SEARCH

SEARCH BY CITATION

Keywords:

  • aging;
  • cellular senescence;
  • glycogenesis;
  • glycogen synthase;
  • GSK3;
  • reactive oxygen species

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  9. Supporting Information

Glycogen biogenesis and its response to physiological stimuli have often been implicated in age-related diseases. However, their direct relationships to cell senescence and aging have not been clearly elucidated. Here, we report the central involvement of enhanced glycogenesis in cellular senescence. Glycogen accumulation, glycogen synthase (GS) activation, and glycogen synthase kinase 3 (GSK3) inactivation commonly occurred in diverse cellular senescence models, including the liver tissues of aging F344 rats. Subcytotoxic concentrations of GSK3 inhibitors (SB415286 and LiCl) were sufficient to induce cellular senescence with increased glycogenesis. Interestingly, the SB415286-induced glycogenesis was irreversible, as were increased levels of reactive oxygen species and gain of senescence phenotypes. Blocking GSK3 activity using siRNA or dominant negative mutant (GSK3β-K85A) also effectively induced senescence phenotypes, and GS knock-down significantly attenuated the stress-induced senescence phenotypes. Taken together, these results clearly demonstrate that augmented glycogenesis is not only common, but is also directly linked to cellular senescence and aging, suggesting GSK3 and GS as novel modulators of senescence, and providing new insight into the metabolic backgrounds of aging and aging-related pathogenesis.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  9. Supporting Information

Most mammalian cells store glycogen as a glucose reservoir because of its rapid conversion into glucose and its potential utilization as a source of energy even in the absence of oxygen (Harris, 2006). Glycogenesis, the biosynthesis of glycogen, occurs in almost every tissue, but particularly in muscle and liver. The priming step in glycogenesis is autocatalytic attachment of glucose to a single tyrosine residue of the enzyme glycogenin, using UDP-glucose as a glucosyl donor. Subsequently, glycogenin extends the glucan chain by linking additional six or seven glucose residues in α-1,4-glycosidic bond. This primed glucan is further elongated by glycogen synthase (GS) up to 60 000 glucose units, and branches arise from frequent α-1,6 glycosidic linkage with the help of branching enzyme (Bollen et al., 1998). Thus, effective stimulation of glycogenesis in response to postprandial glycogenic stimuli (such as glucose, insulin, and parasympathetic nerve impulses), necessitates well-controlled activation of GS through dephosphorylation; this is accomplished by inhibiting its key regulatory kinase, GS kinase 3 (GSK3) (Pagliassotti & Cherrington, 1992; Pagliassotti et al., 1996). The negative modulation of GSK3 on GS activity has been extensively described (Embi et al., 1980; Cohen & Frame, 2001).

Although GS kinase-3 (GSK3) was initially identified as a key regulator of insulin-dependent glycogen synthesis, it is currently well-known as a multifunctional kinase that performs a regulatory role in several cellular functions, including embryonic development, cellular protein synthesis, mitosis, and survival (Embi et al., 1980; Rylatt et al., 1980; Forde & Dale, 2007). Inappropriate modulation of GSK3 activity has recently been implicated in the progression of multiple age-related pathologies such as Alzheimer's disease, noninsulin-dependent diabetes mellitus (NIDDM), inflammation, and cancer (Ban et al., 2003; Eldar-Finkelman & Ilouz, 2003; Shakoori et al., 2005; Takashima, 2006; Jope et al., 2007). Both positive and negative evidence exist for modulation of cellular growth by GSK3. The GSK3 inhibition in Wnt signaling was shown to be involved in cell proliferation and cancer development through β-catenin stabilization (Shakoori et al., 2005; Huang et al., 2007). On the other hand, an essential role of active GSK3 in cell survival was also implied by the death of GSK3β-null mice during embryogenesis, as a result of liver degeneration caused by widespread hepatocyte apoptosis (Hoeflich et al., 2000). This was further supported by the studies that GSK3 protected from TNF-induced cytotoxicity through phosphorylation of NF-κB subunit p65 (Schwabe & Brenner, 2002), and its inhibition potentiated tumor necrosis factor-alpha-related apoptosis-inducing ligand (TRAIL)-induced apoptosis (Liao et al., 2003; Rottmann et al., 2005). The latter evidence allows a possibility that inhibitory phosphorylation of GSK3 may participate in senescence-associated growth suppression. Among several signal cascades to phosphorylate GSK3 (Frame & Cohen, 2001; Forde & Dale, 2007), insulin or insulin-like growth factor-1 (IGF-1)-related pathway is the key regulatory signal to inactivate GSK3 (Cohen & Frame, 2001). Moreover, IGF-1RPI3K/AKT signaling has recently been shown to regulate cellular or organismal lifespan (Kimura et al., 1997; Tissenbaum & Ruvkun, 1998; Bartke, 2001; Tater et al., 2001; Holzenberger et al., 2003), ultimately suggesting the potential involvement of GSK3 inactivation in senescence. However, there is no clear evidence of whether GSK3 is directly involved in cellular senescence or organismal aging.

Failure to actuate cell division in response to a mitotic stimulus is the most characteristic feature of senescent cells (Hayflick, 1965; Campisi & d’Adda di Fagagna, 2007). Nevertheless, the cells remain metabolically active, but in an altered state (Matsumura et al., 1979). Detailed exploration of the metabolic changes critically involved in cellular senescence is quite important to understand the underlying mechanisms of age-associated metabolic diseases. In addition to the potential association of GSK3 inactivation in senescence-associated growth suppression, its central role in activation of glycogenesis prompted us to ask whether glycogenesis itself contributes to senescence. There are a few controversial phenotypic findings describing altered glycogen metabolism in aging and aging-related diseases. Decrease of glycogenic activity was shown to be associated with aging (Khandelwal et al., 1984; Dall’Aglio et al., 1987; Mysliwski & Kmiec, 1992; Kurokawa et al., 1997), whereas increased glycogen accumulation was also observed in cellular senescence and aged tissues (Robbins et al., 1970; Moore et al., 1981; Gertz et al., 1985). Therefore, it is currently unclear whether alterations in glycogenesis are centrally involved in the senescence and aging process. In this study, we present clear evidence that enhanced glycogenesis via GSK3/GS modulation is directly involved in cellular senescence. In addition to glycogen accumulation, GS activation and GSK3 inactivation commonly occur in diverse cellular senescence systems and aged F344 rats. Inhibition of GSK3, using molecular inhibitors, siRNA, or dominant negative GSK3 mutant (K85A), is sufficient to induce senescence phenotypes. Moreover, knock-down of GS using siRNA significantly attenuates senescence-related processes that have been induced by GSK3 inhibitor. Our results suggest that augmenting glycogenesis via GSK3/GS regulation is an important modulator of cellular senescence, providing a new insight into the metabolic background of cellular senescence and aging.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  9. Supporting Information

Glycogenesis is commonly enhanced in cellular senescence and aging rat models

Previously, we developed stress-induced cellular senescence systems of human hepatocyte-derived Chang cells using deferoxamine (DFO) and H2O2, in addition to generating replicative senescence of human diploid fibroblasts (HDFs) (Yoon et al., 2006). To evaluate whether glycogenesis is associated with senescence, we first examined the phosphorylation level of GSK3 and GS, the major regulators of glycogenesis, in DFO-induced senescence. Phosphorylation levels (indicative of the inactive state) at both serine 9 (S9) of GSK3β and serine 21 (S21) of GSK3α increased, but the GS phosphorylation (at S640) did not look changed on the Western blot. Eventually, we found that the ratio of phosphorylated (inactive) GS to total GS was clearly decreased as a result of the increased expression of total GS protein levels (Fig. 1A,a). The dramatic induction of total GS level in this senescence is quite noteworthy. Acquisition of senescent phenotypes was confirmed by enlarged and flattened cellular morphology and senescence-associated β-galatosidase activity (SA-β-gal) (Fig. 1A,b) and irreversible growth arrest at G1 phase (Yoon et al., 2002). To prove further whether these GSK3/GS modulations are truly related with glycogen production, we monitored cellular glycogen contents with amylase-sensitive periodic acid Schiff's reactivity (PAS). As shown in panel c of Fig. 1A, cellular glycogen level was progressively augmented in a time-dependent manner. The credibility of PAS levels for glycogen was verified by observation of increased dense glycogen particles with electron microscopy (Fig. 1A,d). The levels of accumulated glycogen were compared with that of the insulin-stimulated one in high-glucose medium. Interestingly, DFO-treated glycogen accumulation was distinctly progressive and much higher than the maximum level of the insulin-stimulated condition (supporting information Fig. S1). Similar results were obtained in H2O2-induced senescence of Chang cells (Fig. 1B) and in mid-old HDFs with a doubling time of 1 week (Fig. 1C). We further examined the phosphorylation status of AKT and S6 K to identify the common upstream modulator of GSK3 in these senescent cells (supporting information Fig. S2). However, no common upstream signal to inactivate GSK3 was found in the two senescence systems, implying that diverse signals activated by different senescent stimuli may be converged on GSK3 phosphorylation.

imageimage

Figure 1. Enhanced glycogenesis in cellular senescences and aged F344 rats. (A) Chang cells were treated with 400 µm DFO for the indicated periods. (a) Western blot analysis for expression levels and phosphorylation status of GS and GSK3. The expression ratios for total GS/tubulin (inline image), pGS/GS (inline image), pGSK3α/GSK3α (inline image), and pGSK3β/GSK3β (inline image) are shown in the right panel. (b) SA-β-gal assay. Glycogen accumulation was visualized by PAS-stained images (c) and electron microscopic images (d). Purple color of (c) indicates glycogen stained by PAS reagent. The specificity of PAS staining to glycogen was confirmed by pretreatment of amylase (+). The arrows of (d) indicate glycogen particles. (B) Chang cells were treated with 200 µm H2O2 for the indicated periods. (a) Western blot analysis for expression levels and phosphorylation status of GS and GSK3. Senescence phenotype and glycogen contents were confirmed by SA-β-gal activity (b) PAS staining (c), respectively. (C) Young HDFs (PD16) and old HDFs (PD73) were applied to Western blot analysis for expression levels and phosphorylation status of the GS and GSK3 (a). PAS staining (PAS) and SA-β-gal activity (SA-β-gal) are shown (b). (D) Liver tissues of the aging process of F344 rats (3 individuals for each age) were extracted. (a) Western blot analysis for expression levels and phosphorylation status of GS and GSK3. The expression ratios for pGS/GS (inline image), pGSK3α/GSK3α (inline image), and pGSK3β/GSK3β (inline image) are shown in the right panel. (b) PAS staining of F344 rat liver tissues. *Upper band of the second rat at 18 month is an unidentified protein.

We also screened liver tissues in F344 rats spanning a range of ages. Strikingly, the ratio of phosphorylated inactive GS to total GS progressively decreased whereas the augmented phosphorylation of GSK3 (both α and β) was observed in rats aged 12 and 18 months (Fig. 1D,a). As expected, the highest level of glycogen was found in rats aged 15 months and minor increase in rats aged 24 months (Fig 1D,b). These results suggest that glycogen accumulation, by enhanced glycogenesis through GSK3 inactivation and GS activation, is a common event in the process of cellular senescence and aging. However, the conflicting coexistence of active (dephosphorylated) GS and active (dephosphorylated) GSK3 in 24-month aged rats and the minor increase of glycogen in this situation remain unexplainable. Detailed mechanisms on these phenomena are currently under investigation.

GSK3 inhibition by pharmacological inhibitors induces cellular senescence, and accompanies augmented glycogen synthesis

Next, we tested whether direct inhibition of GSK3 activity could induce senescence phenotypes in order to evaluate the potential involvement of enhanced glycogenesis in cellular senescence. GS phosphorylation was decreased when SB415286, a small molecule inhibitor of GSK3 (Coghlan et al., 2000), was applied to Chang cells at concentrations < 15 µg mL−1, but rather increased at higher concentrations of SB415286 (20 and 25 µg mL−1) (Fig. 2A,a). Regardless of GS phosphorylation status, total cellular growth rate declined in a dose-dependent manner (Fig. 2A,b). Although no significant cell death was observed at SB415286 concentrations < 12.5 µg mL−1, dead cells were significantly increased in a dose-dependent manner at concentrations > 15 µg mL−1 (Fig. 2A,b). These results suggest that there may be a strong unknown regulation on GS phosphorylation in these cytotoxic conditions in spite of the presence of GSK3 inhibitor.

imageimage

Figure 2. Pharmacological GSK3 inhibitors induce cellular senescence, accompanied by glycogen accumulation. (A) Chang cells were treated with the indicated concentrations of SB415286 for 2 days. (a) Western blot analysis for phosphorylation status of GS. (b) Cell viability was obtained by counting trypan blue-negative live cells (open fraction of bar) and positive dead cells (closed fraction of bar). Total cell numbers are obtained from the sum of live and dead cells. (B) Dose-dependent (a) and time-dependent (b) SA-β-gal activities of Chang cells exposed to subcytotoxic doses of SB415286 are shown. Chang cells were treated with the indicated concentration of SB415286 for 3 days (a) or treated with 12.5 µg mL−1 of SB415286 for the indicated periods (b) and then cells were applied to SA-β-gal activity. The lower panel displays phosphorylation status of GS protein by Western blot analysis to prove the effect of SB415286. (c) Representative images of the SA-β-gal activity and PAS staining of Chang cells exposed to 12.5 µg mL−1 SB415286. (C) Chang cells were treated with the indicated concentrations of LiCl for 3 days. Representative images of SA-β-gal activity (a), cell population of SA-β-gal positive cells (b), and GS phosphorylation status by Western blot analysis (c) are shown. (D) Young HDFs (PD 18) were exposed to 400 µm DFO, 12.5 µg mL−1 SB415286 (SB41), or serum-free (SF) medium for 3 days. Representative images of SA-β-gal activity and PAS staining are shown. **p < 0.01 vs. control.

Next, we applied SA-β-gal staining to cells with the subcytotoxic doses to examine gain of senescence phenotypes because subcytotoxic doses of the inhibitor arrested the cellular growth rate without cell death. As expected, at the subcytotoxic doses (from 7.5 µg mL−1 to 15 µg mL−1) of SB415286, cells acquired senescence phenotypes in dose- and time-dependent manners, as shown by the gain of SA-β-gal activity and flattened, enlarged cellular morphology (Fig. 2B). Clearly, GS was activated by dephosphorylation in these conditions, resulting in the progressive accumulation of glycogen along with the senescence phenotypes, such as SA-β gal and induction of p16 and p21 (Fig. 2B and supporting information Fig. S3). These results suggest that the delayed growth rate by the subcytotoxic doses of the GSK3 inhibitor is due to the result of progression into senescence. Acquisition of senescence phenotypes by GSK3 inhibition was confirmed again with LiCl, another conventional GSK3 inhibitor (Fig. 2C).

We further examined the effect of SB415286 in primary young HDFs, in comparison with the effects of DFO and the growth arrest effect by serum starvation to discern whether the glycogen accumulation was merely a result of growth arrest. As shown in Fig. 2(D), acquisition of senescence phenotypes and glycogen accumulation were similarly observed with SB415286 and DFO, but not with serum-free media. Additionally, induction of senescence phenotypes by SB415286 and LiCl was also demonstrated in the human cervical adenocarcinoma-derived HeLa cell line (supporting information Fig. S4). From these results, it could be inferred that subcytotoxic doses of GSK3 inhibitors are potent senescence inducers and GSK3 inactivation may be critically involved in cellular senescence.

The doses of SB415286 capable of inducing senescence phenotypes are relatively high compared to the characterized optimal concentration (10 µm) that specifically inhibits GSK3 in vitro (Coghlan et al., 2000), although several recent reports have used similar or higher concentrations (MacAulay et al., 2005; Wang et al., 2006). Therefore, we repeatedly applied 3.6 µg mL−1 SB415286, to mimic the effect of continuous daily application of the drugs in vivo. Single treatment of 3.6 µg mL−1 SB415286 could not induce senescence. But three repeated applications with 3.6 µg mL−1 SB415286 were enough to induce senescence phenotypes (gain of SA-β-gal and delayed growth rate). Increase of SB415286 to five repeated applications augmented the effect (Fig. 3 and supporting information Fig. S5A), indicating that the phenotypes achieved at the doses (7.5 to 12.5 µg mL−1) used in our experiments are not simply nonspecific side effects induced by higher inhibitor concentration. Therefore, we decided to use 12.5 µg mL−1 of SB415286 as an optimal concentration to induce senescence phenotypes in roughly 50% of the cell population on day 3.

image

Figure 3. Repeated treatment of low dose SB415286 induces both senescence phenotypes and glycogen accumulation. (A) Experimental scheme for repeated treatment of SB415286. Chang cells were treated with 3.6 µg mL−1 SB415286 in fresh medium every day for three (×3) or five (×5) days. After 2.5 days cells were split into two wells to maintain cells subconfluent. (B) (a) Cells were applied to SA-β-gal assay and SA-β-gal positive cells were counted and presented as percent of total cells. (b) Cell viability is shown by counting live cells. (C) PAS-stained cells were counted and presented as percent of total cells. **p < 0.01 vs. control.

Download figure to PowerPoint

Enhanced glycogenesis and senescence phenotypes induced by GSK3 inhibitors are irreversible

It is generally accepted that a key characteristic of senescent cells is irreversible gain of senescence phenotypes. Thus, we examined the irreversibility of SB415286-induced senescence. Surprisingly, not only SB415286-induced SA-β-gal activity, but also glycogen accumulation was irreversible depending on the duration of drug exposure (Fig. 4A,B; supporting information Fig. S5B). A similar effect on GS dephosphorylation and delayed cellular growth was observed (Fig. 4B,b,c). In Fig. 1, enhanced glycogenesis via GS dephosphorylation was displayed in H2O2-induced senescence. This result implies that reactive oxygen species (ROS) may be involved in GS activation and consequent augmentation of glycogenesis. Therefore, we monitored intracellular ROS levels using 2′,7′-dichlorodihydrofluorescein (DCF) fluorescence to elucidate whether the irreversible effect on both glycogenensis and senescence phenotypes is the result of ROS generation. As expected, ROS was increased by SB415286 in a dose-dependent manner, and the ROS was maintained even after release from SB415286 treatment (Fig. 4C), suggesting that the initial ROS generated by SB415286 can sustain the active, dephosphorylated state of GS, thereby continuously enhancing glycogenesis. The gain of senescence phenotype in the same experimental condition was confirmed again by increase of cellular granularity (Fig. 4D), another senescence marker (Yoon et al., 2005). These results suggest that GSK3 inhibition irreversibly induces senescence phenotypes and that ROS may play a role as modulator of glycogenesis activation. Therefore, glycogen accumulation could potentially be utilized as a novel marker of senescence.

imageimage

Figure 4. Glycogenesis and ROS production activated by SB415286 is irreversible along with acquisition of senescence phenotypes. Chang cells were treated with 12.5 µg mL−1 SB415286 for 1, 1.5, or 2 days and released by replenishment with fresh media without drug for the indicated periods. (A) Experimental scheme for irreversibility of the effect of SB415286. (B) (a) SA-β-gal activity. (b) Cell viability was compared by counting live cells and presented as percent of control. (c) PAS staining. (d) Western blot analysis. (C) Intracellular ROS levels. (a) Mean values of the arbitrary DCF fluorescence unit were analyzed. (b) Representative profiles of the FACS results. (D) Cellular granularity. (a) Cell populations in the R1 region of (b) were analyzed as percent of total cells. (b) Representative profiles of side-scatter (SSC) results. **p < 0.01; *p < 0.05 vs. control.

GSK3 inactivation induces senescence phenotypes with accompanying glycogen accumulation

To eliminate any possible nonspecific effect of GSK3 inhibitors on acquisition of senescence, we introduced siRNA against GSK3α and GSK3β. Partial knock-down of GSK3α or GSK3β decreased GS phosphorylation and combined application of siRNAs for both GSK3α and GSK3β further effectively attenuated GS phosphorylation (Fig. 5A,a). All three conditions significantly induced senescence phenotypes and glycogen accumulation, and combined treatment showed more effective results (Fig. 5A,b,c; supporting information Fig. S6A). In addition, these combined effects were clearly reversed by ectopic expression of both GSK3α and GSK3β (Fig. 5B).

imageimage

Figure 5. Inactivation of GSK3 induces senescence, accompanying glycogen accumulation. (A) Chang cells were transfected with the indicated amounts (pmole per 60 mm plate) of siRNAs for GSK3α and/or GSK3β (si-GSK3α, si-GSK3β) for 3 days. siRNA for random sequence was used as negative control (si-NC). (a) Western blot analysis. The expression ratio for pGS/GS is shown in the lower panel. (b) SA-β-gal activity. (c) PAS staining. ##p < 0.01 vs. control; **p < 0.01 vs. si-NC. (B) Chang cells pretransfected with pcDNA3, pcDNA-GSK3α-HA, and pcDNA-GSK3β-HA plasmids (2.5 µg per 60 mm plate) were treated again with the indicated amounts (50 or 100 pmole per 60 mm plate) of siRNAs for 3′UTR of GSK3α (si-GSK3α) and 5′-UTR of GSK3β (si-GSK3β) for 3 days to knock-down endogenous GSK3. (a) SA-β-gal activity and PAS staining. **p < 0.01. (b) Western blot analysis. Arrows indicate ectopically expressed HA-tagged GKS3α and β. (C) Chang cells were transfected with pIRES-EGFP, pIRES-GSK3β-HA, pIRES-GSK3β(S9A)-HA, or pIRES-GSK3β(K85A)-HA plasmids (2.5 µg per 60 mm plate) for 3 days. (a) SA-β-gal activity and PAS staining. **p < 0.01. (b) Western blot analysis.

Next, we further examined the effect of overexpression of GSK3β dominant negative mutant (K85A) (Dominguez et al., 1995). Although overexpressed wild-type GSK3β did not affect any cellular function of Chang cells, GSK3β-K85A obviously and effectively induced senescence phenotypes with glycogen accumulation (Fig. 5C and panel a of supporting information Fig. S6B). Similar results were obtained in HDFs (panel b of supporting information Fig. S6B). However, it is also noteworthy that the active GSK3β-S9A mutant triggered both senescence phenotypes and glycogen accumulation, but to a weaker extent. These findings demonstrate that GSK3 inactivation is sufficient to induce senescence phenotypes, and imply a potential role of GSK3 as a modulator of cellular senescence.

Glycogenesis via GS is genuinely involved in cellular senescence

To discern whether glycogenesis through GS activity is truly involved in induction of cellular senescence and not an epiphenomenon, GS was knocked down using siRNAs prior to challenge with SB412586. Interestingly, the knock-down of GS partially attenuated, but significantly, the induction of senescence phenotypes (SA-β gal and p21 induction) by SB415286, in addition to abolishing glycogenesis (Fig. 6; supporting information Fig. S7). These results clearly indicate that glycogenesis through GS activity plays a role as a contributing factor in inducing cellular senescence, and not an unrelated phenomenon.

image

Figure 6. Knock-down of GS attenuates both senescence and glycogenesis induced by SB415286. Chang cells were transfected with the indicated amounts of two different siRNAs for GS (si-GS, #1 and #2) at 1.5 days prior to treatment with 12.5 µg mL−1 SB415286 for 2 days. siRNA for random sequence was used as negative control (si-NC). (A) Western blot analysis. (B) SA-β-gal activity. (C) PAS staining. *p < 0.05 and **p < 0.01 vs. without SB415286; ##p < 0.01 vs. SB415286-treated si-NC.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  9. Supporting Information

GSK3 has long been studied for its role as a negative regulator of glycogen synthesis. Insulin released in response to increased blood glucose initiates a signaling cascade involving insulin receptor (IR) and insulin receptor substrate (IRS), thereby activating PI3 K and AKT in turn (Rylatt et al., 1980; Forde & Dale, 2007). Activated AKT phosphorylates GSK3 at S21 and S9 residues of the GSK3α and GSK3β isoforms, respectively, (Sutherland et al., 1993) and then the phosphorylated inactive GSK3 displays limited kinase activity, resulting in the accumulation of dephosphorylated active GS. Consequently, glucose transported into cells from the blood is converted into glycogen and stored for future use. This insulin or insulin-like signaling cascade, involving IGF-1 receptor, IRS, PI3 K, and AKT, has recently been implicated in controlling lifespan and senescence (Holzenberger et al., 2003; Miyauchi et al., 2004; Kenyon, 2005), implying that the GSK3 and GS couple, one of the major downstream targets of the insulin/IR/PI3 K/AKT cascade, may participate in senescence. However, the relationship of GSK3 and subsequent glycogen metabolism to senescence has not been clearly elucidated. The results presented by our study clearly indicate that glycogen accumulation via GSK3 inactivation and subsequent GS activation is common in both cellular senescence and the aging process of F344 rats. Irreversible enhancement of glycogenesis by GSK3 inhibition and its involvement in cellular senescence further emphasizes the potential role of GSK3 as a modulator of cellular senescence, and the direct involvement of glycogenesis in cellular senescence.

How is the augmented glycogenesis irreversibly sustained during senescence? One prominent mechanism to explain cellular senescence is cumulative irreversible damage by toxic oxygen intermediates like ROS and the resulting activation of stress-related signaling by continuously increased ROS generation (Ames et al., 1993; Finkel & Holbrook, 2000). Therefore, we hypothesize that increased ROS may affect the GSK3/GS modulation of glycogen synthesis. This hypothesis was supported by the exogenous H2O2-induced enhancement of glycogenesis through GSK3 inactivation and GS activation in cellular senescence. Moreover, in addition to SB415286-induced GS activation, resultant glycogen accumulation, increased ROS generation was irreversibly sustained in senescent cells, even after removal of the drug. Recent studies describe that phosphorylation of Akt, the upstream modulator of GSK3, is directly modulated by ROS through a redox-mediated mechanism (Pelicano et al., 2006) and Erk 1/2, another kinase modulated by redox-mediated activation, is a negative regulator of GSK3 (Locher et al., 2002; Wang et al., 2006). Therefore, the potential link of ROS to glycogenesis can be explained by activation of redox-sensitive upstream kinases, such as Akt or Erk 1/2, by ROS generated during progression into senescence.

It is important to evaluate whether GSK3 inactivation is closely related to senescence. Induction of senescence phenotypes and augmented glycogenesis by conventional GSK3 inhibitors (SB415286 and LiCl) in diverse cell lines, such as human liver-derived Chang cells, human cervical adenocarcinoma-derived HeLa cells, and primary HDFs, suggest both a potential role for GSK3 as a modulator of senescence and direct involvement of glycogenesis in senescence. We also tested SB216763, another widely used GSK3 inhibitor (Coghlan et al., 2000; Wang et al., 2006). SB216763 clearly induced SA-β-gal activity, the most generally accepted senescence marker, along with weak glycogen accumulation, but did not increase cell size (supporting information Fig. S8). Similar efficacy and specificity of both SB415286 and SB216763 at blocking GSK3 activity has been noted, although SB216763 has a slightly lower Ki than SB415286 (9.1 nm vs. 30.75 nm) as well as a higher EC50 on glycogen synthesis (30 µm vs. 10 µm) (Coghlan et al., 2000), implying that effects of these two drugs on downstream targets of GSK3 may differ. Wang et al. reported that SB216763 does not activate eIF2B, a GSK3 target, whereas LiCl and SB415286 do at concentrations displaying similar GSK3 inhibition (Wang et al., 2002), supporting the idea that SB216763 is not sufficient to trigger the necessary protein translation to increase cell mass in cellular senescence, thereby displaying small cell morphology. In addition to the GSK3 inhibitors, knock-down of GSK3 using siRNA and overexpression of inactive GSK3β-K85A mutant triggered senescence with glycogen accumulation, which provides clear evidences for the involvement of GSK3 inactivation in cellular senescence. However, our present results are contradictory to the recent observations reported by other groups, describing that active GSK3 contributes to senescence through modulating p53 activity (Zmijewski & Jope, 2004), cyclin D1 (Kortlever et al., 2006), or HIRA activity (Ye et al., 2007). This discrepancy possibly comes from different cellular contextual environments or experimental conditions. We have also found some diverse effects of the GSK3 inhibitors on cellular morphology and destination, depending on doses and types of the inhibitors even in the same cell as aforementioned. More interestingly, overexpression of active GSK3β-S9A mutant could weakly drive senescence with glycogen accumulation. All these findings imply that well-balanced GSK3 activity may be more important to maintain cellular function and abnormal modulation of the activity can be a stressor to cells. Nevertheless, our results obviously support that GSK3 inactivation acts as the more effective inducer of senescence.

Although GSK3 inactivation is sufficient to induce cellular senescence, enhanced glycogenesis may not be a contributing factor, but a epiphenomenon, as there are increasing numbers of GSK3 targets being identified. However, knock-down of GS significantly attenuated the SB415286-induced senescence. This result proves that glycogenesis via GS activation is truly involved in cellular senescence. How can glycogenesis via GS activation contribute to senescence induction? One possible explanation is that in contrast to temporal glycogen accumulation by insulin, the irreversibly and progressively increasing glycogen accumulation may constitute sufficient stress to cells by occupying most cytoplasmic space and disrupting appropriate localization of and/or rapid communication between intracellular organelles, thereby triggering and maintaining cellular senescence. This hypothesis is further supported by the finding that dysregulation of glycogen metabolism is implicated in the development of some diseases, including type 2 diabetes mellitus (Shulman et al., 1990; Cline et al., 1994), although the direct effect of imbalanced glycogen metabolism on cellular function and fate has yet to be elucidated. Another possibility may come from unknown roles for GS besides that in glycogen synthesis. This theory is currently under further investigation.

Taken together, our results demonstrate that coupled GSK3 inactivation/GS activation and subsequent glycogen accumulation are common in cellular senescence and aging. In addition, GSK3 inhibition is sufficient to induce senescence and GS knock-down prevents progression of senescence, implying that enhanced glycogenesis is critically involved in cellular senescence through GSK3 inactivation and GS activation. Our results suggest novel roles for GSK3 and GS as central modulators of or contributors to cellular senescence, and provide new insight into the metabolic background of cellular senescence and aging.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  9. Supporting Information

Cell culture and cell viability

Chang cells were obtained from the American Type Culture Collection (Rockville, MD, USA) and were grown in Dulbecco's modified eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS, Gibco) and antibiotics at 37 °C in a humidified incubator with 5% CO2. The primary culture of normal HDF was isolated from the foreskin of a 4-year-old boy by the method described previously (Kim et al., 2003). Confluent cells were transferred to a new dish, and the numbers of population doublings (PD) as well as the doubling time were continuously counted. Young and old cells used in the present experiments were defined as the HDF with a doubling time of around 24 h and over 1 week, respectively.

To monitor cell viability, cells (3 × 104) were seeded on 12-well plates. Upon completion of the experiment, cells were harvested by trypsinization and counted with a hemocytometer after staining with 0.4% (w/v) trypan blue (Gibco BRL) to exclude dead cells. To evaluate cellular viability, the population of trypan blue-negative live cells and positive dead cells were counted.

Introduction of siRNA or cDNA into cells

Oligonucleotides for GSK3α siRNA (5′-GAAAGACGAGCUUUACCUATT-3′), GSK3β siRNA (5′-CUCAAGAACUGUCAAGUAATT-3′), GS siRNA (5′-GAAUCCUUAUCCAGGCUAATT-3′ or 5′-AGGCCAAGCUGUGCGCAAATT-3′), and negative control siRNA (5′-UAGCGACUAAACACAUCAA-3′) were produced by Samchully Pharm.Co. (Seoul, Korea). To specifically knock-down endogenous GSK3 in the cells transfected with exogenous GSK3 cDNA, siRNAs for 3′-untranslated region (UTR) of GSK3α (5′-GACUAGAGGGCAGAGGUAATT-3′) and for 5′-UTR of GSK3β (5′-GUGCCGAUCUGUCUUGAAGTT-3′) were produced by Samchully Pharm.Co. siRNA duplexes were introduced into cells using Oligofectamine (Invitrogen, Grand Island, NY, USA) according to manufacturer's instruction.

The human GSK3β wild-type and mutant expression plasmids, pcDNA-GSK3β-HA, pcDNA-GSK3β(S9A)-HA, and pcDNA/GSK3β(K85A)-HA, were generously provided by Dr Eui-Ju Choi (Korea University, Seoul, Korea) and cDNA fragments harboring GSK3-HA were subcloned into EcoRI and ApaI sites of pIRES2-EGFP (BD Bioscience Clonetech, San Jose, CA, USA), generating pIRES-GSK3β-HA, pIRES-GSK3β(S9A)-HA, and pIRES-GSK3β(K85A)-HA. The GSK3α expression plasmid, pcDNA-GSK3α-HA, was generated by cloning the GSK3α cDNA fragment amplified against human GSK3a clone (clone ID 3903896, Open Biosystem, Huntsville, Al, USA) with a primer set of 5′-TAAGAATTCCATGAGCGGCGGCGGGCCTTC-3′ and 5′-ATTAGATCTGGAGGAGTTAGTGAGGGTAGG-3′ into EcoRI/BamHI sites of pcDNA3-HA vector. cDNAs were introduced into cells using FuGENE HD transfection reagent (Roche Diagnostics, Indianapolis, IN, USA) according to manufacture's instruction.

Senescence-associated β-galactosidase (SA-β-gal) assay

SA-β-gal activity was assayed at pH 6.0 as described by Dimri et al. (1995) with a slight modification. Briefly, cells were washed twice with PBS (phosphate buffered saline), fixed to plates by 3.5% formaldehyde for 5 min, and then incubated overnight in freshly prepared staining solution [40 mm citrate-phosphate buffer, pH 6.0, containing 1 mg mL−1 of 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal; Sigma, St. Louis, MO, USA), 5 mm potassium ferrocyanide, 5 mm potassium ferricyanide, 150 mm NaCl, and 2 mm MgCl2]. The stain was visible 12 h after incubation at 37 °C. By counting the numbers of the blue-colored and total cells under using Image J software (NIH, Bethesda, MA, USA), percentage of the cells stained blue was estimated to compare the degree of senescence-associated cells.

Detection of glycogen accumulation in cells and tissues

Staining of glycogen was performed by the Periodic Acid-Schiff (PAS) reagent (Sigma, St. Louis, MO, U.S.A.). Cells were fixed in 3.5% formaldehyde for 10 min at room temperature and stained with 500 µL of 0.5% Schiff's solution for 15 min. After washing cells with PBS twice, stained cells were observed and photographed. To determine the origin of the staining, another set of cells were treated with 0.5% amylase (Sigma) for 90 min at 37 °C to degrade intracellular glycogen before PAS staining. By counting the numbers of the purple-colored and total cells under using Image J software (NIH, Bethesda, MA, USA), percentage of the cells stained purple was estimated.

To stain the liver tissues, sectioned frozen blocks were stained in 0.5% Schiff's solution for 2 min and further counterstained with hematoxylin. For negative controls, parallel liver sections were pretreated with 2%α-amylase (Sigma) at 37 °C for 15 min prior to PAS staining.

Determination of intracellular ROS level and granularity

To determine intracellular level of ROS, we used DCFH-DA fluorogenic probe (Molecular Probe, Eugene, OR, USA). Cells were treated with SB415286 for the indicated periods and further incubated in media containing 10 µm DCFH-DA for 15 min at 37 °C. Stained cells were washed, resuspended in PBS, and analyzed by flow cytometry (FACS Vantage, Becton Dickinson Corp.). Mean values of artbitrary fluorescence unit of 10 000 cells were presented and H2O2-treated cells were used as positive control.

Cellular granularity was evaluated by analyzing side scatter of the DCFH-DA stained cells using flow cytometry.

Western blot analysis and antibodies

Cell lysates were applied to Western blot analysis as previously described (Yoon et al., 2006). Antibodies against phospho-GSK3α (pGSK3α, cat. 9337), phospho-GSK3β (pGSK3β, cat. 9336), and pGS (cat. 3893) were purchased from Cell Signaling (Beverly, MA, USA). Antibodies for GSK3 (sc-7291), p16 (sc-759), p21 (sc-397), and p27 (sc-527) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA), and phospho-GS antibodies were from Cell Signaling (#3891) and Chemicon (AB3379; Temecula, CA, USA). Antibody against α-tubulin (Ab-1, Oncogene, Boston, MA, USA) was used as primary antibodies. The expression ratios for total GS/tubulin, pGS/GS, pGSK3α/GSK3α, and pGSK3β/GSK3β were obtained by densitometric analyses of the Western blots.

Electron microscopy

Cells were fixed in Karnovsky's fixative solution (1% paraformaldehyde, 2% glutaraldehyde, 2 mm calcium chloride, 100 mm cacodylate buffer, pH 7.4) for 2 h, washed with cacodylate buffer, and postfixed in 1% osmium tetroxide and 1.5% potassium ferrocyanide for 1 h. Cells were then stained en bloc in 0.5% uranyl acetate, dehydrated through a graded ethanol series and embedded in Poly/Bed 812 resin (Pelco, CA, USA). Cells were sectioned using Reichert Jung Ultracut S (Leica, Cambridge, UK). After staining cells with uranyl acetate and lead citrate, cells were observed and photographed under transmission electron microscope (Zeiss EM 902 A, Leo, Oberkohen, Germany).

Acknowledgments

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  9. Supporting Information

This work was supported by the Korean Science and Engineering Foundation (KOSEF) Grant funded by the Korean government (MOST) (R13-2003-019-01004-0), Korea Research Foundation Grant funded by the Korean Government (KRF-2007-313-C00575 & KRF-2005-041-C00354). We thank the Aging Tissue Bank (Pusan, Korea) for the supply of the aged tissues.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  9. Supporting Information

Fig. S1 Comparison of glycogen accumulation in DFO-induced senescence with that of insulin-treated cells. Chang cells were treated with 400 µM DFO or 100 nM insulin in low (5 mM) glucose- or high (25 mM) glucose-medium for the indicated periods. Glycogen accumulation was detected by PAS staining. In low glucose medium, maximum glycogen accumulation was obtained 12 h after exposure to 100 nM insulin, and thereafter slowly decreased. In high glucose medium the maximum level of accumulated glycogen was maintained up to 48 h.

Fig. S2 Phosphorylation status of GSK3 upstream regulators. Chang cells were treated with 400 µM DFO or 200 µM H2O2 for the indicated periods. (a) Western blot analysis for phosphorylation status of AKT and S6K using antibodies specific to phosphorylated AKT (cat. 4051, Cell Signaling, Beverly, MA, USA) or phosphorylated S6 K (cat. 9205, Cell Signaling).

Fig. S3 Expression of inhibitory cell cycle regulators of SB415286-induced senescence. Chang cells were treated with 12.5 µg mL−1 SB415286 for the indicated periods. Western blot analysis for expression levels of p21, p16, and p27 proteins.

Fig. S4 GSK3 inhibitors induce cellular senescence of Hela cells, accompanied by glycogen accumulation. Hela cells were treated with the indicated concentrations of SB415286 or LiCl for 3 days. (A) SA-β-gal activity. The lower panel shows phosphorylation status of GS protein of the GSK3 inhibitor-treated cells by Western blot analysis. (B) Representative images of SA-β-gal activity of Hela cells exposed to SB415286 (12.5 µg mL−1) or LiCl (25 mM) for 3 days. (C) Representative images of PAS staining of the SB415286-treated cells. The specificity to glycogen was confirmed by pretreatment of amylase (+).

Fig. S5 Effect of repeated treatment of low dose SB415286 and irreversible effect of SB415286. Experimental conditions are the same as those of Fig. 3(A) and Fig. 4B. Representative images of PAS and SA-β-gal staining are shown.

Fig. S6 Effect of GSK3 inactivation. (A) Experimental conditions are the same as those of Fig. 5A. Representative images of PAS and SA-β-gal staining are shown. (B) Experimental conditions are the same as those of Fig. 5C. (a) Representative images of PAS and SA-β-gal staining of Chang cells are shown. GFP fluorescence images indicate transfected cells. (b) The same experiments were performed in HDFs.

Fig. S7 Effect of GS knock-down. Experimental conditions are the same as those of Fig. 6. Representative images of PAS and SA-β-gal staining with GS siRNA (#1) are shown.

Fig. S8 SB216763 induces glycogen accumulation and SA-β-gal activity along with delayed cellular growth rate. Chang cells were treated with the indicated concentrations of SB216763 or LiCl for 3 days. (A) Cell viability was obtained by counting trypan blue-negative live cells (open bar) and positive dead cells (closed bar) and presented as percentage of total cells. (B) Phosphorylation status of GS protein of SB216763-treated cells by Western blot analysis. (C) Representative images for SA-β-gal and PAS staining of Chang cells exposed to the indicated concentrations of SB216763 for 3 days.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing material) should be directed to the Aging Cell Central Office.

FilenameFormatSizeDescription
ACEL_436_sm_FigS1-2.tif484KSupporting info item
ACEL_436_sm_FigS3.tif87KSupporting info item
ACEL_436_sm_FigS4.tif535KSupporting info item
ACEL_436_sm_FigS5.tif577KSupporting info item
ACEL_436_sm_FigS6.tif879KSupporting info item
ACEL_436_sm_FigS7.tif427KSupporting info item
ACEL_436_sm_FigS8.tif369KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.