.
O P E N A C C E S S S O U R C E : Cell Metabolism
Highlights
• The lncRNA Hnscr is highly expressed in htNSCs of young mice but decreases during aging
• Hnscr depletion promotes the senescence of htNSCs and aging-like phenotypes
• Hnscr attenuates htNSC senescence by binding to YB-1 to prevent its degradation
• Theaflavin 3-gallate mimics Hnscr and ameliorates aging-related physiological disorders
Summary
Age-dependent loss of hypothalamic neural stem cells (htNSCs) is important for the pathological consequences of aging; however, it is unclear what drives the senescence of htNSCs. Here, we report that a long non-coding RNA, Hnscr, is abundantly expressed in the htNSCs of young mice but decreases markedly in middle-aged mice. We show that depletion of Hnscr is sufficient to drive the senescence of htNSCs and aging-like phenotypes in mice. Mechanistically, Hnscr binds to Y-box protein 1 (YB-1) to prevent its degradation and thus the attenuation of transcription of the senescence marker gene p16INK4A. Through molecular docking, we discovered that a naturally occurring small compound, theaflavin 3-gallate, can mimic the activity of Hnscr. Treatment of middle-aged mice with theaflavin 3-gallate reduced the senescence of htNSCs while improving aging-associated pathology. These results point to a mediator of the aging process and one that can be pharmacologically targeted to improve aging-related outcomes.
Context and Significance
Understanding the mechanisms of aging is important for developing strategies to combat the many chronic comorbidities associated with it. One of the causes of aging is the exhaustion of stem cells, including hypothalamic neural stem cells (htNSCs), because of their senescence. Here, Xiang-Hang Luo and colleagues investigate the role of a long non-coding RNA, termed Hnscr, which is highly expressed in htNSCs but decreases markedly with age. They found that Hnscr depletion is sufficient to cause senescence of these stem cells and aging-like phenotypes in mice. They also identified a natural small compound, theaflavin 3-gallate, which a key component of black tea, that mimics the action of Hnscr, and its use in mice greatly reduced the senescence of htNSCs while preventing aging-related physiological disorders.
Introduction
The hypothalamus, which has crucial functions in neuroendocrine regulation, communicates with peripheral tissues and responds to environmental and nutritional cues to modulate various aspects of physiological functions. Recently, it was discovered that the hypothalamus plays critical roles in supervising the process of aging (Zhang et al., 2013, Zhang et al., 2017), and hypothalamic neural stem/progenitor cells (htNSCs) mediate this process (Zhang et al., 2017). These cells exist in the hypothalamic third-ventricle wall and mediobasal hypothalamic (MBH) parenchyma (Li et al., 2012) and undergo age-dependent loss, becoming almost entirely absent in advanced aging (Zhang et al., 2017). It is critical to understand the mechanisms that govern the loss of htNSCs to develop strategies to potentially improve the outcomes of comorbidities that are associated with aging. Here, we identified a long noncoding RNA (lncRNA), Hnscr, that is abundantly present in the htNSCs of young mice but is substantially diminished in the htNSCs of middle-aged mice.
Long non-coding RNAs (LncRNAs) are classified as transcripts longer than 200 nucleotides that have limited potential to encode proteins (Akerman et al., 2017, Knoll et al., 2015, Zhao et al., 2014). LncRNAs regulate the expression of target genes at the transcriptional level (Arab et al., 2014), post-transcriptional level (Guo et al., 2016), and post-translational level (Liu et al., 2015) by binding with DNA, RNA, or protein complexes or by serving as the precursors of miRNAs (Lu et al., 2017). LncRNAs regulate a wide range of biological processes via diverse mechanisms (Guttman and Rinn, 2012, Ransohoff et al., 2018, Wong et al., 2018). Thus, stemming from our screening information, we investigated the relationship between Hnscr and htNSCs, revealing that senescence of these cells is controlled by this lncRNA, and, thus, loss of Hnscr contributes to aging-associated pathologies.
Results
Hnscr Levels Decrease in htNSCs of Mice during Aging
To assess the different characteristics of htNSCs from young and middle-aged mice, we analyzed hypothalamic tissue from 3- and 18-month-old C57BL/6J mice for neurosphere formation (Li et al., 2012, Zhang et al., 2017). Neurospheres derived from the hypothalamus of middle-aged mice were much fewer, smaller, and poorly grown than those derived from young mice (Figures 1A–1D ). Given the important roles of lncRNAs in regulating many aspects of cell biology, including survival, we were interested in exploring the patterns of lncRNAs in these cells under different age conditions. Indeed, htNSCs abundantly contain miRNAs and RNAs of larger sizes (Zhang et al., 2017). We performed RNA sequencing to identify differences in lncRNAs between htNSCs derived from young mice versus those derived from aged mice (Figure 1E). A total of 1,122 differentially expressed lncRNAs with at least a 2.5-fold change were identified. We selected 5 lncRNAs with the following properties: (1) showed high expression in htNSCs that strikingly decreased during aging; (2) showed homology and conservation between human and mice; and (3) had no splicing or overlap with any coding genes present in the National Center for Biotechnology Information’s RefSeq database, University of California Santa Cruz Genome Browser, or Ensembl genome browser annotations. Among these differentially expressed lncRNAs, Gm31629 showed the best evolutionary conservation, which had several orthologous genomic regions of greater than 200 base pairs (bp) and showed similar locations in the genome between human and mice, and therefore was selected for further study.
Figure 1Expression of Hnscr Decreases in htNSCs of Mice during Aging
(A–D) Representative images (n = 18 photographs from 3 experiments) of neurospheres generated from the hypothalami of young mice (3 months old) and middle-aged mice (18 months old)
(A). Scale bars, 100 μm. Their quantitation is shown in (B); relative size in ©; and cell output over 5 passages in (D).
(E) Heatmap of RNA-seq profiling of lncRNA expression in htNSCs from young mice (3 months) and middle-aged mice (18 months). Fold change > 2.5, false-discovery rate < 0.20.
(F) Relative Hnscr expression in htNSCs from 3-month-old and 18-month-old mice as determined by qRT-PCR. The value of the expression in the 3-month-old mice was set at an arbitrary value = 1.
(G) Age-associated changes of Hnscr expression in the htNSCs of WT mice. The value of the expression in the 3-month-old mice was set at an arbitrary value = 1.
Data are shown as mean ± SEM, (n = 6–7 in [A]–[D] and [F]–[G]; n = 3 in [E]). ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001 by two-tailed Student’s t test or one-way ANOVA.
The decreased expression of Gm31629 in the htNSCs of aged mice was further confirmed by quantitative real-time PCR analysis (Figure 1F). We measured the expression pattern of Gm31629 in the cerebral cortex, hypothalamus, htNSCs, liver, muscle, white adipose tissue, brown adipose tissue, kidney, and heart, finding that Gm31629 expression was particularly high in htNSCs and the hypothalamus (Figure S1). We thus referred to Gm31629 as Hnscr, given that it is htNSC-related. In addition, we found that the levels of Hnscr were consistently negatively correlated with increasing age (Figure 1G).
Depletion of Hnscr induces aging-like phenotypes
These above data led us to explore whether Hnscr might play an important role in the loss of htNSCs and the aging process. To evaluate whether Hnscr could be physiologically relevant to aging, we generated Hnscr null mice via gene targeting with the TetraOne technique. When young, the Hnscr null mice and their littermate controls were indistinguishable; for example, they were similar in body weight (Figure S2A) and in physiology, such as in behavior and reproductive fitness (Figures S2B–S2K). Then, we followed up the effects of Hnscr knockout on whole-body physiology in middle-aged mice through behavioral analyses. Hnscr null mice exhibited lower muscle endurance performance, coordination, treadmill running, novel object recognition, and sociality function (Figures 2A–2E ). Further, Morris water-maze analyses demonstrated that the Hnscr null mice had a larger decline in cognitive ability than the control mice had (Figure 2F).
Figure 2Hnscr Knockout Increases the Senescence of htNSCs and Aging
(A–F) The physiological changes including muscle endurance performance (A), coordination (B), treadmill running ©, novel object recognition (D), and sociality function (E) of 18-month-old Hnscr null mice and littermate WT mice.
(G and H) Representative micro-CT images (n = 12 photographs from 2 experiments) (G) and quantitative analysis (H) of femur of 18-month-old Hnscr null mice and littermate WT mice. Tb. BV/TV, trabecular bone volume per tissue volume.
(I) Lifespan of Hnscr null mice and littermate WT mice.
(J) Relative Hnscr expression in htNSCs of 18-month-old Hnscr null mice and littermate WT mice. The value of the expression in the WT mice was set at an arbitrary value = 1.
(K–N) Representative images (n = 21 photographs from 3 experiments) of neurospheres generated from the hypothalami of 18-month-old Hnscr null mice and littermate WT mice (K). Scale bar, 100 μm. Their quantitation is shown in (L); relative size in (M); and the cell output over 5 passages in (N).
(O and P) Representative images (n = 18 photographs from 3 experiments) of Tuj1 immunostaining of dissociated htNSCs from 18-month-old Hnscr null mice and littermate WT mice (O). Scale bar, 100 μm. The percentage of Tuj1+ cells is shown in (P).
(Q and R) Representative images (n = 21 photographs from 3 experiments) of SA-βGal staining of dissociated htNSCs from 18-month-old Hnscr null mice and littermate WT mice (Q). Scale bar, 100 μm. The percentage of senescent cells is shown in ®.
(S and T) Representative images (n = 14 photographs from 2 experiments) of hypothalamic sections of 18-month-old Hnscr null mice and littermate WT mice co-stained with SOX2 and BMI1
(S). Scale bar, 100 μm. The relative percentage of SOX2+BMI1+ cells is shown in (T).
Data are shown as mean ± SEM, (n = 10 in [A]–[F], n = 29–30 in [I], and n = 6–7 in [G]–[H] and [J]–[T]). ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001 by two-tailed Student’s t test or one-way ANOVA.
Lifespan analysis was performed by Kaplan–Meier survival analysis.
Metabolically, glucose-tolerance tests (GTTs) and insulin-tolerance tests (ITTs) showed that Hnscr null mice had impaired glucose tolerance (Figure S3A), although insulin resistance was similar between the null mice and controls (Figure S3B). Reproductive fitness was also examined, showing that the weight of testis and sperm count and sperm motility of Hnscr null mice were lower than those of the controls (Figures S3C–S3E). Micro-computed tomography (CT) analysis showed that aging-related bone-mass loss was more evident in the Hnscr null mice than in the controls (Figures 2G and 2H). Consistent with all these observations, the lifespan of Hnscr null mice decreased significantly in comparison with wild-type control mice (Figure 2I), further supporting the observation that Hnscr null mice developed an aging-like phenotype.
Depletion of Hnscr Promotes htNSC Senescence at Middle Age
In the meanwhile, we analyzed the ability to derive neurospheres from the hypothalami of Hnscr null mice. The neurospheres from middle-aged Hnscr null mice were much fewer and smaller, whereas htNSCs in these neurospheres were poorer in both proliferation and differentiation compared with age-matched controls (Figures 2J–2P). Senescence-associated β-galactosidase (SA-βGal) staining further revealed that the percentage of senescent cells were higher in htNSCs derived from Hnscr null mice at middle age than in htNSCs from age-matched control mice (Figures 2Q and 2R). Consistently, Hnscr null mice displayed an accelerated rate of age-dependent loss of htNSCs compared to the control mice (Figures 2S and 2T). We also examined the number of neurons in the MBH and subtypes, including pro-opiomelanocortin (POMC) neurons and neuropeptide Y (NPY) neurons, but did not find a major difference between Hnscr null mice and control mice (Figures S3F–S3I). This result agreed with our observation that loss of Hnscr had minimal effects when animals were young, because the function of these cells for hypothalamic neurogenesis is more relevant to the developmental stage and young condition rather than aged condition. Taken together, loss of Hnscr leads to an induction of htNSC senescence during middle age.
Depletion of Hnscr in htNSCs Leads to Senescence and Aging-like Disorders
To more directly explore the role of Hnscr in affecting htNSCs senescence in vivo, we generated a mouse model with Hnscr specifically depleted in htNSCs. The 14-month-old wild-type (WT) mice were obtained by bilaterally intra-MBH injection of Sox2 promoter-driven adeno-associated viruses (AAVs) expressing microRNA miR30-based short hairpin RNAs (shRNAs) that target Hnscr (AAV-Sox2-shHnscr), as illustrated in Figure 3A. One month later, the hypothalamic tissue of mice was isolated for neurosphere culture. Quantitative real-time reverse transcription PCR (qRT-PCR) analysis confirmed that the expression of Hnscr was largely abrogated in the htNSCs of mice injected with AAV-Sox2-shHnscr (Figure 3B). Based on neurosphere assays, cell-proliferation-rate analysis, cell-differentiation assays, and SA-βGal staining, htNSCs derived from AAV-Sox2-shHnscr-injected mice were more prone to senescence than those derived from the control mice (Figures 3C–3J). Consistent with the senescence, the number of htNSCs were lower in the hypothalamus of AAV-Sox2-shHnscr-injected mice than those in the control mice (Figures 3K and 3L). Four months after injection of the AAVs, these mice were subjected to various physiological and histological analyses that collectively indicate aging; to summarize, the AAV-Sox2-shHnscr-injected mice exhibited impaired muscle functions, cognition, sociality, metabolism, and reproductive fitness (Figures 3M–3R and S4A–S4E), although these phenotypes were generally weaker than whole-body knockout of Hnscr presented in Figure 2. Injection of AAV-Sox2-shHnscr did not affect the number of hypothalamic neurons (Figures S4F–S4I). Taken together, these results suggest that disruption of Hnscr expression promotes the senescence of htNSCs and contribute to accelerated aging-related physiological decline in mice.
Edited by Engadin, 09 July 2020 - 09:35 PM.
