J. Neurochem. (2012) 120, 419–429.
Mitochondrial dysfunction is a prominent feature of Alzheimer’s disease (AD) brain. Our prior studies demonstrated reduced mitochondrial number in susceptible hippocampal neurons in the brain from AD patients and in M17 cells over-expressing familial AD-causing amyloid precursor protein (APP) mutant (APPswe). In the current study, we investigated whether alterations in mitochondrial biogenesis contribute to mitochondrial abnormalities in AD. Mitochondrial biogenesis is regulated by the peroxisome proliferator activator receptor gamma-coactivator 1α (PGC-1α)-nuclear respiratory factor (NRF)-mitochondrial transcription factor A pathway. Expression levels of PGC-1α, NRF 1, NRF 2, and mitochondrial transcription factor A were significantly decreased in both AD hippocampal tissues and APPswe M17 cells, suggesting a reduced mitochondrial biogenesis. Indeed, APPswe M17 cells demonstrated decreased mitochondrial DNA/nuclear DNA ratio, correlated with reduced ATP content, and decreased cytochrome C oxidase activity. Importantly, over-expression of PGC-1α could completely rescue while knockdown of PGC-1α could exacerbate impaired mitochondrial biogenesis and mitochondrial deficits in APPswe M17 cells, suggesting reduced mitochondrial biogenesis is likely involved in APPswe-induced mitochondrial deficits. We further demonstrated that reduced expression of p-CREB and PGC-1α in APPswe M17 cells could be rescued by cAMP in a dose-dependent manner, which could be inhibited by PKA inhibitor H89, suggesting that the PKA/CREB pathway plays a critical role in the regulation of PGC-1α expression in APPswe M17 cells. Overall, this study demonstrated that impaired mitochondrial biogenesis likely contributes to mitochondrial dysfunction in AD.
amyloid precursor protein
nuclear respiratory factor
peroxisome proliferator activator receptor gamma-coactivator 1α
mitochondrial transcription factor A
Alzheimer’s disease (AD), the most common neurodegenerative disorder of the elderly, is characterized by senile plaques, neurofibrillary tangles, and progressive death of neuronal cells at selective brain regions (Smith 1998). The molecular mechanisms underlying the pathogenesis of AD are still unclear. Mitochondrial dysfunction is a prominent and early feature of AD (Zhu et al. 2006) with reduced energy metabolism as one of the best documented early abnormalities in AD (Blass 2000). Key mitochondrial enzymes of oxidative metabolism (i.e. cytochrome C oxidase, α-ketoglutarate dehydrogenase complex and pyruvate dehydrogenase complex) are deficient in AD (Manczak et al. 2004; Qin et al. 2009). Early deficits in synaptic mitochondria were found in an AD mouse model in a recent study (Du et al. 2010). In addition, damaged mitochondrial DNA (mtDNA) including DNA mutation and DNA defects are also found to be involved in AD (Lin et al. 2002). Both amyloid precursor protein (APP) and amyloid-β (Aβ) locate in mitochondria (Anandatheerthavarada et al. 2003; Caspersen et al. 2005; Manczak et al. 2006; Hansson Petersen et al. 2008; Chen and Yan 2010; Calkins et al. 2011) and Aβ not only contributes to significant oxidative damage of mtDNA, but also leads to impaired mtDNA gene expression (de la Monte et al. 2000; Brooks et al. 2007). More recent studies suggest abnormal mitochondrial dynamics including excessive mitochondrial fragmentation and abnormal mitochondrial distribution plays a critical role in mitochondrial dysfunction in AD (Su et al. 2010; Wang et al. 2010, 2008a,b, 2009; Manczak et al. 2011).
During the life cycle of mitochondria, mitochondrial biogenesis plays an essential role in maintaining mitochondrial homeostasis to meet the physiological needs of eukaryotic cells. However, the status of this important aspect of mitochondrial life and function in AD is unclear (Onyango et al. 2010). The factors regulating mitochondrial biogenesis include nuclear respiratory factor 1 (NRF 1) and nuclear respiratory factor 2 (NRF 2), which control the nuclear genes to encode mitochondrial protein, and mitochondrial transcription factor A (TFAM), which drives transcription and replication of mtDNA. The expression of NRF 1, NRF 2, and TFAM are regulated by peroxisome proliferator activator receptor gamma-coactivator 1α (PGC-1α) (Wu et al. 1999), the master regulator of mitochondrial biogenesis. A recent study showed reduced PGC-1α levels in AD (Qin et al. 2009), implicating an impaired mitochondrial biogenesis signaling in AD. Interestingly, our prior studies demonstrated reduced mitochondrial number in susceptible hippocampal neurons in the brain from AD patients (Hirai et al. 2001) and in M17 cells over-expressing familial AD (FAD)-causing APP mutant (Wang et al. 2008b). To explore whether a reduced mitochondrial biogenesis contributes to reduced mitochondria and mitochondrial dysfunction in AD, we investigated the profile of mitochondrial biogenesis signals in AD cases and in M17 cells over-expressing FAD-causing APP mutant and how they may affect mitochondrial function.
Materials and methods
Brain tissue from the hippocampal and cerebellum regions was obtained postmortem using approved IRB protocol from AD (n = 15; ages = 65–91; postmortem interval = 3.6 ± 0.4 h) and control (n = 14; ages = 53–91; postmortem interval = 3.3 ± 0.5 h) patients, as assessed on clinical and pathological criteria established by CERAD and an NIA consensus panel (Khachaturian 1985; Mirra et al. 1991).
Cell culture, transfection, and treatment
M17, a widely used human neuroblastoma cell line, was maintained in Opti-MEM medium (Invitrogen, Eugene, OR, USA), supplemented with 5% fetal bovine serum and 1% penicillin–streptomycin, in a humid incubator with 5% CO2 at 37°C. The cell lines stably expressing human Swedish mutation (K670M/N671L) APP695 or control vector were described in the previous research (Wang et al. 2008b). The micro-RNA (miR) RNA interference (RNAi) vector was generated via the pcDNA 6.2-GW/EmGFP-miR constructs (Invitrogen). 50 μg/mL blasticidin (Invivogene biotech, Eugene, OR, USA) was used for PGC-1α knockdown stable line selection. The control vectors and plasmids of PGC-1α were reported in previous studies respectively (Puigserver et al. 1998; Lerin et al. 2006). Transfection was performed using lipofectamine 2000 following the instructions (Invitrogen). Stable cell lines were maintained with 100 μg/mL of geneticin, 200 μg/mL of hygromycin or 25 μg/mL of blasticidin. APPswe cells were treated and incubated for 48 h with 0.5, 1, 5, 10 mM cAMP. Cells were also stimulated by cAMP at 5 mM, with or without 10 μM PKA inhibitor H89 (Cell Signaling, Beverly, MA, USA).
Real-time PCR to detect mtDNA copy number
mtDNA copy number was measured by real-time PCR method using the Step one plus Real-Time PCR System (Applied Biosystems, Carlsbad, CA, USA) with the SYBR Green detection method. Total intracellular DNA in different cell lines was extracted using QIAamp DNA mini kit (QIAGEN, Germantown, MD, USA) following the manufacturer’s instruction. Each real-time PCR (20 μL total volume) contained 3 μL of template DNA (50 μg), 10 μL of 2× SYBR Green Real-time PCR Master (Applied Biosystems), 1 μL of each of the forward and reverse primers and 5 μL of ultrapure water. The primers for a subunit of human electron transport chain used for mtDNA amplification were: forward, 5′-CAAACCTACGCCAAAATCCA-3′ and reverse, 5′-GAAATGA ATGAGCCTACAGA-3′ (Kukidome et al. 2006; Xu et al. 2009). Primers for Human nuclear 18S, used for internal control, were: forward, 5′-ACGGACCAGAGCGAAAGCA-3′ and reverse, 5′-GACATCTAAGGGCATCACAGAC-3′. Relative amounts of mtDNA and nuclear DNA (nDNA) copy numbers were compared. The ddCt (mtDNA to 18S) represents the mtDNA copy number in a cell.
Brain samples from hippocampus and cerebellum of AD and age-matched control or cells were lysed with 1× cell lysis buffer (Cell Signaling), plus 1 mM phenylmethylsulfonyl fluoride (Sigma, St Louis, MO, USA) and Protease Inhibitor Cocktail (Sigma). Bicinchoninic acid assay (Pierce, Rockford, IL, USA) was used to determine protein concentration. Proteins were separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and were electrotransferred to Immobilon-P (Millipore, Billerica, MA, USA). After blocking with 10% non-fat dry milk, transferred membrane was incubated with primary and secondary antibodies (Zhu et al. 2000). The following antibodies were used in this study: PGC-1α (C terminal) from Calbiochem (San Diego, CA, USA); NRF 1, NRF 2α, NRF 2β, TFAM, and cytochrome B from Santa Cruz Biotechnology (Santa Cruz, CA, USA); Two close bands were recognized by NRF2β antibody which represented the two splice variants that differ in their C-termini. Lamin A/C, p-CREB, and CREB from Cell Signaling; PKA RI, PKA RIIβ, and PKAc from BD Laboratories (Franklin Lakes, NJ, USA); actin from Chemicon (Atlanta, GA, USA). Blots were developed with Immobilon Western Chemiluminescent HRP Substrate (Millipore) or the immunoCruz (Santa Cruz Biotechnology).
Cell nuclear protein extraction
Cell nuclear proteins were extracted using Nuclear and Cytoplasmic Extraction Kit (Thermo Scientific, Rockford, IL, USA) according to the instructions.
ATP and cytochrome C oxidase activity
ATP levels were measured by the ATP Determination Kit (Invitrogen), which was described in the previous study (Wang et al. 2008b). Cytochrome C oxidase activity was determined using the cytochrome C oxidase activity kit (Sigma) following the instructions.
Mitochondrial biogenesis proteins are reduced in hippocampal tissues from AD patients
As mitochondrial biogenesis is regulated by the PGC-1α-NRF 1-TFAM pathway, we first measured the expression level of PGC-1α in hippocampal tissues. Real-time PCR studies of 15 AD and 10 age-matched controls suggested a 50% decrease in PGC-1α mRNA levels in AD (not shown). Immunoblot analysis also revealed significantly reduced total protein levels of PGC-1α in AD brains (Fig. 1a). As PGC-1α is a transcription factor, we also measured the nuclear levels of PGC-1α protein. Immunoblot of nuclear fraction revealed significantly reduced PGC-1α levels in AD brain compared with age-matched controls, suggesting this pathway may be impaired in AD brain (Fig. 1b). Indeed, further immunoblot analysis of proteins downstream to PGC-1α, that is, NRF 1, NRF 2α, NRF 2β, and TFAM, confirmed significant reduction in the levels of all four downstream proteins in AD brain compared with age-matched normal controls: the levels of NRF 1, NRF 2α, NRF 2β, and TFAM were reduced by 38.2% (p < 0.01), 53.3% (p < 0.01), 40.7% (p < 0.01), and 43% (p < 0.01), respectively in AD samples (Fig. 1c). However, in contrast to that in hippocampus, no significant decrease in any of these proteins was found in cerebellum, the unaffected brain region, from AD patients (not shown). These data suggest a reduced mitochondrial biogenesis in affected brain regions in AD.
Reduced mitochondrial biogenesis causes mitochondrial dysfunction in PGC-1α knockdown cells
We hypothesize that reduced mitochondrial biogenesis contributes to mitochondrial dysfunction in AD neurons. To study the effect of reduced PGC-1α and downstream effectors on mitochondrial function in neuronal cells, a mixed clone (KDm) and a single clone, clone 4 (KD4), of M17 cells stably transfected with PGC-1α miR RNAi were established (Fig. 2a). Immunoblot analysis confirmed reduced expression of PGC-1α in both cell lines. The levels of NRF 1, NRF 2β, and TFAM were also reduced in both PGC-1α KDm and KD4 cell lines, whereas the level of NRF 2α was constant (Fig. 2b).
To directly determine whether mitochondrial biogenesis is affected, we measured mtDNA relative to nDNA (mtDNA/nDNA) by real-time PCR and found that mtDNA/nDNA was significantly reduced in PGC-1α KDm and KD4 cells (Fig. 2c). Moreover, reduced expression level of mitochondrial protein, cytochrome B, was also found in these cells (Fig. 2d).
Mitochondria are the major source of ATP. We investigated the effect of reduced PGC-1α on ATP production by firefly fluorescence assay and found reduced ATP levels in PGC-1α KDm and KD4 cells compared with EGFP controls (Fig. 2e), which correlates with an obvious reduction of cytochrome C oxidase activity in these cells (Fig. 2f), suggesting that impaired mitochondrial biogenesis negatively impacts mitochondrial function in neurons.
Impaired mitochondrial biogenesis in APPswe M17 cells
Mutations in the APP gene including the Swedish mutation (APPswe) cause FAD. We previously demonstrated that over-expression of APPswe causes significant decrease in mitochondrial number and a slight but significant increase in mitochondrial size (Wang et al. 2008b). In the current study, we investigated whether impaired mitochondrial biogenesis could contribute to decreased mitochondrial number in APPswe M17 cells.
Compared with wild-type (WT) or vector-control M17 cells, APPswe M17 cells demonstrated significantly reduced levels of nuclear PGC-1α protein (Fig. 3a). Immunoblot and quantification analysis also revealed that the protein levels of NRF 1, NRF 2β, and TFAM were reduced by more than 50% in APPswe cells compared with WT and vector-control M17 cells (p < 0.01), whereas the level of NRF 2α did not change significantly (Fig. 3b).
Consistent with a reduced mitochondrial biogenesis signaling, we also found that mtDNA/nDNA ratio (Fig. 3c), along with expression level of mitochondrial protein, cytochrome B (Fig. 3d), was significantly reduced in APPswe M17 cells. Functionally, APPswe M17 cells displayed significant reduced ATP level and cytochrome C oxidase activity compared with WT and vector-control M17 cells (Fig. 3e and f).
Manipulation of PGC-1α expression modulates mitochondrial biogenesis and dysfunction in APPswe M17 cells
To assess whether PGC-1α reduction played a causal role in the abnormalities in mitochondrial biogenesis and function in APPswe M17 cells, we investigated whether over-expression of PGC-1α rescues APPswe M17 cells. APPswe M17 cells were transiently transfected with PGC-1α plasmid. Forty-eight hours after transfection, expression of mitochondrial biogenesis proteins and mitochondrial parameters were determined. Over-expression of PGC-1α was confirmed by immunoblot (Fig. 4a). Indeed, expression of the downstream proteins including NRF 1, NRF 2β, and TFAM was also significantly increased (*p < 0.05, Δp < 0.01, Student’s t-test) in APPswe cells over-expressing PGC-1α (Fig. 4b). Consistently, mtDNA/nDNA (Fig. 4c) and expression of cytochrome B (Fig. 4d), were also significantly increased to the level comparable to WT M17 cells, suggesting the restoration of mitochondrial biogenesis. Most importantly, the over-expression of PGC-1α led to the rescue of mitochondrial function as reflected by increased ATP level (Fig. 4e) along with increased cytochrome C oxidase activity (Fig. 4f) in APPswe M17 cells.
APPswe M17 cells were also transfected with the miR RNAi of PGC-1α. Reduced expression of PGC-1α was confirmed by immunoblot (Fig. 5a). As expected, down-regulation of PGC-1α led to significantly reduced levels of NRF 1, NRF 2β, and TFAM (Fig. 5b) along with further reduction in mtDNA/nDNA ratio (Fig. 5c), and decreased cytochrome B levels (Fig. 5d) in APPswe M17 cells. Functionally, this led to further reduction in ATP content (Fig. 5e) as well as cytochrome C oxidase activity (Fig. 5f).
PKA/CREB pathway plays a critical role in the reduced expression of PGC-1α in APPswe cells
The expression of PGC-1α is regulated by transcription factor CREB (Lonard and O’Malley 2006). Indeed, we found that the level of CREB phosphorylated at Ser133 was significantly decreased in APPswe M17 cells (Fig. 6a), suggesting reduced activity of CREB in these cells. CREB can be phosphorylated and activated by the PKA pathway. PKA is composed of catalytic subunits (C) and regulatory subunits (R, including RI and RII), with the regulatory units blocking the catalytic centers of the catalytic units. Indeed, immunoblotting and quantification analysis revealed significantly higher levels of PKA RI but no changes in catalytic subunit in APPswe cells (Fig. 6b), suggesting that reduced PKA activity may be involved in the decreased CREB phosphorylation/activation and PGC-1α in APPswe M17 cells. To corroborate the role of PKA, we investigated the effect of cAMP, a PKA activator, on the phosphorylation of CREB and expression of PGC-1α in APPswe M17 cells. APPswe M17 cells were harvested after being treated with various concentrations of cAMP for 48 h and CREB and PGC-1α levels were determined by immunoblot. Indeed, in the concentration range used in this study, cAMP induced a dose-dependent increase in the levels of both p-CREB and PGC-1α in APPswe M17 cells (Fig. 6c and d). Furthermore, cAMP-induced phosphorylation of CREB (Fig. 6e) and increased expression of PGC-1α (Fig. 6f) were completely abolished by 10 μM H-89, a specific PKA inhibitor, thus demonstrating that PKA played a critical role in the reduced expression of p-CREB and PGC-1α in APPswe M17 cells.
Mitochondria dysfunction plays an important role in AD, but the mechanism is still unclear (Zhu et al. 2004; Reddy and Beal 2005, 2008; Swerdlow 2009; Swerdlow et al. 2010). In this study, we demonstrated significant alterations in the expression pattern of proteins involved in mitochondrial biogenesis in hippocampal tissues from AD brain compared with age matched control brain. We found that the expression levels of NRF 1, NRF 2, and TFAM along with nuclear levels of PGC-1α are significantly reduced in AD hippocampus, suggesting that mitochondrial biogenesis is likely impaired in AD. Consistent with this notion, we also found reduced expression levels of PGC-1α, NRF 1, NRF 2, and TFAM along with significantly reduced mtDNA/nDNA ratio and mitochondrial contents in M17 cells over-expressing FAD-causing the APPswe mutant. Importantly, over-expression of PGC-1α almost completely restored, while knockdown of PGC-1α exacerbated, mitochondrial biogenesis deficits and mitochondrial dysfunction in APPswe M17 cells, suggesting reduced mitochondrial biogenesis likely contributes to mitochondrial dysfunction. Moreover, we further demonstrated reduced PKA/CREB signaling in APPswe M17 cells and specific activation of this pathway restored PGC-1α levels and suggest that the impaired PKA/CREB signaling plays an important role in the reduction of PGC-1α and mitochondrial biogenesis in APPswe M17 cells.
Previously, we found a slight but significant reduction in mitochondrial number in hippocampal neurons from biopsied AD brain (Hirai et al. 2001). Similarly, significantly reduced mitochondrial number was also reported in M17 cells over-expressing the APPswe mutant (Wang et al. 2008b). Although alterations in mitochondrial dynamics and/or mitophagy may contribute to changes in mitochondrial number and mass (Moreira et al. 2007), changes in mitochondrial biogenesis could also impact such mitochondrial parameters. Given that PGC-1α is the key convergent component of the complex regulation pathways of mitochondrial biogenesis (Wenz 2009), such a notion is further supported by the recent finding that PGC-1α expression decreases in the AD brain as a function of dementia (Qin et al. 2009). However, given the pleotrophic roles of PGC-1α, it remains to be determined whether and how the mitochondrial biogenesis signaling is altered and whether such alterations contribute to mitochondrial dysfunction in AD. In this study, we not only confirmed reduced nuclear PGC-1α in AD brain, but further demonstrated decreased expression of NRF 1, NRF 2, and TFAM, factors downstream of PGC-1α in AD brain and AD cell models, suggesting that the mitochondrial biogenesis signaling is indeed impaired in AD. A reduced mitochondrial biogenesis is further indicated by reduced mtDNA/nDNA ratio and reduced levels of critical mitochondrial contents in APPswe M17 cells. Consistent with our prior studies, APPswe M17 cells demonstrated mitochondrial dysfunction as reflected by reduced ATP production. To explore the causal relationship between mitochondrial biogenesis deficits and mitochondrial dysfunction, we over-expressed PGC-1α in APPswe M17 cells which almost completely restored the expression of NRF 1, NRF 2, and TFAM and mtDNA/nDNA ratio and mitochondrial contents, indeed, this also resulted in the attenuation of mitochondrial dysfunction in APPswe M17 cells. In fact, another group also reported that over-expression of TFAM could protect SH-SY5Y cells from Aβ-induced mitochondrial dysfunction (Xu et al. 2009). However, knockdown of PGC-1α exacerbated mitochondrial biogenesis deficits and mitochondrial dysfunction in APPswe M17 cells. Overall, these data suggest that impaired mitochondrial biogenesis is likely involved in the mitochondrial dysfunction and play an important role in the pathogenesis of AD. In this regard, it is notable that genes encoding proteins involved in mitochondrial biogenesis have been associated with AD. For example, TFAM locates on chromosome 10, on which several genes were reported to be linked with sporadic AD (Blacker et al. 2003). The moderate possible risk of AD associated with TFAM genotypes and haplotype was found (Gunther et al. 2004). Evidence also showed that variations in TFAM were involved in the pathogenesis of sporadic late-onset AD in the Han Chinese population (Zhang et al. 2011).
Increased mitochondrial biogenesis through PGC-1α activation has been suggested to be a potential therapeutic approach for mitochondrial disorders in cell and animal models of mitochondrial disease (Wenz et al. 2008; Srivastava et al. 2009). This study supports PGC-1α as a potential therapeutic target to increase mitochondrial biogenesis and improve mitochondrial function in AD. Indeed, such a notion is supported by a recent study suggesting that sirtuin 1 and resveratrol, a sirtuin 1-activating molecule, confers significant neuroprotection in a mouse model of AD and tauopathies likely through inhibition of PGC-1α acetylation and promotion of PGC-1α activity (Kim et al. 2007). In this regard, it is important to understand how PGC-1α levels are regulated in APPswe M17 cells. The proximal promoter of PGC-1α contains a functional CREB binding site and PGC-1α levels are regulated by the activity of CREB (Herzig et al. 2001). We found that although the overall levels of CREB remain unchanged, the level of active CREB phosphoryated at Ser133 is significantly reduced in APPswe M17 cells, correlating with reduced PGC-1α expression in these cells. Furthermore, we demonstrated increased levels of PKA regulatory subunits in APPswe M17 cells, suggesting a likely reduced PKA activity in these cells. The fact that the treatment of PKA activator, cAMP, led to increased p-CREB and PGC-1α in a dose-dependent manner which could be completely abolished by specific PKA inhibitor H89 suggesting that the PKA/CREB pathway plays an important role in reduced expression of PGC-1α in APPswe M17 cells. Indeed, treatment of H89 could lead to further decreased levels of p-CREB and PGC-1α in APPswe M17 cells. Perhaps, not coincidentally, CREB protein level and phosphorylation level are also decreased in AD and a reduction in phosphorylation of CREB resulting from inactivation of PKA has also been demonstrated in the postmortem brains of patients with AD (Yamamoto-Sasaki et al. 1999), suggesting that reduced PKA/CREB signaling may also be responsible for reduced PGC-1α expression in AD brain. Therefore, it may be desirable to activate the PKA/CREB signaling pathway as therapeutics to increase the levels or activity of PGC-1α to improve mitochondrial biogenesis and mitochondrial function in AD. Nevertheless, it must be cautioned that PKA is also involved in the hyperphosphorylation of tau protein in AD. Therefore, the activation of PKA to stimulate PGC-1α-dependent mitochondrial biogenesis may cause complication.
Overall, this study demonstrated reduced mitochondrial biogenesis signaling in AD brain and AD cell model. Induction of mitochondrial biogenesis alleviates mitochondrial dysfunction. Based on these data, we suggest that impaired mitochondrial biogenesis contributes to mitochondrial dysfunction in AD and enhancing mitochondrial biogenesis may represent a potential pharmacologic approach for the treatment of AD.
This work was supported by the Dr Robert M. Kohrman Memorial Fund, the National Institutes of Health (AG031852 to XWZ), and the Alzheimer’s Association (IIRG-10-173471 to GP; IIRG-10-173358 to XWZ). Dr Zhu was a consultant for and received grant support from Medivation. Dr Perry is a consultant for Takeda Pharmaceuticals and Neurotez and owns stock options in Neurotez and Voyager. No other authors have any disclosures.