Growth factor erv1-like modulates Drp1 to preserve mitochondrial dynamics and function in mouse embryonic stem cells

Abstract

The relationship of mitochondrial dynamics and function to pluripotency are rather poorly understood aspects of stem cell biology. Here we show that growth factor erv1-like (Gfer) is involved in preserving mouse embryonic stem cell (ESC) mitochondrial morphology and function. Knockdown (KD) of Gfer in ESCs leads to decreased pluripotency marker expression, embryoid body (EB) formation, cell survival, and loss of mitochondrial function. Mitochondria in Gfer-KD ESCs undergo excessive fragmentation and mitophagy, whereas those in ESCs overexpressing Gfer appear elongated. Levels of the mitochondrial fission GTPase dynamin-related protein 1 (Drp1) are highly elevated in Gfer-KD ESCs and decreased in Gfer-overexpressing cells. Treatment with a specific inhibitor of Drp1 rescues mitochondrial function and apoptosis, whereas expression of Drp1-dominant negative resulted in the restoration of pluripotency marker expression in Gfer-KD ESCs. Altogether, our data reveal a novel prosurvival role for Gfer in maintaining mitochondrial fission-fusion dynamics in pluripotent ESCs.

Figure 1.

A role for Gfer in the maintenance of ESC pluripotency. (A) Representative immunoblots depicting protein levels of Gfer and Actin in Lac-Z, Gfer1, and Gfer2 KD ESCs (n = 3). (B) Digital optical section ApoTome images (magnification, 200×) of day 6 lacZ, Gfer1, and Gfer2 KD ESCs depicting alkaline phosphatase activity and levels of Nanog, Oct-4, and SSEA1 (all in red). Numbers in black and green represent average signal intensities ± SD quantified from at least 30 colonies (n = 3), *p = 0.001. (C) Average mRNA levels of Nanog, Oct-4, and Gfer in day 6 WT, WT-lacZ-KD, Gfer1-KD, and Gfer2-KD ESCs. The mRNA levels measured by qRT-PCR were normalized to actin mRNA (n = 3), *p = 0.01. (D) Embryoid bodies formed by WT-lacZ-KD and Gfer1-KD ESCs cultured in the absence of feeder cells and LIF in the media for 2 or 5 d. Magnification, ×100; scale bar, 5 μm. Average surface area in μm² ± SD and average numbers of day 5 EBs per well ± SD from each genotype are shown (n = 3); *p = 0.005.

Figure 2.

Down-regulation of Gfer diminishes proliferation and cell survival in WT ESCs. (A) Average cumulative cell yield from n = 3, when WT-lacZ-KD, Gfer1-KD, and Gfer2-KD ESCs were cultured for seven passages (data from five passages are shown); *p = 0.001. (B) Self-renewal by lacZ, Gfer1, Gfer2, and Nanog KD ESCs, initially mixed (P0) at 3:1 ratio with WT ESCs, measured as percentage (%) of GFP-positive cells at all five passages by flow cytometry. Average percentage of GFP-positive ESCs from three experiments is shown; *p = 0.003. (C) Representative histogram (i) and graphs (ii) showing average (n = 4) percentage of apoptotic cells, measured by annexin V/7-AAD reactivity in WT-lacZ-KD, Gfer1-KD, and Gfer2-KD ESCs, at 72 h after cell sorting for lenti-GFP virus expression. Error bars, SD; *p = 0.0002.

Figure 3.

KD of Gfer initiates mitochondrial dysfunction triggering mitophagy in ESCs. (A) Representative histograms depicting TMRE fluorescence (PE channel) in lacZ and Gfer1/2 KD ESCs. Numbers in red (lacZ-KD) and green (Gfer1/2-KD) are average (n = 3) percentage (%) of cells with TMRE fluorescence intensities within the gated region (dotted box) ± SD; *p = 0.0001. Mitochondrial membrane depolarization was assessed by loss of TMRE retention/fluorescence. (B) Histographic representation of the assay measuring the release of cytochrome c in the indicated ESC genotypes. Red (lacZ-KD) and green (Gfer1/2-KD) numbers are average (n = 3) percentage of cells within the gated region (black line) ± SD; *p = 0.001, measuring loss of PE-conjugated anti-cytochrome c antibody staining. (C) Percentage of cells showing positive staining (gated with a black line) for a PE-conjugated active caspase-3 antibody. Results represent average (n = 3) percentage of cells showing positive staining in the PE channel ± SD, *p = 0.01, for lacZ (red) and Gfer1/2-KD (green). (D) Representative immunoblot (n = 3) analyses of active caspase-3 and Bax in indicated ESC genotype. Gfer and actin levels are also shown. (E) Digital TEM images depicting ultrastructural details in lacZ (left) and Gfer-1 KD (right) ESCs. Scale bars, 5 μm at ×2650 and 1 μm at ×7100 and ×25,000 magnifications. Black arrowheads, representative autophagosomes; white arrowheads, degenerating mitochondria.

Figure 4.

Gfer KD does not affect cell viability or mitochondria in differentiated cells. (A) Graphs showing average (n = 3) percentage of apoptotic cells, measured by annexin V/7-AAD reactivity in indicated MEF genotypes, at 72 h after lentivirus infection. Error bars, SD; *p < 0.05. (B) Immunoblots (n = 2) showing Gfer, caspase 3 (18 kDa), Bax, and actin levels in indicated MEFs. (C) Representative histograms depicting TMRE fluorescence (PE channel) in LacZ and Gfer1/2 KD MEFs. Numbers in red (LacZ-KD) and green (Gfer1/2-KD) are the average (n = 3) percentage of cells with TMRE fluorescence intensities within the gated region (dotted box) ± SD; *p < 0.05. (D) Digital TEM images depicting ultrastructural details in LacZ (left) and Gfer-1 KD (right) MEFs. Magnifications, ×2650, ×5600, and ×25,000. (E) Digital TEM images depicting ultrastructural details at ×5600 magnification in LacZ (left) and Gfer-1 KD (right) NIH 3T3 cells.

Figure 5.

Overexpression of Gfer alters mitochondrial morphology and elevates pluripotent marker expression. (A) Representative digital ApoTome images (magnification, ×100) showing DsRed-positive mitochondria in ESCs infected with MSCV-IRES-GFP-control (control), FG12-GFP-Gfer1 shRNA, or MSCV-IRES-GFP-Gfer (MSCV-GFER) viruses. Scale bar, 1 μm; n = 3. (B) Digital TEM images at ×25,000 magnification (scale bar, 0.5 μm) depicting ultrastructural details of mitochondria in control (left) and MSCV-Gfer (both images on the right) ESCs. (C) Average mRNA levels of Nanog, Oct-4, and Gfer in day 6 control and MSCV-Gfer ESCs. The mRNA levels measured by qRT-PCR were normalized to actin; (n = 3); *p = 0.03 for Nanog and 0.006 for Oct4.

Figure 6.

Inhibition of Drp1 rescues mitochondrial dysfunction and apoptosis in Gfer-KD ESCs. (A) Representative immunoblots (n = 3) depicting Drp1, Gfer, and actin levels in control, Gfer-KD, or MSCV-Gfer ESCs. (B) TMRE fluorescence in lacZ-KD ESCs (red), Gfer1-KD ESCs (green), and Gfer1-KD ESCs treated for 24 h with 25 μM mdivi-1(dotted black). Numbers in red (lacZ-KD), green (Gfer1-KD), and black (Gfer1-KD + 25 μM mdivi-1) are average percentage of cells with TMRE fluorescence intensities within the gated region (gray box) ± SD; (n = 3); *p = 0.005. (C) Representative histograms (n = 3) showing percentage of apoptotic cells, in control or MSCV-Gfer ESCs (top panel) and in Gfer-KD ESCs or Gfer-KD ESCs treated for 24 h with vehicle or 25 μM mdivi-1 (bottom panel).

Figure 7.

Expression of Drp1^DN restores pluripotency in Gfer-KD ESCs. (A) Digital ApoTome (optical section) images (magnification, ×630) depicting Nanog, Oct-4, and SSEA1 expression (all in red) in lacZ-KD and Gfer-KD ESCs and in Gfer-KD ESCs expressing Drp1^DN. Blue, DAPI; green, numbers representing average signal intensities ± SD quantified from 35 colonies (n = 3); *p = 0.0001. (B) Day 6 embryoid bodies formed by lacZ-KD and Gfer-KD ESCs and by Gfer-KD ESCs expressing Drp1^DN. Magnification, ×200; scale bar, 2 μm. Average surface area in μm² ± SD from each genotype is shown (n = 3); *p = 0.001.