SEARCH

SEARCH BY CITATION

Keywords:

  • 17-AAG;
  • hormesis;
  • neurodegenerative diseases;
  • neuroprogenitor;
  • stem cell;
  • stress adaptation

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussions
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

J. Neurochem. (2011) 117, 703–711.

Abstract

Stem cell-based approaches provide hope as a potential therapy for neurodegenerative diseases and stroke. One of the major scientific hurdles for stem cell therapy is the poor survival rate of the newly formed or transplanted neural stem cells. In this study, we found that low-dose treatment with the Heat shock protein 90 (Hsp90) inhibitor 17-allylamino-17-demethoxygeldanamycin (17-AAG), a heavily investigated anti-cancer drug, prevented neural progenitor cells from either naturally-occurring or stress-induced apoptosis, although it induced apoptosis at higher doses. This stress adaptation effect mediated by low-dose 17-AAG is accompanied by activation of multiple cell survival pathways, including the stress response pathway (induction of Hsp70), the MAPK pathway, and the PI3K/Akt pathway. When administered in vivo, 17-AAG led to Akt and glycogen synthase kinase 3β phosphorylation, and more 5-bromo-2′-deoxyuridine positive cells in the mouse brain. These findings could have profound implications in stem cell therapy for neurodegenerative diseases and stroke.

Abbreviations used:
17-AAG

17-allylamino-17-demethoxygeldanamycin

β amyloid peptide

BrdU

5-bromo-2′-deoxyuridine

DMEM

Dulbecco’s modified Eagle’s medium

EB

embryoid bodies

ERK

extracellular signal regulated kinase

FGF

fibroblast growth factor

GA

geldanamycin

GSK-3β

glycogen synthase kinase 3β

Hsp90

heat shock protein 90

NECs

neuroepithelial cells

NP

neuroprogenitor

PBS

phosphate-buffered saline

PI3K

phosphotidylinosotol 3-kinase

SVZ

subventricular zone

The main rationale for cell-based therapies to treat neurodegenerative diseases and stroke is to replace the damaged CNS tissue in an organotypic appropriate manner and/or to provide a trophic support to the tissues at risk (Haas et al. 2005; Chang et al. 2007; Li et al. 2008; Lindvall and Kokaia 2010). However, many issues surrounding cell-based therapies remain to be solved prior to their clinical application. One of them is the low survival rate of the stem cells, either endogenously generated or transplanted (Li et al. 2008; Suzuki and Svendsen 2008; Hedlund and Perlmann 2009). It is crucial to find ways to preserve these endogenously generated or transplanted neural stem/progenitor cells.

Heat shock protein 90 (Hsp90) is one of the most abundant cellular chaperone proteins that functions in a wide range of cellular processes, including neuroprotection and cell apoptosis (Terasawa et al. 2005). In the past few years, several inhibitors have been developed to specifically block Hsp90. One of them is 17-(allylamino)-17-demethoxygeldanamycin (17-AAG), a geldanamycin (GA) derivative currently under heavy clinical trial against multiple malignancies (Terasawa et al. 2005; Gallo 2006; Waza et al. 2006). Apart from its potential applications in cancer treatment, 17-AAG is also known to alleviate neuronal degeneration by targeted degradation of disease-causing protein (Waza et al. 2005, 2006).

In addition, several studies have shown that GA promotes the survival and migration of the osteoclast (Koga et al. 2006; Yano et al. 2008), and exerts neuroprotective effect in ischemic brain (Kwon et al. 2008; Wen et al. 2008). However, the neural protective role of 17-AAG, especially its function on neural stem cells, has not been thoroughly investigated.

Hormesis, or stress adaptation, is a common paradigm found throughout nature (Rattan 2004; Norgaard et al. 2006; Calabrese et al. 2007). An initial exposure of a cell or an organism to a mild stressful stimulus results in an adaptive response by which a second exposure to the same stimulus produces a minimal response (Wheeler and Wong 2007). A more interesting phenomenon is cross-tolerance, by which an initial exposure to a stressful stimulus results in an adaptive response such that the cell or organism acquires resistance to a different stress. Several studies have reported that a mild stress produces stress adaptation or cross-tolerance in cells that potentiates them to endure much severe insults (Geraci et al. 2006; Kraft et al. 2006; Madhavan et al. 2006; Norgaard et al. 2006; Mattson 2008; Kendig et al. 2010). Recently, the term hormesis has been defined as a dose-response relationship in which there is an inhibitory response at higher doses but a stimulatory response at lower doses, resulting in a U-shaped or inverted U-shaped dose response (Calabrese et al. 2007; Mattson 2008; Kendig et al. 2010).

In this study, we set to answer whether the Hsp90 inhibitors, especially 17-AAG, mediated the stress adaptation effect in neuroprogenitor (NP) cells in vitro and the newly generated neural cells in vivo. To our surprise, we found that both GA and 17-AAG shown a typical stress adaptation phenomenon in NP survival and death.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussions
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Materials

ES-J1 mouse embryonic stem cells, derived from the inner cell mass of a male agouti 129S4/SvJae embryo(Li et al. 1992), were kindly provided by Dr. Erhard Bieberich (Medical College of Georgia, Augusta, GA) and the feeder cells were obtained from the Medical College of Georgia ES core facility (Dr. A. Eroglu, Medical College of Georgia, Augusta, GA). Knockout DMEM (Dulbecco’s modified Eagle’s medium), knockout serum replacement, ES-qualified fetal bovine serum, N2 supplement, and fibroblast growth factor (FGF)-2 were from Invitrogen (Carlsbad, CA, USA). 17-AAG was from LC laboratories (Woburn, MA, USA), Dulbecco’s modified Eagle’s medium/Ham’s F12 50/50 mix was purchased from Cellgro (Herndon, VA, USA). ESGRO leukemia inhibitory factor was from Chemicon International (Temecula, CA, USA). 5-Bromo-2′-deoxyuridine (BrdU), Hoechst 33258 and horseradish peroxidase-conjugated secondary IgGs were from Sigma-Aldrich (St Louis, MO, USA). C16 ceramide was from Avanti Polar Lipids (Alabaster, AL, USA). β-amyloid (25–35) was from rPeptide (Bogart, GA, USA) and β-amyloid (35–25) was from Anaspec, Inc (San Jose, CA, USA). Rabbit anti-phospho glycogen synthase kinase 3β (GSK-3β) (serine 9), rabbit anti-Akt and mouse anti-phospho AKT at Thr308 were from Cell Signaling Technologies (Beverly, MA, USA). Mouse anti-BrdU was from BD Biosciences (San Jose, CA, USA). All reagents used were of analytical grade or higher.

In vitro differentiation of ES cells to generate NPs

Pre-implantation blastocyst-derived mouse ES cells J-1 were grown on γ–irradiated feeder fibroblasts, and neural differentiation was induced by serum deprivation of embryoid bodies (EBs) as previously described (Bieberich et al. 2003; Wang et al. 2005). In brief, attached EBs were dissociated and EB-derived cells plated on polyornithin/laminin-coated dishes to generate NPs. At this stage, cells were grown in DMEM/F12 with N2 supplement and FGF-2 (NP medium). EB-derived cells acquire all characteristics of genuine NPs, including expression of Nestin, Sox1, and fibroblast growth factor receptor 1 (Bieberich et al. 2003).

Primary neuroepithelia culture

Primary cultured neuroepithelial cells (NECs) were derived from telencephalons of E14.5 (embryonic day 14.5) mouse embryo brain and cultured in N2-DMEM/F12 [N2-supplemented DMEM/Ham’s Nutrient Mixture F12] containing bFGF (basic FGF; Peprotech, Rocky Hill, NJ, USA) on dishes coated with poly-l-ornithine and fibronectin (Sigma-Aldrich)(Nakashima et al. 1999; Yanagisawa et al. 2006). Mice used for the cell preparation were treated according to the guidelines of the IACUC (Institutional Animal Care and Use Committee) of the Medical College of Georgia to minimize pain or discomfort. NECs cultured for 6 days were replated for treatment with 17-AAG.

Cell viability, proliferation, and cell death assays

Proliferation and apoptosis assays were performed with NP cells treated with or without Hsp90 inhibitors. The number of live NP cells was measured by a microplate reader (Cell Counting Kit-8, Dojindo Molecular Technologies, Inc, Rockville, MD, USA). Briefly, the Cell Counting Kit-8 solution was added to the treated cell culture, incubated for 2–4 h, and their absorption at wavelength 450 nm read by a microplate reader. The absorption at 450 nm was correlated to the live cell number present in the well. For flow cytometry, GA treated NPs were stained with terminal transferase uridyl nick end labeling (TUNEL) reagents (Roche Applied Sciences, Indianapolis, IN, USA) and counted by a Becton Dickinson FACSCalibur Flow Cytometer (Block Scientific, Inc., Bohemia, NY, USA). For the ceramide and β amyloid peptide (Aβ) treatment, NP cells were first treated with 10 nM 17-AAG for 1–2 h before adding 2 μM C16 ceramide or 5 μM Aβ (25–35) peptide. The reverse peptide βA (35–25) was used as control for the Aβ amyloid treatment experiments.

Western blotting, immunofluorescence, RNA extraction, and RT-PCR

The western blotting analysis, immunofluorescence, RNA extraction and RT-PCR were performed as previously described (Wang et al. 2003, 2005, 2006; Bieberich et al. 2004). The Epifluorescence microscopy was accomplished using a Zeiss Axiophot microscope equipped with a Spot digital camera (Carl Zeiss, Inc., Thornwood, NY, USA). RT-PCR Primers for Nestin: forward, 5′-ggagacctggaacatgaatc-3′; and reverse, 5′-cactcagagtgaacttcagc-3′.

BrdU labeling, tissue preparation, and quantification of BrdU-labeled cells

Four-month-old C57BL6 mice were used for these studies. Mice were intraperitoneally (i.p.) injected 17-AAG (0.2 mg and 1 mg/kg) twice within 48 h. Sixteen hours after the second injection of 17-AAG, BrdU (50 mg/kg, i.p.) was administered twice a day (at 2-h intervals) for 2 days to label newly formed cells. Control mice were injected the same way but with only vehicle solution [1% or 5% dimethylsulfoxide in phosphate-buffered saline (PBS)].

Twenty-four hours after the last BrdU injection, mice were anesthetized by isoflorine and perfused transcardially with PBS and then 4% paraformaldehyde in PBS. Brains were removed and cryoprotected. For each brain, 10 μm coronal sections covering the subgranular zone (SGZ) and subventricular zone (SVZ) were cut and every other section was collected (about 60 slides with three brain slices on each slide). Every other odd numbered slide (45 sections, 15 slides per brain) was processed for BrdU immunostaining. The dried sections were treated with 10 g/mL DNAse I for 30 min at 37°C. After washing with PBS, the sections were incubated with anti-BrdU antibody for 2 h at 25°C or overnight at 4°C. They were further incubated with secondary antibody for 1 h and then mounted and covered with cover glasses for epifluorescence or confocal microscopic analysis. The even numbered slides were H&E stained to clarify cytoarchitecture. The number of BrdU positive cells on the stained sections was counted double blindly and added up for each brain (resulted in number N). The total number of BrdU positive cells in each brain was then calculated as 8 × N.

Statistical analysis

Data from individual experiments were presented as mean ± SD (or SEM). Student’s t-test was used to compare the means between the treatment group and the control group. p < 0.05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussions
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Low dose of Hsp90 inhibitors protected NPs from naturally-occurring apoptosis

In order to study the Hsp90 inhibitors’ role in NP cells, we first determined the protein’s expression level in normal NP cells. Figure 1 shows both Hsp90α and Hsp90β were abundantly expressed in the NP cells during the first 48 h of plating (NP-24 h and NP-48 h), although there appeared to be some difference between the two isoforms during the attached EB2 stages.

image

Figure 1.  Hsp90s are expressed in mouse neural progenitor cells. Mouse ES cells J1 were differentiated according to established protocols. Samples were collected at stages indicated and western blot analysis performed. Note both Hsp90a and Hsp90β were abundantly expressed in NPs, although differentially expressed at other stages. ES, feeder free ES cells; EB1, floating embryoid bodies; EB2, attached embryoid bodies; NP, neuroprogenitor.

Download figure to PowerPoint

To elucidate the role of Hsp90 in regulating NP cell survival, we treated the NP cultures with two doses of GA (10 nM and 100 nM) and compared them to vehicle-treated control cultures. Surprisingly, only the 100 nM GA was found to induce apoptosis shown by condensed nuclei and morphology changes (41 ± 5.6% of total number of NPs, Fig. 2a and b), while the 10 nM GA protected NPs against naturally-occurring baseline apoptosis (16 ± 1.3%) (Fig. 2a and b). The observation that control NPs show about 30% baseline apoptosis is consistent with our previous findings (Bieberich et al. 2004; Wang et al. 2005). This dose-dependent apoptosis induction by GA was confirmed by flow cytometry analysis of TUNEL stained NPs (Fig. 2c). By western blotting analysis, 100 nM GA seemed to induce cleavage of caspase 3, while 10 nM GA prevented cleavage of caspase 3 (Fig. 2d-i and d-ii), which confirms the morphology and flow cytometry observations. Caspase 9 was also activated by 100 nM GA, but not by 10 nm GA (Fig. 2d-i and d-ii), indicating that the GA-induced apoptosis in NP might occur via the intrinsic mitochondria apoptosis pathway. These data implies that GA induces a hormetic effect in NPs.

image

Figure 2.  High dose of GA induced apoptosis, while low dose of GA protected NPs from naturally occurring baseline apoptosis. NP cells were treated with GA for 24 h. (a) Cells were fixed and stained with Hoechst 33258 (Hoe). Con, control; GA10, GA 10 nM; GA100, GA 100 nM. (b) Percentage of apoptotic cells in GA-treated NPs. Cells with condensed nuclei and morphology changes were counted as apoptotic. N = 5. (c) Percentage of apoptotic cells in GA-treated NPs measured by flow cytometry after TUNEL staining. N = 3. (d-i) High dose of GA activated caspase 3 and caspase 9, while low dose of GA did not. (d-ii) Densitometry quantification of the bands in (d-i). N = 3, *p < 0.05, error in (b), (c), and (d-ii) denote SD.

Download figure to PowerPoint

Low-dose 17-AAG increased NP cell viability and protected NPs from stress-induced apoptosis

The cell protective effect induced by a low dose of GA should increase the viability of NPs. Since GA is toxic to liver, we performed most of our functional studies using a safer drug, 17-AAG (Waza et al. 2005, 2006). To determine 17-AAG effect on cell viability, we first measured the relative live-NP cell number affected by 48-h 17-AAG treatment. Figure 3(a) shows that the effect of 17-AAG on relative live-NP cell amount is hormetic: 5 nM and 10 nM treatment increased cell number by more than 60%, and 250 nM treatment killed the cells. This is consistent with data obtained from GA treatment (Fig. 2). It is noteworthy that the 24-h treatment of 17-AAG showed a similar effect (data not shown). To evaluate whether the increased-cell-number is a result of enhanced proliferation or not, we did BrdU labeling of the NP cells. Microplate readings of the BrdU incorporation showed that both 17-AAG and GA did not affect cell proliferation (Fig. 3b). These data suggest that the low-dose 17-AAG treatment enhances NP viability.

image

Figure 3.  Low-dose 17-AAG increased NP viability. (a) Relative live-cell counting. NPs were treated with indicated concentrations of 17-AAG for 48 h and the relative number of live cells was measured using a microplate reader. Background (no cells) was subtracted and the reading from 0 nM 17-AAG treatment was set as 100%. 17-AAG concentrations: Con, 0 nM; AAG1, 1 nM; AAG5, 5 nM; AAG10, 10 nM; AAG250, 250 nM. N = 3–5, error bars are SD. (b) Cell proliferation assay. NPs were treated as in (a) and the cell proliferation was measured by a microplate reader. N = 4, p > 0.05. (c-i) High dose of 17-AAG induced NP cell apoptosis, while lower dose did not, evidenced by cleaved caspase 3. (c-ii) Densitometry quantification of the bands in (c-i). N = 3, error bars represent SEM. (d-i) Low-dose 17-AAG protected NECs (prepared from E14 mouse brains) from naturally-occurring apoptosis. (d-ii) Densitometry quantification of the bands in (d-i). N = 3, error bars represent SEM. *p < 0.05, **p < 0.01. Similar Figure denotations apply to the rest of figures, with Figure x-ii being densitometry quantifications of the bands in Figure x-i.

Download figure to PowerPoint

To investigate the mechanism of this hormetic effect mediated by Hsp90 inhibitors, we performed western blot analyses. We found that only high dose 17-AAG induced caspase 3 and caspase 9 activation, while lower dose did the opposite (Fig. 3c). This effect of low-dose 17-AAG was further validated using primarily cultured NECs from E14 mouse brains which shown that low-dose 17-AAG reduced activated caspase 3 (Fig. 3d).

To determine whether 17-AAG induces cross-tolerance in NP cells, the cells were treated with two neurodegenerative disease-related apoptosis inducers: amyloid Aβ peptide and ceramide. Amyloid Aβ is a peptide of 39–43 amino acids that appears to be the main constituent of amyloid plaques in the brains of Alzheimer’s disease and other neurodegenerative disease patients (Hartmann et al. 1997; Tillement et al. 2010). And ceramide accumulation caused by alteration of the sphingolipid metabolism has been linked to neural cell death in Alzheimer’s disease brains (Cutler et al. 2004; Satoi et al. 2005; Wang et al. 2008). Our data shown that 5 μM Aβ peptide (25–35) and 2 μM C16 ceramide induced massive NP cell apoptosis determined by condensed nuclei and morphological changes (Fig. 4a). Excitingly, this cytotoxicity was prevented by treatment with 10 nM 17-AAG (Fig. 4a). This observation was validated by detection of activated caspase3 (Fig. 4b and c). Taken together, these data suggest NPs acquire cross-tolerance after low-dose 17-AAG treatment.

image

Figure 4.  Low-dose 17-AAG protected NPs from stress-induced apoptosis. (a) Percentage of apoptotic cells in β-amyloid and C16 ceramide treated NPs as measured by condensed nuclei and morphological changes. Con: Aβ peptide 35–25; AB: Aβ peptide 25–35; C16, C16 ceramide. (b-i and b-ii) Low-dose 17-AAG protected NPs against Aβ peptide induced apoptosis measured by cleaved caspase 3. βA and BA: Aβ peptide 35–25 as control. (c-i and c-ii) Low-dose 17-AAG protected NPs against C16 ceramide-induced apoptosis measured by cleaved caspase 3. In (a), (b-ii), and (c-ii), N = 3, error bars represent SEM. *p < 0.05, **p < 0.01.

Download figure to PowerPoint

Low dose of Hsp90 inhibitors induced Hsp70 expression, Akt phosphorylation, prolonged GSK-3β phosphorylation and transient ERK1/2 phosphorylation

Hsp90 inhibitors are well known apoptosis induces (Terasawa et al. 2005). However, our data show that they have pro-survival effect when used at lower concentrations. In order to understand how this happens, we investigated in depth its cellular mechanisms. GA or 17-AAG has been shown to induce expression of molecular chaperones Hsp70 and 40, which protect cells against further damages (Gallo 2006; Waza et al. 2006). To determine whether this effect occurred in NPs, cell lysates were subjected to western blot analysis. Figure 5(a) shows that treatment with both 10 nM and 100 nM GA induced over-expression of Hsp70, with a stronger effect with 100 nM GA (Fig. 5a, left panel). The observation that 100 nM GA induced elevation of Hsp70 but still induced apoptosis implies the existence of other cell protection pathways involved in this process. It has been reported that GA stimulates Akt and extracellular signal regulated kinase (ERK) via transient activation of Src kinase (Koga et al. 2006). We tested the phosphorylation of GSK-3β, a substrate of Akt kinase (Bhat et al. 2004; Koga et al. 2006), and found that GSK-3β phosphorylation was up-regulated by 24-h treatment of 10 nM GA, but down-regulated by 100 nM GA (Fig. 5a, left panel).

image

Figure 5.  Low-dose Hsp90 inhibitors activated multiple cell protection pathways. (a-i and a-ii) NPs were treated with GA and 17-AAG for 24 h. Cell lysates were subjected to western blot analysis for phospho-GSK-3β (pGSK-3β), total GSK-3β (tGSK-3β), Hsp70, and β-actin. (b-i and b-ii) NPs were treated with 17-AAG for 24 h. Cell lysates were tested for phospho-Akt (pAkt) and total Akt (tAkt). (c-i and c-ii) NPs were treated with 10 nM 17-AAG for times indicated. Cell lysates were analyzed for pGSK-3β, tGSK-3β, and β-actin. (d) PI3K inhibitors LY290004 and Wortmannin blocked the GSK-3β phosphorylation induced by low-dose 17-AAG *p < 0.05, **p < 0.01.

Download figure to PowerPoint

The regulation of GSK-3β phosphorylation by 17-AAG was investigated in a wider dose range in NPs. We found that GSK-3β phosphorylation was up-regulated by a lower dose of 17-AAG (10 nM, 1 nM has little effect) (Fig. 5a, middle panel) and down-regulated by higher doses (250 and 500 nM) (Fig. 5a, right panel). This hormetic regulation of GSK-3β phosphorylation by 17-AAG is consistent with that of GA. To further understand this dose dependent regulation of GSK-3β phosphorylation by 17-AAG, the protein level and phosphorylation status of Akt were measured. Figure 5(b-i) and (b-ii) show that 250 nM 17-AAG induced Akt degradation (to 37.4 ± 7.5% of control level), which is consistent with previous findings (Zhang and Burrows 2004). Strikingly, 10 nM 17-AAG up-regulated Akt phosphorylation to 175.6 ± 9.4% of control level (Fig. 5b-i and b-ii).

Since GSK-3β activation (dephosphorylation) induces neuronal cell death and is currently a drug target for CNS disease therapies (Bhat et al. 2004; Chen et al. 2004; Aghdam and Barger 2007), we determined the time course of GSK-3β phosphorylation affected by low-dose 17-AAG. We found that the GSK-3β phosphorylation was initiated as early as 1 h and persisted for up to 48 h after cells were treated with 10 nM 17-AAG (Fig. 5c-i and c-ii, 1 h data not shown). To validate that low-dose 17-AAG protected NPs through the phosphotidylinosotol 3-kinase (PI3K)/Akt pathway, we inhibited the PI3K pathway with two inhibitors, LY294002 and Wortmannin. Figure 5(d-i) and (d-ii) show that both LY294002 and Wortmannin blocked GSK-3β phosphorylation mediated by 17-AAG. The above data indicate that low-dose 17-AAG activates the PI3K/Akt pathway, thus exerting a cell protective effect against apoptosis both acutely and for a prolonged time.

Another survival factor, ERK1/2, was transiently phosphorylated by 10 nM 17-AAG within half an hour (Figures S1 and S2). The involvement of MAPK pathway was confirmed by treatment with a Mek inhibitor U0126 (Figures S3 and S4).

To address whether 17-AAG treatment alters the genotype of the NPs, we tested the level of Nestin, a NP cell marker. Data from RT-PCR and immunofluorescence shown that the level of Nestin was not affected (Figure S5 and not shown), indicating low-dose 17-AAG does not affect the genotype of NPs.

17-AAG induced GSK-3β and Akt phosphorylation and increased the number of BrdU positive cells in mouse brains

The in vitro experiments discussed above indicate that treatment with low-dose 17-AAG protects NP cells against apoptosis by activating multiple pro-survival pathways. To evaluate whether 17-AAG would have a similar effect in vivo, we injected 17-AAG (0.2 mg/kg body weight) i.p. into 4-month-old C57BL/6 mice. Whole brain tissues were collected 24 h after the injection and western blot analyses performed to test the GSK-3β and Akt phosphorylation level. Figure 6(a-i) and (a-ii) show that GSK-3β and Akt phosphorylation were up-regulated by 17-AAG (5.3 ± 0.6 and 4.5 ± 0.4-fold, respectively) when normalized to total GSK-3β or Akt level, which is consistent with the in vitro results.

image

Figure 6.  17-AAG induced GSK-3β and Akt phosphorylation and protected newly generated neural cells in mouse brains. Mice were i.p. injected with vehicle control, 0.2 or 1.0 mg/kg body weight 17-AAG. (a-i and a-ii) 24 h after 0.2 mg/kg 17-AAG injection, whole brain tissues were collected and subjected to western blot analysis for pGSK-3β, tGSK-3β, pAkt, and tAkt. N = 3, error bars represent SEM. (b) Representative SVZ images of coronal sections from 17-AAG injected mouse stained with BrdU. Con, vehicle-treated mice; AAG, 17-AAG treated mice. Scale bar, 100 μm. (c) Cell counting of BrdU positive cells in the SVZ and SGZ of the hippocampus of the mouse brain. N = 4, error bars represent SD, *p < 0.05.

Download figure to PowerPoint

To test whether this GSK-3β and Akt phosphorylation translates into protection of newly generated neural cells in vivo, we monitored neurogenesis in 17-AAG injected mice by BrdU. Figures 6(b) and S6 show images of BrdU immunostaining of the SVZ and the SGZ. Stereological cell counting shows that 0.2 mg/kg 17-AAG significantly increased the number of BrdU labeled cells by 1.74 ± 0.2-fold in the SVZ, and 1.83 ± 0.3-fold on the SGZ of the dentate gyrus, when compared to vehicle-injected samples (Fig. 6c). A higher dose 17-AAG (1 mg/kg) did not show a significant effect. There are two possible explanations for how 17-AAG treatment could result in more BrdU positive cells in vivo: the treatment could enhance cell proliferation, or could protect the newly generated cells. The in vitro studies showed 17-AAG did not increase NP cell proliferation (Fig. 3b). These data indicate that 17-AAG most likely protect the newly generated neural cells in the mouse brain. However, since the western blot data shown in Fig. 6(a-i) and (a-ii) was performed using whole mouse brain extract, we believe that not only the newly generated neural cells, but the other brain tissue were also protected.

Discussions

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussions
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

One of the scientific challenges for stem cell therapy in treating neurodegenerative diseases and stroke is the low cell survival rate and functionality of the newly generated or grafted cells (Norgaard et al. 2006; Li et al. 2008). Hormesis, or stress adaptation, provides a new target as chemoprevention or nutritional neuroprotection for these cells (Calabrese et al. 2008b).

The stimulation of bioprotective processes, including anti-apoptosis, antioxidative stress, DNA repair, and immune system activation are believed to underlie the cell hormetic response (Calabrese et al. 2008a,b; Kendig et al. 2010). The Hsp90 inhibitors GA and 17-AAG have been reported to transiently activate osteoclast c-Src signaling and promotes growth of prostate carcinoma cells in bone by stimulating the pro-survival kinases Akt through PI3K (Koga et al. 2006; Yano et al. 2008). In the nervous system, GA has been shown to have neuroprotective effect in ischemic brain (Kwon et al. 2008; Wen et al. 2008). However, no study has been done on their role in neurogenesis and neural stem cell survival, which may lead to a long term functional recovery of neurodegenerative diseases (Hess and Borlongan 2008).

In this study, we sought to explore the possibility of a hormetic effect of 17-AAG in NPs in vitro and the newly generated neural cells in vivo. In consideration of the complexity of the neurodegenerative disorders, Hsp90 inhibitors were chosen in our study since Hsp90 has many client proteins and thus affects multiple signal pathways. We were hoping to identify an agent that can induce cross-tolerance, a more interesting hormetic phenomenon (Norgaard et al. 2006; Wheeler and Wong 2007).

When we treated NP cultures with GA and 17-AAG, we found that these two agents protected NPs form naturally occurring apoptosis at lower doses, although killed NPs at higher doses. Of particular interest is the observation that low-dose 17-AAG protects against ceramide and β–amyloid induced apoptosis, since ceramide accumulation and β-amyloid deposition have been linked to many forms of neurodegenerative disorders(Cutler et al. 2004; Ohtani et al. 2004; Satoi et al. 2005; Wang et al. 2008). We found that this cell protective function of low-dose 17-AAG is mediated by: (i) increased expression of neuro-protective chaperons, such as Hsp70; (ii) activated pro-survival kinases ERK and Akt; and (iii) prolonged phosphorylation and subsequent inactivation of GSK-3β, which regulate cell proliferation and survival. Furthermore, we found that low-dose 17-AAG induced prosurvival factors in vivo, such as phosphorylation of Akt and GSK-3β, which indicate that it not only protects the newly generated cells, but also other nerve tissues. This is an exciting display of hormesis in vivo. Its underlying mechanism will be further studied.

Several anti-tumor agents have been found to prolong survival in neurodegenerative disease mouse models (Ferrante et al. 2004; Ravikumar et al. 2004; Waza et al. 2005, 2006), although their mechanisms underlying the protective effect may vary. Most of these agents have cytotoxic effects on normal cells, which must be overcome prior to any clinical applications against neurodegeneration. To advance the current understanding of 17-AAG, this study found that 17-AAG exposes a dose dependent protection of NP cells by inducing stress adaptation and cross adaptation in NPs. This phenomenon opens an important avenue for treatment of neurodegenerative disorders by 17-AAG; and also could aid in the strategy design in its use for cancer therapy. To further understand the hormetic effect mediated by 17-AAG, we will determine whether low-dose 17-AAG will protect transplanted NP cells in future studies, which is another key aspect of the cell replacement therapies.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussions
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

This work is supported by a Scientist Training Program award from Medical College of Georgia and partly by a Scientist Development Grant award from American Heart Association to GW. We like to thank Drs. Erhard Bieberich, Somsankar Dasgupta, and F. C. Alex Chiu for their suggestions and technical support. We thank Haiyan Qin for her technical support.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussions
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussions
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Figure S1. Low-dose 17-AAG induced ERK activation. NPs were treated with 17-AAG for 24 h. Cell lysates were subjected to western blot analysis for phospho-ERK (pERK) and total ERK (tERK).

Figure S2. Densitometry quantification of the bands in Figure S1. N = 3, error bars represent SEM.

Figure S3. The Mek inhibitor U0126 blocked the GSK-3β phosphorylation induced by low-dose 17-AAG. NPs were treated with indicated for 24 h. Cell lysates were subjected to western blot analysis for pERK and tERK.

Figure S4. Densitometry quantification of the bands in Figure S3. N = 3, error bars represent SEM.

Figure S5. RT-PCR analysis of Nestin in control and 10 nM 17-AAG treated NPs. NPs were treated with 17-AAG for 24 h. Total RNA was extracted and RT-PCR performed for Nestin. Beta actin was used as a loading control.

Figure S6. Low magnification images of the BrdU staining of coronal brain sections from 17-AAG injected (or vehicle control) mice. The BrdU staining is green, and the nuclei staining in SGZ is blue. SVZ, subventricular zone; SGZ, subgranular zone. Scale bars, 200 μm.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JNC_7239_sm_FigsS1-S6.pdf822KSupporting 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.