Uncovering molecular mechanisms of regulated cell death in the naked mole rat

M. musculus fibroblasts are more sensitive to DNA damage compared to fibroblasts from H. glaberH. glaber

To compare the sensitivity of skin fibroblasts from H. glaber vs. M. musculus towards DNA damage, metabolic assays were carried out. For this comparative analysis, immortalized skin fibroblasts from both H. glaber and M. musculus, e.g. NSF8 and 3T3 cell lines, respectively, (termed thereafter NMR and mouse cells) were taken in the exponential growth phase (Supplementary Figure 1). NMR and mouse cells were treated in a time- and dose-dependent manner with MMS, 5FU and Eto (Figure 1). A notable decrease in the metabolic activity of NMR cells was detected at higher concentrations of MMS, 5FU and Eto compared to mouse cells. This was observed for all time intervals. Moreover, upon treatment with the same concentrations of MMS, Eto and 5FU, a longer exposure time was needed to reduce the metabolic activity of NMR cells in comparison to mouse cells (Figure 1). It has to be noted that the concentrations of MMS, 5FU and Eto, that are commonly used in the studies of DNA damage-induced apoptosis, are in the range between 10 and 70 μM. Indeed, a notable metabolic activity loss of mouse fibroblasts was observed upon treatment with 10 to 70 μM MMS, 5FU and Eto (Figure 1). However, incubation using the same concentration range of DNA damaging agents, led to an only marginal decrease of the metabolic activity in the case of NMR fibroblasts. Moreover, the drop of metabolic activity of NMR fibroblasts was observed starting from the concentration of 100 μM only. In particular, a strong decrease in metabolic activity (20-40%) was observed in 16 hours for the treatment of NMR fibroblasts with 280 μM of MMS, 320 μM of 5FU and 120 μM of Eto (Figure 1, right side). Taken together, we can conclude that M. musculus fibroblasts are more sensitive to DNA damage-induced metabolic activity loss compared to fibroblasts from H. glaber.

M. musculusfibroblasts are characterized by a higher caspase activity upon treatment with DNA damage agents compared to fibroblasts from H. glaber

The differences in metabolic response to DNA damage allowed suggesting that H. glaber and M. musculus might have a different sensitivity to DNA damage-induced cell death. Activation of effector caspases-3 and -7 is a hallmark of apoptosis. To examine effector caspase activation in fibroblasts from M. musculus vs. H. glaber caspase-3/7 activity was measured in a time- and dose-dependent manner. For mouse fibroblasts, we selected concentrations of MMS, 5FU and Eto that cause strong changes in metabolic activity (Figure 1). Treatment with these concentrations of all three DNA damaging agents caused an increase of caspase-3/7 activity in mouse fibroblasts. The maximal value of caspase activity was reached in 16 h after addition of MMS and 5FU (Figure 2A, 2B), and from 16h to 20 h for Eto treatment (Figure 2A–2C). However, the observed increases in caspase activity were not very high: 1,6 fold for MMS, 2-fold for 5FU and 3-fold for Eto (Figure 2D). Moreover, the maximal values of caspase activity were relatively robust towards the dose of DNA damaging agent. This indicates that mouse fibroblasts show only moderate levels of caspase-3/7 activity upon incubation with MMS, 5FU and Eto.

To compare caspase activation in fibroblasts from H. glauber vs. M. musculus the same concentrations of MMS, 5FU and Eto were used for NMR cells. In line with the observed lower sensitivity in metabolic response, fibroblasts from H. glaber showed lower levels of caspase activity compared to fibroblasts from M. musculus. Only a small induction of caspase-3/7 activity was detected upon MMS treatment of NMR cells, while almost no increase of caspase activity was observed upon Eto and 5FU treatment (Figure 2A–2D). In particular, upon treatment with 70 μM MMS, the increase in caspase activity was observed already at 8 hours after MMS treatment, peaking at 16 hours. A similar dependence was observed upon higher concentrations of MMS (Figure 2A). Interestingly, the same time dependence was also found for 5FU and Eto treatments, though caspase-3/7 activity was rather weak within the selected concentration range of 5FU and Eto.

Collectively, the DNA damage-induced caspase activity of fibroblasts from H. glaber was much lower than that of M. musculus. Next we aimed to compare the cell death responses of H. glaber vs. M. musculus fibroblasts upon MMS, 5FU and Eto treatment.

Flow cytometry indicates a higher rate of cell death in fibroblasts from M. musculusvsH. glaber

The cell death induction in fibroblasts from M. musculus vs H. glaber was compared using flow cytometry analysis with FITC-AnnexinV and propidium iodide (PI) staining. The measurements were performed at the time point of maximal caspase activity assuming that under these conditions cells show a strong apoptotic response. Accordingly, NMR and mouse cells were exposed to Eto as well as 5FU for 20 h; and 5FU for 16 h. The relative number of viable (AnnexinV^-/PI^-), apoptotic (AnnexinV⁺/PI^-) and late apoptotic/necrotic (AnnexinV⁺/PI⁺) cells was then measured. Importantly, we assumed that a population of double positive AnnexinV⁺/PI⁺ cells might comprise both late apoptotic or secondary necrotic cells as well as cells that undergo regulated necrosis [20]. The latter is based on reports that DNA damage might induce the regulated type of necrosis, necroptosis, in particular upon high concentrations of DNA damaging agents [21].

In accordance with the results from caspase activity and metabolic assays, NMR cells turned out to be more resistant towards DNA damage-induced cell death compared to mouse cells. In NMR cells, the relative number of viable cells decreased at higher concentrations of reagents compared with mouse cells (Figure 3). Furthermore, early apoptotic cells (AnnexinV⁺/PI^-) and double positive AnnexinV⁺/PI⁺ cells were detected upon treatment with higher concentrations of DNA damaging agents in NMR cells than in mouse cells. In particular, strong differences were observed upon 5FU treatment, where NMR cells remained largely resistant (Figure 3). Differences were also observed upon treatment with MMS and Eto. In particular, strong differences were monitored upon incubation with 70 and 140 μM of MMS as well as 15 and 30 μM of Eto. Importantly, treatment with all three DNA damaging agents led to the detection of a population of early apoptotic cells (AnnexinV⁺/PI^-) indicating that NMR cells undergo apoptosis. Though, as mentioned above, this population was much smaller in NMR cells compared to mouse cells treated with the same concentration of DNA damaging agents. Upon increase of the concentration of DNA damaging agents, a population of double positive AnnexinV⁺/PI⁺ cells were observed. As mentioned above, we assumed that this population might comprise both late apoptotic and necroptotic cells.

To examine whether high concentrations of DNA damaging agents induce the necroptotic type of cell death in NMR cells, we applied a pharmacological inhibitor of necroptosis (nec-1). Nec-1 inhibited the loss of metabolic activity induced by 280 μM of MMS (Supplementary Figure 2). These results indicate that high concentrations of DNA damaging agents induce the regulated necrotic pathway. Taken together, we have shown that MMS, Eto and 5FU induce apoptotic and likely necrotic types of cell death in NMR cells. Moreover, NMR cells were less sensitive towards DNA damage-induced cell death compared to mouse cells.

Electron microscopy supports the apoptotic and necrotic phenotype of cell death in fibroblasts from M. musculus and H. glaber

The morphology of dying cells is a key feature indicating the type of cell death. We performed electron microscopy (EM) analysis of NMR and mouse fibroblasts treated with toxic agents to verify the presence of different morphological features of cell death. The ultrastructure of both NMR and mouse cells in control samples (Figure 4A, 4B) did not show any signs of destruction. In particular, the cytoplasm with average electron density contained a well-developed endoplasmic reticulum and other organelles. Large nuclei in NMR cells were more indented in comparison to those in mouse cells. The nuclei of both cell kinds contained several nucleoli. Control samples of NMR and mouse cells contained very few necrotic cells (0-1 per 100 cells).

Incubation of NMR and mouse cells with 280 μM of MMS during 16 h resulted in a number of swollen cells in both cell types, which might represent the first step towards necrosis (Figure 4C). However, the plasmalemma of these cells had retained its integrity, and therefore the state of these cells could not be attributed to necrosis, but rather to late apoptosis. In addition, a number of necrotic cells with disrupted plasmalemma, nucleus and other organelles were observed (Figure 4D). The ultrastructure of necrotic cells did not differ in NMR and mouse cells, as in the case of apoptosis, the signs of which were observed in both cell types incubated with MMS. This indicated that these stimulation conditions led to the induction of both types of cell death: apoptosis and regulated necrosis. Furthermore, it has to be noted that examining the samples after 16 hours of treatment, we observed the cells at different stages of apoptosis. This explains differences in apoptotic cells observed in ultrathin section, in particular, in morphology of the nucleus and the presence of apoptotic blebbing (Figure 4E, 4F). The obtained data were in good agreement with flow cytometry analysis and observations that upon incubation with 280 μM of MMS for 16 h, the majority of cells in both NMR and mouse fibroblasts are double positive (Annexin V⁺/PI⁺). The latter population might contain both late apoptotic or secondary necrotic as well as necroptotic cells, which is in good agreement with the EM analysis.

Examination of mouse cells incubated with 160 μM of 5FU for 16 h revealed many destroyed cells with morphologic signs of necrosis (3-4.5 per 100 nucleus containing cells), while necrotic cells in NMR samples were rare (0.5-1 per 100 cells). Cells with signs of apoptosis (Figure 5A, 5B) were relatively rare in both NMR and mouse samples. This was in accordance with the flow cytometry analysis, in which a lower sensitivity of NMR cells towards 5FU had been detected. Moreover, these data fit well to the suggestion that upon high concentrations of DNA damaging agents, cells rather undergo regulated necrosis, which was observed in these experiments.

For EM analysis with Eto, we took the lower concentration of Eto to examine the initial steps of cell death induction in NMR and mouse fibroblasts. Cells of NMR and mouse samples under Eto (60 μM) treatment for 16 h mostly did not show visible signs of destruction, however, some NMR cells were observed to swell. Moreover, cytoplasm and nucleus had less electron density than untreated cells (Figure 5C). Necrotic cells were not numerous in either NMR or mouse samples treated with Eto (0.8-1.0 per 100 cells). A similar frequency was observed with respect to apoptotic cells (Figure 5D). These results are in accordance with the data obtained with flow cytometry, which indicated a very low number of dying cells with this concentration of Eto and the induction of both cell death pathways.

Thus, electron microscopic examination of ultrathin sections of NMR and mouse cells incubated with MMS, 5FU, and Eto revealed a correspondence of the morphological characteristics of the cell samples to those established by flow cytometry analysis. These analyses have demonstrated that both pathways, apoptosis and regulated necrosis, are induced upon MMS, 5FU, and Eto treatment, which is in the good agreement with caspase activity and flow cytometry assays.

Cell culture

The establishment of cell cultures derived from NMR fibroblasts was described earlier [7]. Immortalized NMR skin fibroblasts, NSF8, were kindly provided by Dr. V. Gorbunova (University of Rochester, Department of Biology). NMR cells were cultured in MEM supplemented with 15% FBS (Gibco), 10% AmnioMAX II Complete Medium (Gibco), 5 ng/ml bFGF, 10⁵ U/L penicillin, 100 mg/L streptomycin, and 2.5 mg/L amphotericin β at 32° C in the presence of 5% CO₂. The cell line was deposited in the cell bank of the Institute of Molecular and Cellular Biology, SB RAS (“The General Collection of Cell Cultures”, No. 0310-2016-0002).

The 3T3 mouse cell line was kindly provided by Dr. V. Fishman (Sector of Genomic Mechanisms of Ontogenesis, Institute of Cytology and Genetics, Russia). Cells were cultured in MEM supplemented with 15% FBS (Gibco), 10⁵ U/L penicillin, 100 mg/L streptomycin, and 2.5 mg/L amphotericin β at 37° C in the presence of 5% CO₂.

All experiments were carried out with dividing cells in the exponential growth phase. Cell growth curves (Supplementary Figure 1) were used to determine the parameters of cell growth.

Metabolic assay

Cells were seeded in 96-well plates at a concentration of 3 × 10³ cells/well in 200 μl culture medium and cultured as described above for 24 (NSF8 cells: NMR fibroblasts) or 12 (3T3 cells: Mouse fibroblasts (MMU)) h. Afterwards, the compounds dissolved in DMSO were added, and the cells were incubated for various time intervals. The EZ4U test (Biomedica) was carried out according to the manufacturer’s instructions. Briefly, 20 μl of EZ4U labeling mixture was added, and the cells were incubated for 2–3 h. Absorbance of the samples was measured at 492 nm using a microplate reader CLARIOstar Plus (BMG LABTECH GmbH). All experiments were repeated in the triplicate.

Apoptosis detection

NMR (NSF8 cells) and MMU (3T3 cells) cells were seeded in 6-well plates and cultured as described above. When the cell cultures reached 80% confluence, the cells were treated with the DNA damaging compounds for 16 (5FU) or 20 (Eto and MMS) h. Before treatment with trypsin, detached cells were collected. Adherent cells were rinsed with PBS and detached from the plate with trypsin. Soybean trypsin inhibitor (Sigma) was added up to 10 μg/ml concentration to inhibit trypsin-initiated proteolysis. The detached and trypsinized cells were combined and pelleted by centrifugation (5 min at 500 g). FITC-Annexin V Apoptosis detection kit and FACSCanto II (BD Biosciences, San Jose, USA) flow cytometer were used to detect cell death. The analysis was performed on a FACSCantoII flow cytometer (Becton Dickinson) using the FACSDiva Software (BD Biosciences). Cells were initially gated based on forward scatter vs. side scatter to exclude small debris. Ten thousand events from each population were collected. The inhibitors nec-1 (Sigma-Aldrich, USA) and z-VAD-fmk (BD Biosciences) dissolved in DMSO were added 2 hours before addition of DNA damaging reagents.

Caspase activity detection

Cells were seeded in 96-well plates at a concentration of 3 × 10³ cells/well in 200 μl culture medium and cultured as described above to reach 60% confluence. Caspase activity was measured using the Caspase-Glo 3/7 assay kit according to the manufacturer's instructions (Promega, Madison, WI). Briefly, 200 μl of Caspase-Glo 3/7 reagent was added to each well; the plate was shaken gently for 30 s and incubated at room temperature for 2 h. Luminescence of the samples was measured using an Infinite M200 plate reader (Tecan, Research Triangle Park, USA). The proluminescent substrate containing DEVD tetrapeptide was used. This substrate is capable of being cleaved by caspase-3/7 resulting in a release of aminoluciferin, a luciferase substrate. The activity of effector caspases-3/7 can be estimated according to luminescence intensity.

Electron microscopy

NSF8 and 3T3 cell pellets were fixed after corresponding treatment in 4% paraformaldehyde and post-fixed in 1% osmium tetroxide solution, then routinely processed and embedded in Epon–Araldite mixture (SPI, USA). Semithin sections were prepared from hard blocks, stained with Azur 2, and examined by light microscopy to select the areas for ultrathin sectioning. Ultrathin and semithin sections were cut using an UC7 ultramicrotome (Leica, Germany), contrasted using uranyl acetate and lead citrate (SPI, USA), and examined using a JEM 1400 transmission electron microscope (Jeol, Japan). The images were collected using a side-mounted Veleta digital camera (EM SIS, Germany).

Statistical analysis

Statistical analysis was performed using Statistica 10 software. All experiments were performed at least in triplicate. In the figures, bars show the standard deviation. The level of confidence was estimated using the Mann–Whitney U test.