Physiological or α-processing of amyloid-β precursor protein (APP) prevents the formation of Aβ, which is deposited in the aging brain and may contribute to Alzheimer's disease. As such, drugs promoting this pathway could be useful for prevention of the disease. Along this line, we searched through a number of substances and unexpectedly found that a group of high-energy compounds (HECs), namely ATP, phosphocreatine, and acetyl coenzyme A, potently increased APP α-processing in cultured SH-SY5Y cells, whereas their cognate counterparts, i.e., ADP, creatine, or coenzyme A did not show the same effects. Other HECs such as GTP, CTP, phosphoenol pyruvate, and S-adenosylmethionine also promoted APP α-processing with varying potencies and the effects were abolished by energy inhibitors rotenone or NaN3. The overall efficacy of the HECs in the process ranged from three- to four-fold, which was significantly greater than that exhibited by other physiological stimulators such as glutamate and nicotine. This suggested that the HECs were perhaps the most efficient physiological stimulators for APP α-processing. Moreover, the HECs largely offset the inefficient APP α-processing in aged human fibroblasts or in cells impaired by rotenone or H2O2. Most importantly, some HECs markedly boosted the survival rate of SH-SY5Y cells in the death process induced by energy suppression or oxidative stress. These findings suggest a new, energy-dependent regulatory mechanism for the putative α-secretase and thus will help substantially in its identification. At the same time, the study raises the possibility that the HECs may be useful to energize and strengthen the aging brain cells to slow down the progression of Alzheimer's disease.
A prominent feature of brain aging and Alzheimer's disease (AD) is the deposition of Aβ, a proteolytic fragment of amyloid-β precursor protein (APP). Physiological or α-processing of APP prevents the formation of Aβ (Esch et al. 1990; Walsh and Selkoe 2004), but this pathway is somehow inactivated during aging with a concomitant increase of Aβ, reflecting a proteostasis failure (Lannfelt et al. 1995; Kern et al. 2006). It has been found that activation of APP α-processing inevitably decreases Aβ, and vice versa (Skovronsky et al. 2000; Frautschy et al. 1998; Etcheberrigaray et al. 2004). As such, drugs promoting APP α-processing would potentially benefit the aging brain. Along this line, a range of compounds have been found to exhibit such effects including growth factors, hormones, cytokines, and stimulators of neurotransmitter receptors and signaling transduction pathways. These compounds have been under intense studies to evaluate their therapeutic values (Lichtenthaler 2011; Sun and Alkon 2010; reviews).
In our laboratory, we have also tested a number of substances for their potencies to promote APP α-processing. These substances include, among others, current and experimental medications, herbal extracts, and nutritional supplements. The search has led us to unexpectedly find that a group of high-energy compounds (HECs) is highly effective in promoting APP α-processing. Subsequent experiments showed that they also boosted the survival of cultured neuronal cells impaired by energy inhibitors or oxidative stress.
High-energy compounds contain an energy-rich bond of either ~phosphoryl or ~acyl group, which upon hydrolysis liberates at least 7 kilocalories per mole of free energy (ΔGo) under standard conditions. There are many HECs in the body, the best known of which are adenosine, guanosine, or cytidine triphosphate (ATP, GTP, and CTP), phosphocreatine (PCr), acetyl-coA, phosphoenol pyruvate (PEP), and S-adenosylmethionine (SAM) (Lehninger 1975; Nicholls and Ferguson 2002). In CNS, free energy is the essential driving force for synthesis of macromolecules, fast axonal and dendritic transport, and maintenance of ion gradients (McDaniel et al. 2003; Bonda et al. 2009).
Materials and methods
Human neuroblastoma SH-SY5Y and human skin fibroblasts from young and aged donors were purchased from ATCC (Gaithersburg, MD, USA). ATP, GTP, CTP, PCr, acetyl-coA, PEP, ADP, GDP, Cr, CoA, PEP, CTP, SAM, glutamate, and antibodies to N-terminus of APP (22c11) and β-actin were from EMD Millipore (Temecula, CA, USA). Pre-cast sodium dodecyl sulfate–polyacrylamide gel electrophoresis gels were from Invitrogen (Carlsbad, CA, USA). NaN3, rotenone, MTT toxicity assay kit, Dulbecco's modified Eagle's medium (DMEM), and other chemicals were from Sigma-Aldrich (St. Louis, MO, USA).
Cell culture and APP secretion assays
SH-SY5Y and fibroblast cells were routinely maintained in DMEM supplemented with 10% bovine fetal serum and 100 μg/ml of streptomycine and 100 U/mL of penicillin. For APP secretion assay, the cells were subcultured in the 12-well plate until confluence and then pre-incubated in the serum-free DMEM for 1 h before the addition of the testing agents. After incubation for 2 h at 37°C, conditioned medium was collected and cleared by centrifugation. Proteins were precipitated by trichloroacetic acid and processed for gel electrophoresis and western blotting as described previously (Weidemann et al. 1989; Chen and Fernandez 2004). A near-infrared dye-conjugated secondary antibody (LI-COR Biosci., Lincoln, NE, USA) was used and visualized using a LI-COR Odyssey Infrared Imaging System.
Cell survival and viability MTT assay
SH-SY5Y cells were subcultured in the 12-well plate in DMEM containing 10% fetal bovine serum until 50% confluency. For experiment, the medium was replaced with serum-free DMEM and testing agents were added. The culture was continued in the incubator and parallel samples were taken out at various time intervals for microscopic examination or cell viability assay using an MTT toxicity assay kit (Sigma) according to the manufacturer's instructions with modifications (Chen and Fernandez 2004).
Evaluation of the differences in fold increase of APP secretion was carried out using the computer-assisted paired Student t-test and cell viability was calculated using anova with post hoc test.
HECs promote APP α-processing
To find effective compounds that may benefit the aging brain, we tested a number of substances by measuring their effectiveness in promoting the release of the secreted form of APP by α-processing (sAPPα) from cultured SH-SY5Y neurons. The tested substances included current or experimental drugs, herbal extracts, nutritional supplements, and other compounds. These experiments revealed unexpectedly that among the substances, ATP was highly effective in promoting sAPPα release (Fig. 1). As APP α-processing has been found to be sensitive to energy perturbations (Gabuzda et al. 1994; Webster et al. 1998; Hoyer et al. 2005), the finding prompted us to speculate that the observed effects of ATP could be attributable to its high energy content. Subsequent experiments confirmed this speculation.
It was found that, similar to ATP, two other HECs, acetyl-coA and PCr, also exhibited strong and dose-related effects on APP α-processing. As shown, these three HECs at 1 mM each promoted sAPPα release by 3.3-, 3.8-, and 4.2-fold relative to basal line, respectively (Fig. 1 a and b). The presence of high-energy bonds in these compounds was necessary for the observed effects as their cognate counterparts without the same energy content, namely ADP, creatine (Cr), or CoA, did not show the same effects (Fig. 1b, CoA not shown). Of note was that acetyl-coA possesses an ~acetyl (thioester) energy-rich bond (Su and Abumrad 2009).
Further studies showed that several other HECs, namely PEP, GTP, CTP, and SAM (1 mM each), also exerted similar effects on APP secretion (Fig. 2a). Overall, they displayed varying magnitudes of efficacy that ranged from three- to more than four-fold when tested at the same concentration. ATP seemed to be the most efficient one with an EC50 = 0.48 mM (Fig. 2b). Such effectiveness was significantly greater than that of glutamate (2.3-fold) (Fig. 2a), nicotine, or bradykinin (not shown).
Nevertheless, their stimulatory effects were completely abolished when these HECs were tested in the presence of respiratory chain inhibitor rotenone (20 μM), or cytochrome oxidase inhibitor NaN3 (25 mM) (Fig. 2a, showing their effects on ATP only, the effects on other HECs not shown). Similar effects on APP secretion were also observed when ATP, GTP, and PCr were tested in cultured human kidney cells HEK293 (not shown) or human skin fibroblasts (see below). Thus, the observed effects of the HECs were presumably a general phenomenon in various cell types.
HECs offset the effects of age-related impairments
Mitochondrial function and the HECs levels are decreased in the aging brain and AD (Owen and Sunram-Lea 2011; Foy et al. 2011; Meier-Ruge and Bertoni-Freddari 1999), and the resulting energy crisis underlies various age-related changes including decreased APP α-processing (Lannfelt et al. 1995; Kern et al. 2006). We compared sAPPα release in human skin fibroblast lines obtained from five young (age 17–29) and five aged donors (age 81–96). It was found that the average level of sAPPα release from aged cells decreased to 42.5% of the young cell control in the resting state (Fig. 3a and b). Interestingly, ATP or PEP (0.5 mM each) enhanced the sAPPα release in aged cells to 82.3% and 79.6% of the young-cell level, respectively (Fig. 3a and b; effects of PCr not shown in a).
To see whether the inefficient APP α-processing in aged cells could be linked to energy suppression, we tested the effects of rotenone in the young cells. As shown, treating young cells with 0.2 μM rotenone for 2 h reduced the sAPPα level to 52.3% of the control. And importantly, this reduced level was largely restored by cotreatment of 1.5 mM each of ATP or PEP, which brought the sAPPα level back to 79.2% and 90.3% of the control, respectively (Fig. 4, left panel).
Dysfunctional mitochondria and energy depletion in the aging brain occur with concomitant free-radical accumulation (Bonda et al. 2009). We found that treating young cells with 0.2 mM H2O2 for 2 h decreased sAPPα release to 46.1% of the young cell control and, again, this effect of H2O2 was largely attenuated by cotreatment of the cells with 1.5 mM each of ATP or ACoA, which increased the sAPPα level to 89.3 and 74.5% of the young cell control, respectively (Fig. 4, right panel).
HECs promote cell survival
If the HECs offset age-related impairments to APP processing, it would be possible that the HECs also slow down the cell death induced by the same impairments. This possibility was examined in the death process of SH-SY5Y cells induced by rotenone, an agent that causes neuronal death in cultured cells and animal models for AD and Parkinson's disease (Chaves et al. 2010; Höglinger et al. 2005). In our hands, 0.2 μM rotenone caused a progressive neuronal death with a concomitant cell shrinkage and disappearance of the neuronal processes, as visualized under a microscope (Fig. 5a, upper panel). However, pre-incubation of the cells with PEP (1 mM) for 30 min prior to the addition of rotenone substantially preserved the cell shape and the processes (Fig. 5a, lower panel). A quantitative cell viability assay showed that rotenone alone caused all cells to die by 2 days, but pre-incubation of PEP, PCr, or ATP (1 mM each) enhanced the cell survival to 53%, 42%, or 15% of the untreated control, respectively (Fig. 5b). Meanwhile, acetyl-coA and GTP (1 mM each) also increased cell survival to 35% and 17% under the same experimental conditions (not shown).
Finally, we tested the effects of the HECs on cell death induced by H2O2 and found that while 0.6 mM H2O2 killed all cells by 10 h, pre-treatment with 1 mM each of PCr, PEP, or acetyl-coA considerably increased the cell survival to 49%, 38%, and 21% of the control, respectively, by 20 h (Fig. 5c). At the same time, ATP or GTP also moderately increased cell survival (not shown).
Thus, these HECs displayed significant neuroprotective effects in cultured cells, although with varying degrees of efficacy. Overall, it appeared that PCr, PEP, or acetyl-coA were more efficient than ATP and GTP under the experimental conditions. Although the reasons were unclear, it is possible that this may relate to the energy levels they released, durability of the actions in the cell, or compatibility with the specific cell type.
Free energy in APP processing
In this study, we found that HECs promote sAPPα secretion by three- to four-fold, an efficacy that is greater than that reported for other physiological stimulators, such as glutamate, nicotine, bradykinin, and growth factors (averaged two- to three-fold) (Jolly-Tornetta et al. 1998; Kim et al. 1997; Mills et al. 1997; Solano et al. 2000; Fisher et al. 2003). It thus appears that these HECs are perhaps the most efficient physiological stimulators for the APP secretory pathway when tested in vitro, although perhaps less efficient than phorbol esters (three- to six-fold reported in various cell types) (Efthimiopoulos et al. 1994; Bauxbum et al. 1990).
The HECs may exert their observed actions through diverse pathways. For example, ATP and GTP may interact with a family of purinergic receptors (P2Y) and activate the G protein-related signal transduction cascades (Abbracchio et al. 2009). PCr and Cr are taken up into cell by a sodium-dependent transporter and serve as a primary energy reserve system in the body (Adhihetty and Beal 2008). Acetyl-CoA can directly enter mitochondria and affects oxidative phosphorylation (Su and Abumrad 2009). However, despite the multiple initial paths, our data suggest that the free energy they released appears to be the converging point of these pathways and is responsible for the observed effects on APP α-processing. Consistent with this view, other energy-rich compounds have been reported to exhibit similar effects (Camden et al. 2005; Epis et al. 2008).
APP α-processing is sensitively regulated by a myriad of metabolic or signal-transduction pathways (Lichtenthaler 2011; review) and is also sensitive to energy alterations (Gabuzda et al. 1994; Webster et al. 1998; Hoyer et al. 2005). Now our findings in this study will make the regulatory mechanism of APP α-processing or α-secretase an even more attractive subject for study. By highlighting free energy as an additional regulator, our study has revealed an intriguing, energy-dependent feature of α-secretase. Future studies on this “signature trait” may lead to a better understanding of the true identity of this key enzyme and its central regulatory role in the overall APP metabolism in brain aging and AD (Chen and Fernandez 2004; Nguyen et al. 2012).
HECs protect cells against age-related impairments
Our findings that the impairments of energy suppression and oxidative stress on APP α-processing can be partially attenuated by HECs may suggest a promising approach to reduce Aβ production in the aging brain, as the two processing pathways compete for the same APP pool (Skovronsky et al. 2000; Frautschy et al. 1998; Chen and Fernandez 2004). More importantly, that the HECs substantially rescue the cells from death process invoked by energy inhibition or oxidative stress further suggests that the HECs can enhance the cell resilience against age-related impairments. Thus, these compounds may serve as potential medications for intervention in brain aging and AD.
The precise mechanism underlying the protective effects of the HECs is unclear at present, but it is possible that increased sAPPα or decreased Aβ may each contribute a part. Although this issue may be clarified by using APP mutant cells in future studies, the use of natural cells is this study is more relevant to brain aging and early phase of sporadic AD. Our data suggest that boosting cell bioenergetics is a pertinent mechanism in the protective effects of the HECs. This view is also based on following considerations.
Brain function is highly energy dependent and enhancing energy metabolisms can facilitate brain oxygenation, circuit integration, and antioxidant defense (Bonda et al. 2009; McDaniel et al. 2003). Indeed, while the overall mechanisms of AD have remained highly controversial today, there is a broad consensus suggesting that energy depletion or mitochondrial dysfunction is perhaps the earliest measurable defect in aging, which underlies a wide range of downstream impairments such as oxidative stress, protein aggregations, and ion dyshomeostasis (Castellani et al. 2002; Blass 2000; Beal 1998; Swerdlow 2011).
This energy-centric hypothesis has been substantially supported by an increasing number of studies showing that energy suppression triggers Aβ overproduction and tau aggregation in cultured cells (Hoyer et al. 2005; Chaves et al. 2010; Leuner et al. 2012), causes Aβ deposition and tauopathy in the animal brain (Höglinger 2005; Grünblatt et al. 2007; Leuner et al. 2012), and also directly impairs animal learning and memory (Lannert and Hoyer 1998; Luques et al. 2007). These findings can explain an important empirical knowledge that physical and brain exercises should serve as the first line of defense against age-related cognitive deterioration in the elderly population.
Potential use of HECs for AD prevention
Moreover, given the key role of energy crisis in brain aging, it is conceivable that boosting energy levels would be an ideal point of entry for pharmaceutical intervention in AD. In this regard, our findings may provide a promising basis for future studies. As many HECs in the body have not been fully characterized, it is anticipated that future studies will discover more efficient and brain-specific HECs, or develop their longer lasting derivatives to serve as safe, effective, and cost-economic medications for AD. In this regard, several energy-rich supplements have been found to exert some neuroprotective effects (Owen and Sunram-Lea 2011; Adhihetty and Beal 2008; Tchantchou et al. 2008; Montgomery et al. 2003). It is likely that their long-term and early use in the aging population, enhanced by healthy lifestyle, will result in more prominent effects.
High-energy compounds may also improve the efficacy of current or experimental drugs that target neurotransmitters, Aβ or tau aggregates, free radicals, ion imbalances, inflammation, and other lesions (Salomone et al. 2012; Herrmann et al. 2011; reviews). These drugs have not shown prominent results in the clinic today, but by improving cellular bioenergetics, they may display better effects for AD, a multifactorial disorder that requires multifaceted intervention.
Finally, HECs may also benefit other late-life degenerative conditions such as Parkinson's, stroke, osteoporosis, and muscle dystrophy in which energy depletion is also the underpinning factor. As some HECs have been used to build muscles and bones, it is possible that ‘strengthen the brain like muscle and bone’ may emerge as a useful strategic thinking in our endeavor to energize and protect aging brains, thereby extending their lifespan to catch up with the extended life expectancy of the aging population.
This study is supported by the Merit Review program of The U.S. Department of Veterans Affairs (to M.C.). We thank Drs. Jessica Chang and Bruce Citron for excellent technical assistance.