Figure 1.
Bmi1 is down-regulated in the aging mouse brain.
A) Coronal sections from the cerebral cortex were analyzed by IHC using Bmi1 and NeuN (top panels), or Bmi1 and GFAP (lower panels) antibodies. Bmi1 (brown) is highly expressed in NeuN+ (pink) neurons and its expression decreases with age, while it remains unchanged in GFAP+ (Pink) astrocytes. Note that in aged neurons, Bmi1 labeling was not uniform, with some neurons expressing Bmi1 at moderate levels (blue arrowheads), while others showing nearly undetectable level (red arrowheads). Scale bars; 20 µm. (B) The relative expression of Bmi1 in cortices from young and old brains was analyzed by Q-PCR. Each point represents a comparison between one old and one young mouse cortex. The blue line represents the mean, and the red line represents the standard deviation (n = 15; **P<1.022E−24). (C)Young (40–55 days old) and old (21–26 months old) mice brain samples were analyzed by Western blot for Bmi1 expression. Protein loading was normalized using β-actin and α-tubulin. Data in brackets are levels of Bmi1 and are expressed as Mean ± s.d. (n = 4 brains per group; *P<0.05). (D)Thenumbers of neurons (NeuN+ cells) and non-neuronal cells (NeuN- cells) were counted in young and old mice brain slices. (left panel) Data were presented as absolute neuron number per cortical field. (right panel) Data were expressed as the percentage of neurons versus all cortical cells (neurons + non-neuronal cells). Results are Mean +/− s.d. (n = 3 brains per age, and counts were made on 4 to 9 slices per brain; P = 0.18).
Figure 2.
Antioxidant defenses are reduced in the aging mouse brain.
(A) The relative expression of senescence-associated genes in cortices from young and old brains was analyzed by Q-PCR. Results are Mean ± s.d. (n = 3; *P<0.05; **P<0.01). The dashed line represents the basal gene expression level measured in young mice. (B) The relative expression of antioxidant genes in cortices from young and old brains, and from P25 Bmi1−/− and WT mice was analyzed by Q-PCR. The dashed line represents the basal gene expression level measured in young compared to old and to WT compared to Bmi1−/− mice. Results are Mean ± s.d. (n = 3; *P<0.05; **P<0.01). (C) Coronal sections from the cerebral cortex of young and old mice, and of P25 WT and Bmi1−/− mice were labeled with antibodies against 8-oxo-guanine (8-OG; brown) and GFAP (pink). Note the increase in 8-oxo-guanine labeling in neurons from old and Bmi1−/− mice compared to respective controls. Scale bars; 50 µm. (D) ChIP analysis of young and old brains revealing accumulation of p53 and heterochromatin marks (histone H3 K27me2 or H3 k9me2) at the xCT, Sod1 and Sod2 promoters in old brains. Antibodies against acetylated histone H4 and IgG were used as control. The β-major promoter region of globin was use as negative control. Results are Mean ± s.d. (n = 3; *P<0.05).
Figure 3.
BMI1 is down-regulated in the aging human brain and retina.
(A) Immunohistochemistry on human brain (hippocampus) sections using anti-Bmi1 (brown) and anti-GFAP (pink) antibodies. BMI1 is expressed in neurons, but not in GFAP+ astrocytes, and expression is highly reduced in old brain neurons. Note the virtual absence of BMI1 labeling in some neurons (red arrowheads). Scale bars; 20 µm. (B) Immunofluorescence analysis of BMI1 expression in the human retina (23 years old, frozen sections). BMI1 is highly expressed in human photoreceptors (white arrowheads), which cell body lies in the outer nuclear layer (ONL), while its expression is weaker in neurons of the inner nuclear (INL) and ganglion cell (GCL) layers (red arrowheads). Scale bars; 20 µm. (C) Human retina samples were analyzed by Western blot for BMI1 expression and protein content was normalized using β-actin. BMI1 protein levels are reduced in old retinas (65–75 years). Results are Mean ± s.d. (n = 2–5 retinas per group; *P<0.05). (D)Immunofluorescence analysis of GFAP and P16INK4A expression in young and old human retinas. Note increased GFAP and P16INK4A immunoreactivity in the old retinas. Scale bars; 20 µm.
Figure 4.
BMI1 is highly neuroprotective against topoisomerase I inhibition and mitochondrial poisoning.
(A) Empty plasmid vectors (CMV-GFP) or human BMI1-carrying plasmid (CMV-GFP: BMI1) were transfected in 293FT cells and lysates were analyzed 72 hours later for Bmi1 expression by Western blot. β-actin was used as internal control for normalization of protein loading. Non-transfected cells were used as control (Ctl) for endogenous Bmi1 expression. (B) Experimental scheme showing the procedure used to electroporate plasmid vectors in primary neuronal cultures from e18.5 WT mouse embryo cortices. (C) After 7 days in vitro (DIV), electroporated neurons were exposed to CA, 3-NP or their respective vehicles. 16 hours later, cultures were stained for apoptosis induction (caspase-3 in brown) and expression of GFP (in pink), in order to distinguish neurons carrying or not the transgene. (D) Cell viability was assessed in cultures photographed in (C) as the percentage of GFP+/Caspase-3− cells versus total GFP+ cells. Results are Mean ± s.d. (n = 3; *P<0.05; **P<0.001). (E) After 7 DIV, electroporated WT and p53−/− neurons were exposed to CA or vehicle (DMSO) and analyzed after 16 hours as described in (C). Results are Mean ± s.d. (n = 3; **P<0.001).
Figure 5.
BMI1 over-expression activates AOR genes and suppresses ROS.
(A and B) e18.5 WT neurons were infected with BMI1-expressing (Adeno-BMI1) or GFP-expressing (Adeno-GFP) adenoviruses. 48 hours post-infection, neurons were treated or not with 8 mM of 3-NP and analyzed 6 hours later. (A) ROS levels were measured with the fluorescent dye DCFDA. Data were normalized to the protein contents and expressed as fold change relative to levels in Adeno-GFP-infected cells. Results are Mean ± s.d. (n = 3 independent cultures; *P<0.05; **P<0.01). (B) relative gene expression was analyzed by Real-Time PCR. Data are normalized to levels found in adeno-GFP-infected neurons that were set at 1 (dashed line). Results are Mean ± s.d. (n = 3 independent cultures; *P<0.05; **P<0.01). (C) e18.5 WT or p53−/− neurons were infected with adenoviruses and analyzed for ROS levels as in (A). Data were expressed as fold change relative to levels in Adeno-GFP-infected WT neurons. Results are Mean ± s.d. (n = 3 independent cultures; *P<0.05).
Figure 6.
Bmi1 deficiency during aging influences neurons resistance to genotoxic stresses and mitochondrial dysfunctions.
Proposed model of Bmi1 function in neurons: (A) When over-expressed, Bmi1 represses p53 activity by an unknown mechanism, leading to complete inhibition of p53 pro-apoptotic and pro-oxidant activities and supra-activation of the antioxidant defense system. (B) In young neurons, where Bmi1 expression is robust, Bmi1 partially represses p53 activity, thus allowing modulation of p53-mediated apoptosis and repression of antioxidant response elements (ARE). These elements are present in antioxidant-coding genes activated by the Nrf2 transcription factor. (C) In aging neurons, where Bmi1 expression becomes deficient, p53 is activated (1), leading to induction of apoptosis and inflammation, and in transcriptional repression of antioxidant-coding genes (2). Elevated mitochondrial reactive oxygen species (mROS) concentrations ultimately induce damages to lipids and DNA, which further activate p53 (3), resulting in the formation of a vicious circle. This situation renders old neurons particularly more vulnerable to genotoxic stresses (gs) and mitochondrial dysfunctions. This model is based on data from the present work, and those published previously [20].