.
F U L L T E X T S O U R C E : Cell
Highlights
• Aged OPCs fail to respond to differentiation signals
• Aged OPCs acquire many of the hallmarks of cell aging
• Fasting and the fasting mimetic metformin rejuvenate poor remyelination in aged rodents
• Metformin reverses age-related changes, making OPCs respond to differentiation factors
Summary
The age-related failure to produce oligodendrocytes from oligodendrocyte progenitor cells (OPCs) is associated with irreversible neurodegeneration in multiple sclerosis (MS). Consequently, regenerative approaches have significant potential for treating chronic demyelinating diseases. Here, we show that the differentiation potential of adult rodent OPCs decreases with age. Aged OPCs become unresponsive to pro-differentiation signals, suggesting intrinsic constraints on therapeutic approaches aimed at enhancing OPC differentiation. This decline in functional capacity is associated with hallmarks of cellular aging, including decreased metabolic function and increased DNA damage. Fasting or treatment with metformin can reverse these changes and restore the regenerative capacity of aged OPCs, improving remyelination in aged animals following focal demyelination. Aged OPCs treated with metformin regain responsiveness to pro-differentiation signals, suggesting synergistic effects of rejuvenation and pro-differentiation therapies. These findings provide insight into aging-associated remyelination failure and suggest therapeutic interventions for reversing such declines in chronic disease.
Introduction
The ability to regenerate oligodendrocytes, the myelin-forming cells of the CNS, contrasts with the poor capacity to regenerate neurons in most brain regions (Franklin and Ffrench-Constant, 2017). Generation of oligodendrocytes from oligodendrocyte progenitor cells (OPCs) occurs throughout life and contributes to myelin turnover (Young et al., 2013, Hill et al., 2018, Hughes et al., 2018, Tripathi et al., 2017) and adaptive myelination (Gibson et al., 2014, Mitew et al., 2018, Hughes et al., 2018) as well as to the regenerative process of remyelination that follows demyelination (awadzka et al., 201). As with most regenerative processes, the efficiency of remyelination declines progressively with aging to the extent that it becomes so slow that it eventually fails (Oh et al., 2014, Sim et al., 2002). This has important implications for chronic demyelinating diseases such as multiple sclerosis (MS) that can extend over several decades. Delayed remyelination renders demyelinated axons susceptible to irreversible degeneration, a phenomenon that underpins the progressive neurological decline associated with the later stages of MS (Franklin et al., 2012).
The slowing of remyelination with aging is characterized by impaired recruitment of OPCs into the lesion area as well as their delayed differentiation into oligodendrocytes (Sim et al., 2002). Chronically demyelinated MS lesions either lack OPCs or, more commonly, contain OPCs that have failed to differentiate (Boyd et al., 2013, Chang et al., 2002, Kuhlmann et al., 2008). Increasing the number of OPCs in white matter lesions of aged animals does not improve remyelination (Woodruff et al., 2004), indicating that differentiation of OPCs into oligodendrocytes is the bottleneck for remyelination. The mechanisms that regulate OPC differentiation are dysregulated in the aging brain (Shen et al., 2008), in part because of age-related changes in the cells and molecules in the environment in which remyelination occurs (Cantuti-Castelvetri et al., 2018, Hinks and Franklin, 2000, Natrajan et al., 2015). These changes can be overcome, in principle, by providing a more youthful systemic environment that is permissive for regeneration (Ruckh et al., 2012). Thus, remyelination can potentially be enhanced by pro-differentiation factors lacking in the aged brain. However, it remains unclear whether OPCs undergo intrinsic changes with aging that affect their responsiveness to differentiation signals.
Results
Aged OPCs Differentiate Slowly and Do Not React to Pro-differentiation Compounds
We first asked whether age-related changes in OPCs contribute to the differentiation delay observed in aged animals during remyelination (Sim et al., 2002). Studies of OPC aging have been hampered by the technical challenges of culturing OPCs isolated from the aged adult rodent CNS. Thus, we first optimized existing protocols to establish cultures of adult OPCs from young adult (2–3 months) and aged (20–24 months) rats (Figure S1; Dugas and Emery, 2013). We used magnetic activated cell sorting (MACS) for A2B5 to isolate cells from young and aged adult brains (Figure S1A). The positively selected cells showed minor contamination for CD11b+ microglia (<0.8%) or MOG+ oligodendrocytes (<2%) with no significant difference between preparations from young and aged animals (Figures S1B–S1D). Using immunohistochemistry, we found that A2B5+ cells of both age groups co-expressed PDGFRa, NG2, Sox10, and Olig2, confirming their identity as OPCs (Figures S1E–S1I and S1N–S1R). In contrast, A2B5+ cells from young or aged adults never expressed mature lineage markers such as CNPase and MBP (Figures S1J and S1K), the microglia marker Cd11b (Figure S1L), or the astrocyte marker GFAP (Figure S1M), indicating that cultures were free of mature oligodendrocytes, microglia, and astrocytes. This enabled us to compare the differentiation efficiency of OPCs isolated from the young adult and aged CNS.
We next tested the relative differentiation ability of OPCs from young adults (hereafter referred to as young OPCs) and OPCs from old adults (aged OPCs) when grown in differentiation medium from which the growth factors that maintain proliferation were removed. Although 60% of young OPCs differentiated into mature oligodendrocytes (CNPase+ and MBP+), fewer than 20% of aged OPCs acquired these markers within the same period, revealing a slower inherent capacity for differentiation (Figures 1A–1C). We next assessed how adult OPCs responded to thyroid hormone (T3), a well-established promoter of OPC differentiation (Gao et al., 1998). Although T3 accelerated the differentiation of young OPCs, there was no significant effect on the differentiation of aged OPCs (Figures 1B and 1C). Similar results were obtained with other factors known to have pro-differentiation effects on OPCs derived from newborn animals or on pluripotent stem cells, such as 9-cis-retinoic acid (Huang et al., 2011), miconazole (Najm et al., 2015), and benzatropine (Deshmukh et al., 2013), all of which enhanced differentiation in young adult (Figures 1E and 1F) but not aged OPCs (Figures 1E, 1G, and S4).
Figure 1OPCs Lose Their Inherent Capacity for Differentiation and Their Responsiveness to Differentiation Factors with Aging
(A) Representative images of young adult (2–3 months old) and aged OPCs (20–24 months old) differentiated in the absence of growth factor or in the presence of T3. Increasing maturity was visualized using O4 (early), CNPase (intermediate), and MBP (mature), immunocytochemical markers of the oligodendrocyte (OL) lineage. Scale bars, 50 μm.
(B and C) Quantification of cells over time in culture: CNPase+/Olig2+ cells (B) and MBP+/Olig2+ cells ©. Statistical significance was determined using two-way ANOVA repeated measurements followed by Dunnett’s post test to compare each group against “aged T3.” All data are presented as mean ± SD (n = 3 biological repeats).
(D) Schematic of the experimental design.
(E) Representative images of the differentiation assay performed with young and aged OPCs. Newly formed oligodendrocytes were identified as MBP+/Olig2+ cells. Scale bars, 50 μm.
(F and G) Quantification of the differentiation assay for young (F) and aged OPCs (G). n = 3 biological replicates for each group, one-way ANOVA with Dunnett’s multiple comparisons test for each group against the group differentiating in the absence of growth factors (“w/o GF”).
- Error bars represent SDs. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.
- See also Figures S1–S3.
Cultures of aged OPCs were generally of lower cell density, suggesting decreased survival. Because the density might influence the differentiation rate through secretion of paracrine factors, we asked whether increasing the cellular density might rescue the differentiation ability of aged OPCs or reduce the differentiation rate of young OPCs. Seeding half of the normal number of young OPCs or twice the number of aged OPCs did not result in a significant change in differentiation rates (Figures S2A–S2E), indicating that the differences in cellular density between young and aged cultures do not account for the reduced differentiation of aged OPCs.
Because some aged OPCs differentiated into MBP+ oligodendrocytes, we asked whether the non-differentiating cells were unable to differentiate or were differentiating at a slower rate. To test this, we cultured aged OPCs for 4 weeks in differentiation medium and found a significant increase in differentiation (Figures S2D and S2E), indicating that aged OPCs do not lose their general ability for differentiation but undergo intrinsic changes that significantly slow their differentiation program. These intrinsic changes also cause aging OPCs to become less responsive to factors that induce differentiation, which likely contributes to the failure of oligodendrocyte lineage differentiation, characteristic of many non-remyelinating chronic MS lesions (
Next we characterized molecular alterations responsible for the aged OPC phenotype using RNA sequencing (RNA-seq) to compare the transcriptomes of OPCs isolated from young and aged rats. Approximately 20% of all genes were differentially expressed with aging (1.5-fold change in expression, adjusted p value [p.adj] < 0.05). Among the genes more highly expressed in young adult OPCs were those that are characteristic of adult OPCs in 21-day-old mice (Marques et al., 2016; Figure 2A) and their self-renewal, including Pdgfra, Ascl1, and Ptprz1 (Emery, 2010; Figure 2B; .Table S1). In contrast, aged OPCs expressed higher levels of the early differentiation markers Cnp1, Sirt2, and Enpp6 (Figure 2B). Because we did not find a higher proportion of MOG+ cells or those expressing more mature lineage markers, such as CNPase, in our aged OPC preparations compared with young OPCs (Figures S1J and S1K), we ruled out the possibility that these changes in the transcriptome were caused by contamination with oligodendrocytes. Thus, we concluded that aged OPCs lose their characteristic stem cell signature (Figures 2A and 2B). To identify the cellular processes that might contribute to the aged OPC state, we used ingenuity pathway analysis on genes preferentially expressed in aged OPCs. We found enrichment of terms that are closely linked to organismal and stem cell aging, such as mitochondrial dysfunction, unfolded protein response (UPR), autophagy, inflammasome signaling, and nuclear factor κB (NF-κB and p38 mitogen-activated protein kinase (MAPK) signaling (Figure 2C). Consistent with the predictions made on the basis of the RNA-seq data, we found increased mTOR activity in freshly isolated aged OPCs by detection of the phosphorylated forms of the downstream target p70S6-kinase (Figure 2D). mTOR activity is a crucial regulator of adult stem cell quiescence, activation, and differentiation (Mihaylova et al., 2014, Rodgers et al., 2014) and is linked to cellular aging (Laplante and Sabatini, 2012). Aging is associated with increased and dysregulated mTOR activity, which contributes to DNA damage and cellular senescence (Castilho et al., 2009, Chen et al., 2009, Yilmaz et al., 2006). We therefore predicted that both DNA damage and markers of senescence would increase with adult OPC aging. Consistent with this prediction, single-cell comet assays revealed that aged OPCs had significantly more DNA damage than young OPCs (Figures 2E and 2F). Using our RNA-seq data, we also found that aged OPCs expressed several genes associated with cellular senescence at significantly higher levels than young OPCs (Figure 2G; Tacutu et al., 2018). We found that aged OPCs had 8-fold higher mRNA levels of the senescence marker Cdkn2a (Figure 2H). Last, aged OPCs had lower levels of ATP and reduced cellular respiration (Figures 2I and 2J), likely reflecting a combination of mitochondrial dysfunction and reduced mitochondrial content. Thus, aged OPCs, like other adult stem cells, acquire a variety of hallmarks of aging that likely contribute to loss of their regenerative potential.
Figure 2Aged OPCs Have Reduced Expression of OPC-Specific Genes and Acquire Hallmarks of Aging
(A) Young and aged OPCs were tested for differential expression of OPC-specific genes. The pie chart summarizes the findings as the percentage of genes that were expressed at significantly higher levels in aged or young OPCs (p.adj < 0.05) or that were not differentially expressed (p.adj > 0.05). See also Table S1.
(B) qRT-PCR validation of several genes identified in RNA-seq, comparing freshly isolated young and aged OPCs (n = 3 biological replicates for each age group, two-tailed t test).
© Top 5 pathways identified by ingenuity pathway analysis (Z score > 2 and p.adj. < 0.05) for genes enriched in aged OPCs (p.adj < 0.05; see also Table S2).
(D) Western blot for the downstream mTORC1 pathway target p70S6K and actin loading controls. P, phosphorylated; n = 2 biological samples for each age group.
(E) Representative images for comet assays (alkaline conditions) of freshly isolated young and aged OPCs to visualize the degree of DNA damage. Presence of a tail indicates DNA damage.
(F) Quantification of the comet assay. The categories used for scoring are depicted in the respective boxes. Statistical significance was determined using one-way ANOVA and Turkey’s post test. All data are presented as mean ± SD (n = 3 biological replicates for each age group).
(G) Heatmap of genes from RNA-seq data whose expression is associated with cellular senescence. All depicted genes are differentially expressed (n = 3 biological repeats).
(H) qRT-PCR results visualizing expression of the senescence marker Cdkn2a. Data are presented as mean ± SD (n = 3 biological replicates for each age group, two-tailed t test).
(I) Fold change of the basal oxygen consumption rate (bOCR), measured after overnight culture in vitro (n = 3 biological repeats for each age group, two-tailed t test).
(J) Normalized intracellular ATP content of freshly isolated OPCs (n = 5 biological repeats for each age group, two-tailed t test).
- Error bars represent SDs. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.
.../...
.
