Alzheimer’s disease (AD) is a neurodegenerative disease which is the main cause of dementia worldwide, and it is currently incurable. It is characterized by a loss of memory and progressive cognitive, functional, and behavioral decline that interferes with daily life [1
]. It is now well recognized that the pathogenesis begins up to one or two decades before the onset of the clinical symptoms [2
]. Therefore, understanding the mechanisms that lead to the progression of the disease is essential to establish an early diagnosis and slow or prevent its progression.
The two hallmarks of AD are the accumulation of extracellular senile plaques formed by amyloid beta (Aβ) peptides, and the accumulation of hyperphosphorylated tau aggregates that form neurofibrillary tangles (NFTs) inside neurons [3
]. In addition to plaques and NFTs, oxidative damage to proteins, nucleic acids, and lipids plays a key role in the pathophysiology of the disease, as in most age-related ailments [4
]. It has been reported that Aβ promotes the generation of reactive oxygen species (ROS), either directly or indirectly, by triggering N-methyl-D-aspartate receptor-dependent Ca2+
influxes and leading to mitochondrial dysfunction. However, it has been postulated that Aβ accumulation may be a consequence of oxidative stress and that Aβ and tau act as antioxidants in AD (reviewed by Sutherland et al. [5
]). Although this issue is still unclear, a wide range of studies have shown that the imbalance between the production of ROS, on the one hand, and antioxidant defenses, on the other, contribute considerably to the pathogenesis and progression of AD [4
]. In fact, considerable attention in AD research has been focused on identifying compounds capable of scavenging excess ROS.
In particular, resveratrol (RV; trans-3,4′,5-trihydroxystilbene) and selenium (Se), which are both nutraceuticals with antioxidant properties that can permeate the brain blood barrier, seem to have therapeutic potential as neuroprotective agents [9
]. RV is a polyphenol that is mainly found in some fruits such as blueberries, blackberries and grapes, and also in peanuts. It has been shown that RV mimics the anti-aging and neuroprotective effects of caloric restriction through sirtuin 1 (SIRT1) mechanisms [11
]. RV indirectly activates SIRT1 through cAMP signaling that leads to activation of the 5′ AMP-activated protein kinase (AMPK)/SIRT1 pathway [12
]. Furthermore, both in vitro and in vivo experimental AD studies have suggested that RV activates the SIRT1 pathway as its main neuroprotective mechanism [14
]. However, RV may partially act through other mechanisms, as demonstrated by in vitro treatments in the presence of the SIRT1 inhibitor sirtinol, where RV neuroprotection was only partially abolished [18
]. In this regard, RV has potent antioxidant properties through direct scavenging of ROS. Some clinical trials have shown that resveratrol is safe, well-tolerated, and is capable of decreasing neuroinflammation and modifying some AD biomarkers, such as cerebrospinal fluid Aβ40 and Aβ42 [19
Meanwhile, Se is an essential micronutrient for brain function that plays a critical role in multiple metabolic pathways, including those involved in antioxidant defense in organisms [21
]. Se is a component of antioxidant enzymes, such as glutathione peroxidase, and there are a number of other selenoenzymes and selenoproteins. There are two different commonly occurring forms of Se in nature, selenite (Se (IV)) and selenate (Se (VI)), and both have been studied in the context of the prevention of AD onset and progression. Studies have shown that diets supplemented with these components can play a neuroprotective role in AD experimental models [22
]. For example, Se (IV) can reduce the amount of Aβ plaques [23
] and Se (VI) may reduce hyperphosphorylation of tau [24
]. Studies in humans have found a significant decrease of Se in AD brains or blood cells, compared to controls [25
]. Therefore, both RV and Se diet supplementation are promising strategies to combat aging and AD.
Given that several peripheral and systemic abnormalities interact with the brain and influence the development and progression of the pathology, it has been suggested that AD may be considered a systemic disease [27
]. In fact, many authors have shown that not only do components of the nervous system from AD patients present increased oxidative stress markers compared to healthy controls (HCs), but so too do lymphocytes [28
]. This has prompted some authors to use lymphoblastoid cell lines (LCLs) from AD patients and HCs as a suitable and more feasible model to study the disease in vitro. These human cell lines arise from peripheral B lymphocytes infected in vitro with the Epstein–Barr virus; a process that immortalizes them [29
]. Some studies have already shown alterations in the cell cycle, proliferative activity and Aβ processing, as well as higher oxidative stress in AD than control-derived LCLs [30
Within this context, we use LCLs from AD patients and HCs to investigate the potential role of RV and both Se (IV) and Se (VI) in the reduction of ROS generated after an oxidative injury. We also examine whether these compounds elicit expression changes in genes involved in the antioxidant cell response and other aging-related mechanisms. We found that AD LCLs showed a lower capacity of response against oxidative injuries than HC LCLs, as expected. Furthermore, RV triggered a protective response against ROS under control and oxidizing conditions and increased the expression of gene coding for known antioxidants and anti-aging factors; whereas Se exerted antioxidant effects only in AD LCLs under oxidizing conditions. Our findings support RV as a powerful compound with preventive and therapeutic properties against redox and aging alterations of AD and reinforce the value of LCLs as a human cell model for studying the protective mechanisms of nutraceuticals.
2. Materials and Methods
2.1. Cell Lines
Immortalized lymphocytes from AD patients from the Department of Neurology of the University Hospital Doce de Octubre (Madrid, Spain) and age-matched HCs, were used for this study. AD patients were at a moderate stage of the disease and presented values between 10 and 18 in the Mini-Mental State Examination. Details of the informed consent and technical procedures for the establishment of LCLs from peripheral blood samples were previously reported [32
]. The cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 (Biowest, Nuaillé, France, #L0500), which contained 2 mM L-glutamine, supplemented with 10% fetal bovine serum (FBS; Gibco, Pailey, Scotland, #10270) and 1% penicillin/streptomycin (Gibco, #15070) or 0.1% gentamicin (Gibco, #15750-045). LCLs were grown in suspension inside T25 flasks in an upright position, in 8 mL of completed medium per flask and at a density of 1 × 106
cells/mL. They were maintained in a humidified 5% carbon dioxide incubator at 37 °C. The culture medium was routinely changed every 2 days by removing 4 mL of the medium from above the cells and replacing it with an equal volume of fresh medium. The cell lines were routinely tested for the absence of mycoplasma contamination (Mycoplasma Gel Detection Kit; Biotools, Madrid, Spain, #4542). For the experiments, cells were seeded in 2 mL tubes with 1 mL of medium without FBS for the time required before each experiment at a concentration of 3 × 105
cells/mL. All the cell culture plastic was from Nunc™ (ThermoFisher Scientific, Waltham, MA, USA).
2.2. Characterization of Oxidative Stress by DCFH-DA (2′7′-dichloro-dihydro-fluorescein diacetate) Assay
The most widely used probe for the detection of oxidative species in living cells is 2′7′-dichloro-dihydro-fluorescein diacetate (DCFH-DA). It is a non-fluorescent cell-permeable molecule. Within cells, the acetate groups are hydrolyzed by intracellular esterases, leading to 2′,7′-dichloro-dihydro-fluorescein (DCFH). The presence of ROS, mainly hydroperoxides, oxidizes DCFH to dichlorofluorescein (DCF), which is highly fluorescent [40
]. To use this technique, cells from 2–3 different AD or HC LCLs were seeded at a cell density of 3 × 105
cells/mL in T25 flasks, the day before the experiment, in FBS-free medium. For the test, the cells were gently homogenized with the medium and 1 mL was transferred to different tubes. After centrifugation at 1000 rpm for 5 min, the cells were resuspended with 400 µL of 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-buffered saline solution (HBSS) and 4 µL of 100× DCFH-DA (10 µM final concentration; Molecular Probes, Leiden, The Netherlands, #D-399) or 4 µL of HBSS for the negative controls. Negative controls were used to obtain the background fluorescence data and were processed in parallel throughout the experiment. After suspension, all the samples were incubated for 20 min with gentle shaking at 37 °C in the dark to allow the DCFH-DA to load into the cells. After centrifugation for 5 min at 200 rcf, the supernatant was removed, and the cells were resuspended in 400 µL of HBSS containing the corresponding treatments (performed in triplicates). The different conditions were: control treatment with HBSS only, and different concentrations of the oxidizing agents H2
(200 µM, 500 µM, and 1000 µM; Sigma, St Louis, MO, USA, #216763) and FeSO4
(1 µM, 5 µM, and 25 µM; Sigma, #F7002) in HBSS. Then, all the samples were incubated for 1 h with gentle shaking at 37 °C in the dark. Next, 44 µL of 10× lysis buffer was added to each tube. All the samples were homogenized and centrifuged at 10,000 rcf for 10 min, and 200 µL of the supernatant was transferred to a 96-well plate, making two replicates per sample. Cell fluorescence was determined using a SPECTRAmax GEMINI XS microplate reader fluorimeter (Molecular Devices, San Jose, CA, USA), with a wavelength of excitation and emission of 485 nm and 530 nm, respectively. Finally, in order to control for the cellular protein content, 50 µL of 2 N NaOH solution was added to the cell pellet for future processing using the Bradford protein assay (Bio-Rad Protein Assay Dye Reagent Concentrate; Bio-Rad, Hercules, CA, USA, #500-0006).
2.3. DCFH-DA Assay to Study Se (IV), Se (VI,) and RV Antioxidant Effects
The same protocol as for the DCFH-DA assay described above was applied with some changes. Cells from 2–3 different AD or HC LCLs were seeded at a density of 3 × 105 cells/mL in 2 mL tubes. Twenty µL of each protective treatment was added to the corresponding tubes (performed in triplicate). The concentrations of the protective compounds were obtained from the literature and tested in preliminary studies not to affect cell growth or viability. The different protective treatments were: Se (IV) (Sigma, #S5261) at 5 and 10 µM, Se (VI) (Sigma, #S0882) at 100 and 200 µM, and RV (Sigma, #R5010) at 10 and 50 µM. RV, Se (IV), and Se (VI) were solubilized with DMSO (0.1%) or HBSS, respectively. Afterwards, DMSO was added to all the experimental conditions (0.1%). After overnight incubation (18 h), the samples were centrifuged at 1000 rpm for 5 min; cells were resuspended with 400 µL HBSS and 4µL of 100× DCFH-DA. Eight µL of each protective treatment was again added to maintain the corresponding concentrations. Then, the samples were incubated for 20 min with gentle shaking at 37 °C in the dark. After centrifugation for 5 min at 200 rcf, the supernatant was removed, and the cells were resuspended in 400 µL of the corresponding protective treatment and/or the agents used to induce oxidative stress: H2O2 and FeSO4. The final experimental conditions were: control, 1000 µM H2O2 or 5 µM FeSO4, with 8 µL of Se (IV), Se (VI), or RV. The samples were incubated for 1 h with gentle shaking at 37 °C in the dark. Next, 44 µL of lysis buffer 10× was added to each tube. All the points were homogenized and after centrifugation at 10,000 rcf for 10 min, 200 µL of the supernatant was transferred to a 96-well plate, making 2 replicates for each sample. Finally, in order to control the cellular protein content, 50 µL of 2 N NaOH solution was added to the cell pellet for future processing using the Bradford protein assay.
2.4. Gene Expression Analysis
Cells from 2–3 different AD or HC LCLs were transferred to 2 mL tubes containing FBS-free medium and incubated for 18 h at a concentration of 3 × 105 cells/mL with the corresponding treatment (four replicates per experimental condition). The samples were removed from the incubator and rapidly centrifuged at 1300 rpm for 5 min and washed with 200 µL of PBS. Then, the samples were centrifuged again at the same speed for 5 min, the supernatant was removed, and cells were resuspended with 20 µL of PBS followed by 180 µL of RNA later (Sigma, #R0901). The samples were stored at 4 °C. The different experimental conditions consisted of: non-treatment control, Se (IV) at 10 µM, Se (VI) at 200 µM, and RV at 50 µM. RV, Se (IV), and Se (VI) were solubilized with DMSO (0.1%) and HBSS, respectively. DMSO was added in all experimental conditions. The analysis of gene expression was performed in the absence of oxidative challenges.
2.5. mRNA Purification
The four samples corresponding to the same conditions were pooled in one single tube. Then, RNA later was removed, and RNA was extracted using mirVana™ miRNA Isolation Kits (Life Technologies, Carlsbad, CA, USA, #AM1561), following the manufacturer’s instructions to obtain total RNA, including small RNA. The quantity and quality of the RNA samples was determined using a ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). Samples with low concentrations of RNA were concentrated using the SpeedVac vacuum system (Savant, ThermoFisher, Waltham, MA, USA). Samples were stored at −80 °C until further use.
2.6. cDNA Reverse Transcription
Random-primed cDNA synthesis was performed at 37 °C starting with 0.3 μg of RNA, using high-capacity cDNA Reverse Transcription Kits (Life Technologies, #4368814). All the samples were diluted at a 1:4 ratio and stored at −20 °C until further use.
2.7. Real-Time Quantitative PCR
Gene expression of candidate genes was determined using TaqMan Fluorescein amidite (FAM)-labeled specific probes (Applied Biosystems, Foster City, CA, USA) and Quantimix Easy Probe kits (Biotools, #10.601-4149) in an RFX96TM Real-time system (Bio-Rad). Samples were analyzed in duplicate. Results were normalized to phosphoglycerate kinase 1 (PGK1
) and beta-2-microglobulin (B2M
) gene expression levels. A list of probes is provided in Supplementary Table S1
2.8. Statistical Analysis
Statistical analysis was performed using GraphPad Prism 5.01 software (GraphPad Software, La Jolla, CA, USA). Analysis was via two-way ANOVA. Significance values were given for the two factors: treatment (Tr) and disease (Ds), and for the interaction Tr × Ds. A post hoc Tukey’s test, or Fisher’s Least Significant Difference (LSD) test, was applied after a significant Tr effect (since this factor has more than two levels) and after a significant interaction. All the values are shown as mean ± standard error (SEM). Statistical outliers (≥ two standard deviations from the mean) were removed from the analysis. P-values ≤ 0.05 were considered statistically significant.
In this study, we used LCLs derived from lymphocytes of AD patients and age-matched HCs to investigate the potential protective effects of RV and both Se (IV) and Se (VI) on ROS levels, generated at basal conditions and after an oxidative insult. We also studied the potential mechanisms of action involving modulation of antioxidant and anti-aging genes.
AD LCLs showed lower antioxidant defenses than HC LCLs, with higher ROS levels in response to different concentrations of oxidizing agents (H2
). The increased oxidative stress of AD LCLs is consistent with observations in many studies using fresh lymphocytes from AD patients [4
], suggesting that the immortalized AD lymphocytes are a valuable model to test protective strategies. Furthermore, cellular level oxidative stress might be associated with pathology severity in the donor. In a cell model of familial AD by expression of PSEN1
mutations in fibroblasts, we have previously seen higher basal and H2
-induced ROS in the more aggressive mutations than in those causing a milder AD phenotype [69
]. Here we used LCLs obtained at a moderate stage of sporadic AD and they showed distinct characteristics of oxidative stress. We may speculate that they show intermediate levels of AD-associated oxidative stress and are a sensitive model to test responses to interventions.
Interestingly, in the absence of oxidative insults, RV decreased ROS levels in both AD and HC LCLs. Given that oxidative stress is recognized as a risk factor for developing AD, this finding supports the potential preventive and therapeutic effects of this component by reducing the ROS burden not only in AD patients but also in healthy and at-risk populations. Consistent with this, it has been reported that RV can prevent the deleterious effects triggered by oxidative stress in neuronal cells and brain tissue of experimental AD models [70
]. Likewise, under oxidizing conditions RV triggered a global protective response towards ROS in both AD and HC LCLs. Although the specific mode of action of RV needs further characterization, its antioxidant properties may contribute to the potent neuroprotective effects reported in AD mouse models [14
], and to the promising effects reported in clinical trials with AD patients [19
The Se compounds we used confirmed the antioxidant properties of this element in LCLs, although both Se (IV) and Se (VI) only exerted antioxidant effects in AD cell lines under H2
treatment. This suggests that, compared to RV, Se (IV) and Se (VI) need a stronger oxidizing insult and higher basal levels of ROS to exert beneficial effects at the particular concentrations studied and time of exposure. In fact, whereas Se levels are decreased in AD patients [25
], a recent study has shown that Se status is not associated with cognitive performance in a healthy population [73
]. Similarly, no preventive AD effects have been reported for Se supplementation in a long-term supplementation trial [74
]. Future studies of Se therapeutic properties would clarify whether there is increased protection under greater oxidant/antioxidant imbalance.
Regarding our transcriptional experiments, Se (IV) and Se (VI) did not significantly regulate any of the genes studied. Interestingly, we observed a general upregulation of the following genes caused by RV: CAT
, and TXNIP
. Treatment with the highest concentration of RV (50 µM) consistently upregulated CAT
, and VPS13C
in both AD and HC LCLs (there was a global RV effect for all these genes). Notably, NFE2L2
(also known as NRF2
) is a transcription factor involved in the activation of genes with antioxidant response elements (AREs) such as SOD1
and many members of the glutathione S-transferase family [75
], to counteract endogenously or exogenously generated oxidative stress. It also upregulates the expression of the histone deacetylases SIRT1
] and SIRT3
], according to several studies. The upregulation of NFE2L2
could, therefore, be responsible for the upregulation of GSTZ1
(a chaperone involved in the activation of SOD1
), and both SIRT1
. Similarly, SIRT1
can enhance the activity of the NFE2l2/ARE pathway according to some authors, thus establishing a positive feedback [77
]. These findings are consistent with previous studies of brain ischemia, Parkinson’s disease, and other pathological conditions showing that treatment with RV activates the NFE2l2/ARE pathway [82
]. Furthermore, RV has been proven to activate SIRT1
through the metabolic sensor AMPK, as indicated above, which is consistent with findings in animal studies [15
]. Similarly, the increase in CAT
expression is consistent with some animal and human studies showing an increase in both activity and expression of this enzyme as a consequence of RV treatment [84
] or SIRT1
]. SIRT1 is known to deacetylate SOD1 and promote its activity by facilitating its association with copper chaperone for SOD1 (CCS); therefore, the upregulation of both genes as a consequence of RV treatment may further contribute to ROS elimination by increasing SOD1 activity [86
]. Interestingly, the gene in the sirtuin family SIRT3
is a known key enzyme for the functionality of mitochondria and its decrease is linked to neurodegeneration such as AD [87
is another mitochondrial enzyme that was upregulated by RV. Consistently, a previous study reported that SOD2
expression was induced by RV-dependent activation of the PI3K/Akt and GSK-3β/β-catenin signaling pathways [88
]. RV upregulation of SIRT3
would improve mitochondrial metabolism and energy efficiency, leading to decreased mitochondrial oxidative stress. In some cases, either HC or AD LCLs showed greater responsiveness to the upregulation of specific genes. For instance, the increase of NEF2L2
by RV was significant in HC LCLs, whereas in the case of CAT
the upregulation was significant for AD LCLs. This suggests that some genes could have a different response to RV treatment according to the cellular environment. For example, it is possible that CAT
expression is sensitive to RV in a context of higher oxidative stress, and that NEF2L2
constitutes a target of RV when used as a preventive approach or early treatment (i.e., when there is a low to moderate burden of ROS). Meanwhile, TXNIP
is considered an intracellular amplifier of oxidative stress, since it is a negative regulator of the thioredoxin system; a major cellular thiol-reducing and antioxidant system. According to several authors, RV inhibits TXNIP
]. Unexpectedly, we observed slight upregulation of this gene by RV, which is inconsistent with previous findings and could be counterproductive in the context of AD. Special attention should be paid to this issue in future studies. Finally, genetic studies have implicated loss-of-function mutations in the human VPS13
gene in neurodegenerative disorders by causing defects in membrane lipid homeostasis [67
]. The increase in this gene by RV might help prevent aging-associated lipid imbalances [91
Regarding basal transcriptional differences between HC and AD LCLs, we observed that AD lymphoblasts exhibited general upregulation of CASP1
and downregulation of CCS
. Consistently, CASP1 is activated in AD brains and overexpressed in monocytes from AD patients, and it is in fact considered a therapeutic target against age-dependent cognitive deficits and AD [92
]. Elevated antioxidant enzyme levels have been reported before in AD patients [94
], and the upregulation of SOD2
might indicate compensatory upregulation of mitochondrial antioxidant defenses in response to oxidative stress in AD LCLs. Interestingly, a downregulation of CCS has been reported in several AD models, which diminishes SOD1
activity and increases the expression of the enzyme β-secretase 1 involved in the amyloidogenic processing of the amyloid precursor protein (APP) [94
]. We found that RV treatment upregulated CCS
expression levels, thus reaching and surpassing those of HC LCLs under basal conditions.
Overall, this study confirms a derangement of oxidative defenses in lymphocytes from AD patients, indicating lower resilience to oxidative injuries and age-related oxidative stress, and supports RV as a more powerful compound, with more consistent and robust effects regarding antioxidant and transcriptional outcomes, than Se (IV) or Se (VI) under our experimental conditions. In particular, modulation of gene expression of important anti-aging (SIRT1
) and antioxidant (CAT
) genes by RV seems to be partly contributing to its mechanism of action. The fact that protective mechanisms of RV are activated in cells from both healthy and diseased AD donors is in agreement with the activation of protective mechanisms against aberrant proteostasis in both wild-type and AD transgenic mice after chronic treatment with RV [15
]. In that last study, we found that RV mechanisms yielded a strong neuroprotection against memory loss and AD pathology in transgenic mice and cognitive enhancement in wild-type mice. Here, the activation of anti-aging and antioxidant genes in our peripheral cell models further confirms that RV may be a potent protective agent and an inducer of resilience against aging and AD. Thus, in accordance with our findings and those of previous reports, RV should be studied further and considered a valuable nutraceutical candidate for early therapies aiming to prevent or delay the onset and progression of AD clinical symptoms.
Moreover, our findings reinforce the value of LCLs as a feasible model for understanding the protective mechanisms of nutraceuticals with antioxidant properties (such as RV) against the cumulative burden of oxidative stress and other cell alterations in AD.