inhibition of TSZ-activated whole cell currents upon siRNA knockdown of ANO6. Mean ± SEM (number of cells). #Significant increase of whole cell currents by TSZ (b–d, p = 0.0015–0.04) (unpaired t test). *Significant inhibition of whole cell currents by 5Cl− or TA (b, p = 0.001–0.028; c, p = 0.018) (paired t test). $Significant inhibition of TSZ-induced whole cell currents by siRNA-ANO6 (p = 0.039; unpaired t test)

Full size image

To assess a possible contribution of other anoctamins, we analyzed mRNA expression of TMEM16A-K/ANO1-10 in NIH3T3 cells, and found significant levels only for ANO6 and for ANO10, with a minor expression for ANO8 (Fig. 5a, b). ANO6 has been identified as Ca2+-activated phospholipid scramblase and as Ca2+-activated Cl− channel [9, 21–25]. ANO6 produced outwardly rectifying and time-dependent Cl− currents and changed cell volume and membrane properties in activated platelets and macrophages [8, 26, 27]. In contrast to ANO6, which is a membrane-localized protein, ANO10 is located predominately in cytosolic compartments and is associated with the ER. Nevertheless, it has been shown to produce membrane currents and to participate in the activation of macrophages [28]. However, siRNA-knockdown of ANO6 and/or ANO10 expression had only a minor inhibitory effect in necroptotic cell death (Fig. 5d). This result is further corroborated by the fact that none of the anoctamin inhibitors were able to attenuate necroptotic cell death (Fig. 5e). An array of inhibitors of Ca2+ signaling (such as blockers of Ca2+ release or inhibitors of Ca2+ influx) showed only minor effects on necroptotic cell death. We, therefore, propose that Ca2+ signaling or Ca2+-dependent activation of anoctamins is of minor relevance for necroptosis.

Fig. 5
figure 5

Minor role of TMEM16F in necroptotic cell death. a, b RT-PCR analysis of the expression of TMEM16 proteins in NIH3T3 cells indicates dominant expression of TMEM16F (ANO6) and TMEM16K (ANO10). c Confirmation of successful protein knockdown by siRNA (72 h). d Marginal effects of knockdown of TMEM16F (ANO6) and TMEM16 K (ANO10) on necroptotic cell death. e No effects of inhibitors of TMEM16 proteins on necroptotic cell death. CaCCinhAO1 (10 µM), T16inhAO1 (10 µM), tannic acid (TA, 10 µM), and NS3728 (10 µM). f Percentage of necroptotic cells (Annexin V/7-ADD positivity) induced by TSZ and effects of various compounds on cell death. Removal of extracellular Ca2+ (Ca2+ free), xestospongin C, dantrolene, YM58483, SK&F69365, and the TRPM7-inhibitor NS8593 (10 µM) had no or only marginal effects on cell death. Mean ± SEM (number of cells). #Significant inhibition of necroptosis (d, p = 0.003–0.021; f, p = 0.017) (ANOVA)

Full size image

Discussion

The final step during necroptosis requires RIPK3-dependent phosphorylation of MLKL. The formation of putative pMLKL membrane channels has been claimed, which might lead to cell death due to cell swelling and membrane disintegration [7, 16]. Our present data, however, show little evidence for a nonselective current that would allow permeation of both anions and cations. Moreover, rather than cell swelling, membrane reorganization and cell shrinkage were observed. Finally, evidence for a role of TRPM7 channels during necroptosis of NIH3T3 cells could not be provided in the present study [16]. Our data indicate an only small increase in baseline Ca2+ levels, which takes place even in the absence of extracellular Ca2+, indicating a minor role of Ca2+ influx for necroptotic cell death. Instead, the data indicate oscillatory Ca2+ release from intracellular ER Ca2+ stores that is, however, not essential for induction of necroptosis, because the inhibition of Ca2+ release by dantrolene does not block necroptotic cell death. [7]. Notably, identical results were obtained in TSZ-treated HT29 colonic carcinoma cells, i.e., marginal Ca2+ increase with activation of anoctamin currents, which had both little effect on necroptotic cell death (Supplement 6). It is currently unclear whether the Ca2+ increase reported here is related to pMLKL-induced mitochondrial permeabilization observed earlier [29]. Notably, in another study, TNF-induced necroptosis was observed even in the absence of mitochondria [11]. Finally, we cannot confirm an essential role of Ca2+ influx through Ca2+ permeable TRPC channels as reported earlier [30]. Thus, neither mitochondrial permeabilization, nor Ca2+ release from ER stores seemed to be essential for necroptotic cell death. In addition, our data do not supply evidence for an essential role of TMEM16F/ANO6 or TMEM16K/ANO10 during necroptosis. The recently identified volume-regulated anion channel (VRAC) LRRC8A [4, 5] may also be irrelevant for necroptotic cell death, as neither knockdown of LRRC8A nor the VRAC inhibitor NS3728 interfered with necroptotic cell death (data not shown). Future studies may look into the role of other LRRC8-isoforms, such as LRRC8D [3]. Taken together, the present study demonstrates the activation of anoctamin 6 phospholipid scramblase/ion channel during necroptosis. Activation of ANO6 appears to be a cellular event that takes place parallel to the actual necroptotic cell death, but, nevertheless, may be highly relevant in the physiological/pathophysiological context, e.g., by producing a so-called “eat me” signal [31].

Materials and methods

Cells, cDNA, RT-PCR

Mouse embryonic fibroblasts (NIH3T3) [11], HT29 (human colonic carcinoma cells) [32], 16HBE (human immortalized bronchial epithelial cells) [33], H441 (human airway club cells) [34], and L929 (murine skin fibroblasts) cells [35] were grown as described in earlier publications Generation of cDNA for ANO6 and transfection/expression of ANO6 has been report earlier [9]. RT-PCR analyses were performed using the standard conditions and appropriate primers. Cells were treated with 10 ng/ml (NIH3T3; L929) or 100 ng/ml (HT29, 16HBE, H441) TNF-α, 5 µM smac minetic, and 25 µM Z-VAD, (TSZ) to induce necroptosis.

Western blotting

Protein was isolated from cells using a sample buffer containing 50 mM Tris–HCl, 150 mM NaCl, 50 mM Tris, 100 mM dithiothreitol, 1 % Nonidet P-40, 0.5 % sodium deoxycholate, and 1 % protease inhibitor mixture (Sigma, Taufkirchen, Germany). Proteins were separated by 8.5 % SDS-PAGE and transferred to a polyvinyl membrane (GE Healthcare, Munich, Germany). Membranes were incubated with primary anti-ANO1 rabbit polyclonal AB (Davids Biotech, Regensburg, Germany; 1:1000) or anti-phospho-MLKL antibody (Abcam. USA; 1:2000) overnight at 4 °C. Proteins were visualized using horseradish peroxidase-conjugated secondary antibody and ECL detection.

Measurement of [Ca2+]i

Measurement of the intracellular Ca2+ concentration was performed as described recently [36]. In brief, cells were loaded either with 5 μM Fura2-AM (to measure global cytosolic Ca2+ changes) in Ringer solution at 37 °C for 30 min. Fluorescence was detected at 37 °C, using an inverted microscope IMT-2 (Olympus, Nuremberg, Germany) and a high-speed polychromator system (Visi-Chrome, Puchheim, Germany). The results were obtained at 340/380 nm fluorescence ratio (after background subtraction). After calibration [36], intracellular Ca2+ concentrations were calculated.

Patch clamping

Cells were grown on coated glass cover slips. If not indicated otherwise, patch pipettes were filled with a cytosolic-like solution containing KCl 30, K -gluconate 95, NaH2PO4 1.2, Na2HPO4 4.8, EGTA 1, Ca -gluconate 0.758, MgCl2 1.03, d-glucose 5, ATP 3, and pH 7.2. The intracellular (pipette) Ca2+ activity was 0.1 µM. Coverslips were mounted in a perfused bath chamber on the stage of an inverted microscope (IM35, Zeiss) and kept at 37 °C. The bath was perfused continuously with Ringer solution at a rate of 8 ml/min. Patch pipettes had an input resistance of 2–4 MΩ when filled with the cytosolic-like (physiological) solution. Currents were corrected for serial resistance. The access conductance was measured continuously and was 60–140 nS. Currents (voltage clamp) and voltages (current clamp) were recorded using a patch clamp amplifier (EPC 7, List Medical Electronics, Darmstadt, Germany), the LIH1600 interface, and the PULSE software (HEKA, Lambrecht, Germany) as well as the Chart software (AD Instruments, Spechbach, Germany). The capacitances were recorded continuously and read directly from the instrument. Data were stored continuously on a computer hard disc and analyzed using the PULSE software. In regular intervals, membrane voltage (Vc) was clamped in steps of 20 mV from −100 to +100 mV from a holding voltage of −100 mV. Current density was calculated by dividing the whole cell currents by cell capacitance.

Fluorescence-activated cell sorting

Phosphatidylserine exposure to the outer cell membrane of apoptotic cells or at the inner plasma membrane of necrotic cells and incorporation of 7-AAD into necrotic cells was quantified by fluorescence-activated cell-sorting (FACS) analysis. The ApoAlert annexin V–FITC antibody and the 7-AAD antibody were purchased from BD Biosciences.

HoloMonitor

Quantitative phase microscopy was applied to detect cell morphology and to calculate for cell volume [12]. Quantitative phase shift imaging allows for non-invasive long-term imaging of non-labelled cells in cell culture incubators (37 °C, humidified air, 5 % CO2). Holographic microscopy is used in the HoloMonitorTM time-lapse cytometer (Phase Holographic imaging PHI, Lund, Sweden).

Data and statistics

Data are shown as individual traces or as summaries with mean values ± SEM and number of experiments or cells given in parenthesis. For statistical analysis of unpaired data, ANOVA or unpaired t test was used as appropriate. For the statistical analysis of paired data, paired t test was used. A p value of <0.05 was accepted as statistically significant difference (indicated by #, $ for unpaired data and by * for paired data). Individual p values are given in the figure legends.