Ca2+-mediated mitochondrial reactive oxygen species metabolism augments Wnt/β-catenin pathway activation to facilitate cell differentiation

Abstract

Emerging evidence suggests that reactive oxygen species (ROS) can stimulate the Wnt/β-catenin pathway in a number of cellular processes. However, potential sources of endogenous ROS have not been thoroughly explored. Here, we show that growth factor depletion in human neural progenitor cells induces ROS production in mitochondria. Elevated ROS levels augment activation of Wnt/β-catenin signaling that regulates neural differentiation. We find that growth factor depletion stimulates the release of Ca(2+) from the endoplasmic reticulum stores. Ca(2+) subsequently accumulates in the mitochondria and triggers ROS production. The inhibition of mitochondrial Ca(2+) uptake with simultaneous growth factor depletion prevents the rise in ROS metabolism. Moreover, low ROS levels block the dissociation of the Wnt effector Dishevelled from nucleoredoxin. Attenuation of the response amplitudes of pathway effectors delays the onset of the Wnt/β-catenin pathway activation and results in markedly impaired neuronal differentiation. Our findings reveal Ca(2+)-mediated ROS metabolic cues that fine-tune the efficiency of cell differentiation by modulating the extent of the Wnt/β-catenin signaling output.

Keywords: Calcium Signaling; Dishevelled; Human Neural Progenitor Cells; Mitochondrial Metabolism; Nucleoredoxin; Reactive Oxygen Species (ROS); Redox Signaling; Wnt Pathway.

FIGURE 1.

Induction of differentiation alters intracellular redox balance. A, confocal images of neurons (βIII-tubulin, green) and glia cells (red) at 0 h (proliferating cells) and 2 days after initiation of differentiation. Nuclei are shown in blue. Mature neurons were quantified based on βIII-tubulin labeling. Immature neurons were quantified through the labeling of the neuronal RNA-binding protein Hu-antigens C and D (HuC/D) that is expressed earlier than βIII-tubulin. n = ∼9000 cells per time point. B, confocal images of redox state (grayscale; green in merge) at 0, 0.5, and 2.5 h of differentiation. White arrows indicate faint signal at 0 and 2.5 h. Phospholipids (red) and cell boundaries (dotted white lines) are shown. C, kinetics of the cellular redox state measured as mean fluorescent intensity at 10-min intervals over the 1st h of differentiation. Significant increase appears at 30 min of differentiation. n = ∼150 cells per time point. D, kinetics of the cellular redox state measured as mean fluorescent intensity at 0.5-h intervals over the first 3 h of differentiation. Redox state reaches baseline levels after 3 h. n = ∼150 cells per time point. E, kinetics of the cellular redox state measured as mean fluorescent intensity at 0.5-h intervals over the first 3 h of differentiation using flow cytometry. a.u., arbitrary units. F, confocal images of intracellular redox state (grayscale; green in merge) after three sequential exchanges of proliferating medium in pre-stained proliferating cells. Phospholipids are in red. Mean fluorescent intensities show change in redox state only in differentiating cells. n = ∼50 cells per time point. G, cytotoxic effect of 3 mm H₂O₂ assessed with MTT. *, p ≤ 0.05. Error bars, S.D. Scale, 10 μm.

FIGURE 2.

Changes in mitochondrial ROS metabolism during differentiation. A, confocal images of the mito-ROS levels (glow dark) at 0, 0.5, and 2.5 h of differentiation. B, kinetics of mito-ROS metabolism measured as mean fluorescent intensity at 10-min intervals over the 1st h of differentiation. n = ∼150 cells per time point. C, kinetics of mito-ROS metabolism measured as mean fluorescent intensity at 0.5-h intervals over the first 3 h of differentiation. n = ∼200 cells per time point. prolif., proliferating. D, flow cytometry as parallel determination confirming the variation of ROS levels. *, p ≤ 0.05. Error bars, S.D. Scale, 10 μm; a.u., arbitrary units.

FIGURE 3.

Changes in intracellular Ca²⁺ compartmentalization during differentiation. A, confocal images of ER-Ca²⁺ levels (green) at 0 and 0.5 h of differentiation. White arrows in 0.5-h panel indicate residual ER-Ca²⁺ as punctate signal. Mean fluorescent intensities show Ca²⁺ release 30 min after initiation of differentiation. n = ∼100 cells per time point. B, confocal image of ER-Ca²⁺ levels (green) in proliferating (prolif.) cells treated for 1 h with 1 mm CAF. Cell boundaries are shown by dotted white lines. Only a punctate fluorescence signal remains (white arrow), consisting of the residual ER-stored Ca²⁺. Histogram shows mean fluorescent intensity values of ER-Ca²⁺. n = ∼50 cells per time point. C, cytotoxic assay of CAF treatment (10 mm) using trypan blue exclusion assay. D, confocal images of the Ca²⁺-unbound form of fura red (glow dark) at 0, 0.5, and 3 h of differentiation. The decrease in the signal (0.5 h) reflects the increase of the cyto-Ca²⁺ levels. E, kinetics of the cyto-Ca²⁺ levels. An increase at 0.5 h is followed by a return to baseline after 3 h of differentiation. n = ∼150 cells per time point. F, confocal image of cyto-Ca²⁺ levels (glow dark) after three sequential exchanges of culture medium in proliferating cells. The signal was quantified and compared with control and differentiating cells. n = ∼50 per time point. G, confocal image of Ca²⁺-unbound form of fura red (glow dark) in CAF-treated proliferating cells (10 mm, 1 h). *, p ≤ 0.05. Error bars, S.D. Scale, 10 μm; a.u., arbitrary units.

FIGURE 4.

Changes in mitochondrial Ca²⁺ during differentiation. A, confocal images of mito-Ca²⁺ levels (glow dark) at 0, 0.5, and 3 h of differentiation. B, histogram shows mean fluorescent intensity values of mito-Ca²⁺ levels during the first 3 h of differentiation. An increase in Ca²⁺ influx into mitochondria occurs at 0.5 h of differentiation. n = ∼150 cells per time point. C, confocal image of mito-Ca²⁺ levels (glow dark) in CAF-treated proliferating (prolif.) cells (1 mm, 0.5 h). Histogram shows mean fluorescent intensity values of mito-Ca²⁺. n = ∼100 cells per time point. *, p ≤ 0.05. Error bars, S.D. Scale, 10 μm; a.u., arbitrary units.

FIGURE 5.

Mitochondrial Ca²⁺ fluxes regulate ROS metabolism. A, schematic of the 3-h RuR treatment as follows: 1 h treatment of proliferating cells with RuR + GF is followed by a 2-h treatment of differentiating cells with RuR-GF. The medium is replaced by a drug-free GF-free medium at 2 h of differentiation. B, histogram shows mean fluorescent intensity values of mito-Ca²⁺ levels for low (gray) and high RuR dose (black). n = ∼100 cells per time point. C, kinetics of mito-Ca²⁺ levels in untreated (white) and low RuR dose-treated cells (gray) plotted as mean fluorescent intensities. n = ∼150 cells per time point. D, confocal images of mito-Ca²⁺ levels (glow dark) in untreated and 0.5 μm RuR-treated cells at 0, 1, and 3 h of differentiation. 0.5 μm RuR inhibits Ca²⁺ influx within mitochondria, but its effect is reversible. E, confocal images of mito-ROS metabolism (glow dark) in untreated and 0.5 μm RuR-treated cells at 0, 1, and 3 h of differentiation. RuR prevents the rise of ROS production. F, kinetics of mito-ROS levels plotted in untreated (white) and low RuR dose-treated cells (gray) as mean fluorescent intensities. Drug removal reverses the effect shown by significant elevation of mito-ROS levels at 3 h of differentiation. n = ∼300 cells per time point. G, histogram shows mean fluorescent intensity values of mito-ROS levels for low (gray) and high RuR dose (black). n = ∼100 cells per time point. H, cells were seeded at the same concentration prior to RuR treatment, and the induction of differentiation and cell number was scored at 72 h of differentiation. Histogram shows mean cell density values per individual treatments and untreated cells. *, p ≤ 0.05. Error bars, S.D. Scale, 10 μm; a.u., arbitrary units.

FIGURE 6.

DVL2 and NRX amounts are affected by Ca²⁺-mediated ROS metabolism. A, Western blots of DVL2 and NRX showing a ROS-dependent increase of both protein amounts occurring 1 h after differentiation or after H₂O₂ treatment (1 mm). All bands are from the same blot. The signal intensities were normalized to 1 at 0 h (proliferating cells) and quantified as a fold change. B, confocal images of DVL2 (green) and NRX (glow dark) in untreated and 0.5 μm RuR-treated cells at 0, 1, and 3 h of differentiation. RuR inhibits the increase of both proteins (1 h) and removal of RuR allows for the cytosolic accumulation of both proteins (3 h). Mean fluorescent intensities are quantified in bar graphs. n = ∼200 cells per time point. *, p ≤ 0.05. Error bars, S.D. Scale, 10 μm; a.u., arbitrary units.

FIGURE 7.

Ca²⁺-mediated ROS metabolism modulates DVL2-NRX complex. A, representative pseudocolor images illustrating FRET_eff resulting from the fully corrected FRET signal at 0, 1, and 3 h of differentiation in untreated and 0.5 μm RuR-treated cells and in 1 mm H₂O₂-treated proliferating cells. The color scale at the bottom of the panel is shown in %. B, mean values of FRET_eff of images shown in A. FRET_eff in untreated cells is reduced by half in 1 h after GF depletion, although it remains unchanged in RuR-treated cells. 1 h after RuR removal, DVL2-NRX complex begins to dissociate as shown by ∼30% reduction in FRET_eff, n = ∼100 cells per time point. prolif., proliferating. *, p ≤ 0.05. Error bars, S.D. Scale, 10 μm.

FIGURE 8.

Nuclear accumulation of β-catenin is regulated by Ca²⁺-mediated ROS metabolism. A, confocal images of β-catenin (glow dark) showing that RuR treatment inhibits its nuclear accumulation. Nuclei are shown in green. Mitotic cells are shown by asterisks. a.u., arbitrary units. B, mean intensities of nuclear β-catenin; C, histograms of frequency distribution of the fluorescent intensities show increased nuclear accumulation of β-catenin and formation of two cell populations, with and without nuclear β-catenin. RuR treatment prevents the formation of the population with nuclear β-catenin. n = ∼600 cells per time point. *, p ≤ 0.05. Error bars, S.D. Scale, 20 μm.

FIGURE 9.

ITRP1 and MCU activities regulate the Ca²⁺-mediated ROS metabolism. A, ITPR1 and MCU mRNA levels (fold changes) were analyzed by quantitative real time PCR and expressed at 0, 24, and 48 h after the differentiation was initiated. Data for untreated cells (ctl) were compared with cells transfected with ITPR1, MCU, or untargeted (−) ctl siRNAs. B, confocal images of ER-Ca²⁺ levels (green) in (−) control and (−) ITPR1 at 0 and 1 h of differentiation. The bar graphs present the mean fluorescent intensity values (a.u.). a.u., arbitrary units. ITPR1 silencing leads to the sequestration of Ca²⁺ in the ER, although Ca²⁺ is released in (−) control. n = ∼100 cells per time point. C, confocal images of mito-Ca²⁺ levels (glow dark) in (−) control and (−) MCU at 0 and 1 h of differentiation. The bar graphs show the mean fluorescent intensity values. MCU silencing prevents Ca²⁺ accumulation in mitochondria. n = ∼100 cells per time point. D, confocal images of mito-ROS levels (glow dark) in (−) control, (−) MCU, and (−) ITPR1 at 0 and 1 h of differentiation. The histograms present the mean fluorescent intensity values. Both MCU and ITPR1 silencing block the rise of ROS metabolism. n = ∼100 cells per time point. *, p ≤ 0.05. Error bars, S.D. Scale, 5 μm.

FIGURE 10.

Ca²⁺-mediated ROS metabolism modulates β-catenin-dependent neuronal differentiation. A, AXIN2 and MAP2 mRNA levels detected by quantitative real time PCR at 0 and 48 h after the differentiation was induced. Data for untreated cells (ctl) are compared with cells treated with RuR or LiCl, and with transfected cells as follows: (−) control, (−) MCU, and (−) ITPR1. Results show that both chemical and genetic disruption of Ca²⁺-mediated ROS metabolism down-regulated the gene response of AXIN2 and MAP2. B, confocal images of neurons (green) and glia cells (red) 3 days after initiation of differentiation showing less neurons when cells were treated with 0.5 μm RuR or 10 mm NAC, although 20 mm LiCl increased the neuronal amount. Nuclei are shown in blue. C, bar graph shows the quantification of the neuronal yields at 3 days of differentiation. n = ∼5000 cells per condition. D, intracellular ROS levels measured in untreated and NAC-treated cells using flow cytometry. Fluorescence intensities of dihydrorhodamine 123 (DHR123) were averaged at 0, 1, and 2 h after induction of differentiation. *, p ≤ 0.05. Error bars, S.D. Scale, 50 μm; a.u., arbitrary units. ns, not significant.

FIGURE 11.

Schematic model of neuronal differentiation of human neural progenitors mediated by Ca²⁺-dependent mitochondrial ROS metabolism. Dashed arrows, fluxes; full arrows, stimulation; β-cat, β-catenin.