Late-onset Alzheimer's disease is associated with inherent changes in bioenergetics profiles

Abstract

Body-wide changes in bioenergetics, i.e., energy metabolism, occur in normal aging and disturbed bioenergetics may be an important contributing mechanism underlying late-onset Alzheimer's disease (LOAD). We investigated the bioenergetic profiles of fibroblasts from LOAD patients and healthy controls, as a function of age and disease. LOAD cells exhibited an impaired mitochondrial metabolic potential and an abnormal redox potential, associated with reduced nicotinamide adenine dinucleotide metabolism and altered citric acid cycle activity, but not with disease-specific changes in mitochondrial mass, production of reactive oxygen species, transmembrane instability, or DNA deletions. LOAD fibroblasts demonstrated a shift in energy production to glycolysis, despite an inability to increase glucose uptake in response to IGF-1. The increase of glycolysis and the abnormal mitochondrial metabolic potential in LOAD appeared to be inherent, as they were disease- and not age-specific. Our findings support the hypothesis that impairment in multiple interacting components of bioenergetic metabolism may be a key mechanism contributing to the risk and pathophysiology of LOAD.

Figure 1

Data summary of Seahorse XFp Cell Mito Stress Test assays. (a) Profiles of Mito Stress Test data for OCR, ECAR, and PPR in n = 10 LOAD and n = 20 control fibroblast cell lines with arrows indicating injections into media of the specific stressors oligomycin (Olig.), carbonyl cyanite-4 (trifluoromethoxy) phenylhydrazone (FCCP), and Rotenone/Antimycin A (R/A) (see Supplementary Fig. 1 for additional details). (b) Relative values of parameters and (c) calculated ratios between healthy old versus young control, and LOAD versus young or old control samples. (d) Results from the Seahorse XF Cell Energy Phenotype Test Report Generator shown for OCR (pmol/min), ECAR (mpH/min), and metabolic potential ((stressed OCR or ECAR/baseline OCR or ECAR) × 100%)), and calculated ratios between old versus young control, and LOAD versus young or old control samples. Experiments were performed in n = 3 replicates. *P < 0.1; **P < 0.05; ***P < 0.01 (detailed statistical data can be found in Supplementary Table 2).

Figure 2

Bioenergetics parameters related to energy production and glycolysis. ATP (a) and L-Lactate (b) production were increased in LOAD fibroblasts when compared to controls. Glucose uptake in absence or presence of IGF-1 was decreased in LOAD and old fibroblasts when compared to young cells (c). Shown are relative values and ratios comparing AD with healthy old and young and healthy old with young samples. Two experiments were performed in n = 3 replicates. *P < 0.1; **P < 0.05; ***P < 0.01 (detailed statistical data are provided in Supplementary Table 2).

Figure 3

Bioenergetics parameters related to mitochondrial respiration. Nicotinamide metabolites, including NADt (a), NADH (b), and NAD⁺ (c) were decreased in LOAD samples compared to young and old controls, while the RRs (ratios NAD⁺/NADH) were not different between controls, but slightly increased in LOAD fibroblasts compared to all, young, and old control cells (d). Shown are relative values and ratios comparing AD with old and young and old with young samples. Two experiments were performed in n = 3 replicates. *P < 0.1; **P < 0.05; ***P < 0.01 (detailed statistical data are provided in Supplementary Table 2).

Figure 4

Parameters related to mitochondria mass and membrane integrity. qPCR experiments amplifying a region of the mtMinArc (a), CS (b), and MitoTracker assays (c) showed reduction of mitochondrial DNA amplification, CS activity, and mitochondria staining in healthy old and LOAD cells, respectively, versus young controls. FCM in combination with the MitoPT^® dye JC-1 showed no significant difference in the ratios of red and green fluorescent cells between control and LOAD samples indicating no loss of transmembrane electrical potential gradient (Δψ_m) (d). The assay measures confirmation changes of the positively charged dye according to its localization in the cell, i.e., green fluorescence when accumulating in the cytoplasm and orange/red fluorescence when distributing to mitochondria. Orange/red fluorescence is indicative of dye accumulation in mitochondria due to an intact Δψ_m, while green fluorescence indicates loss of Δψ_m due to mitochondrial damage of the ETC, and proton and pH gradient. Shown are relative values and ratios comparing AD with old and young and old with young samples. Two experiments were performed in n = 3 replicates. *P < 0.1; **P < 0.05; ***P < 0.01 (detailed statistical data are provided in Supplementary Table 2).

Figure 5

Parameters related to mitochondrial DNA integrity and transcription. qPCR experiments amplifying the 4977 “common deletion” within the mtMajArc showed reduced PCR amplification in old and LOAD cells (a), but no differences of calculated ratios over mtMinArc amplification (b). Expression of the MTND4 gene was decreased in LOAD cells when compared to all, young, and healthy old controls (c). ROS levels measured with the OxiSelect (d) or MitoSox (e) assays were slightly increased in LOAD and significantly increased in healthy old cells, compared to young controls. Shown are relative values and ratios comparing LOAD with old and young and old with young samples. Two experiments were performed in n = 3 replicates. *P < 0.1; **P < 0.05; ***P < 0.01 (detailed statistical data are provided in Supplementary Table 2).

Figure 6

mRNA expression profiles of genes related to glycolysis, NAD metabolism, and CAC activity. (a) Gene expression profiles of PFKFB3 and LDHA showed downregulation in healthy old and LOAD cells, but a relative increase in the ratios of LOAD versus old fibroblasts when compared to LOAD or healthy old versus young cells. (b) Gene expression profiles of NMNAT2, NAMPT, PARP1, SIRT1, and SIRT3 demonstrated predominant downregulation in LOAD cells, including a significant decrease or increase of NMNAT2 in LOAD or old controls, respectively. (c) Gene expression of IDH3A, OGDH, MDH2, and cytosolic MDH1 showed significant differences between LOAD and old cells with increased expression of IDH3A, MDH1 and MDH2 in LOAD and decreased expression of IDH3A and MDH2 in old cells. In contrast, OGDH was decreased in LOAD and increased in old fibroblasts. Experiments were performed in n = 3 replicates. *P < 0.1; **P < 0.05; ***P < 0.01 (detailed statistical data are provided in Supplementary Table 2).

Figure 7

Heat maps of data correlations. Data related to energy production, mitochondrial function or integrity (a), and gene expression (b) were correlated within (a,b) or with age (c). Positive (blue) or negative (red) correlations (R values) are shown.

Figure 8

Schematic summary of bioenergetic changes in LOAD and old control fibroblasts. Parameters related to mitochondrial function, mass, integrity, OxPhos, and NAD metabolism are in the blue area and those related to glycolysis in the green area. Arrows indicate up-regulation, down-regulation, or no changes in parameters for LOAD relative to young or old controls, and old relative to young controls. While alterations in glucose uptake, IGF-1 signaling, mitochondrial mass, ROS production, DNA deletion, and membrane stability follow similar patterns in healthy old and LOAD cells, changes in glycolysis, NAD metabolism, CAC activity, expression of MTND4, proton leak, and respiratory potential appear to be more specific to either LOAD or aging. See text for details. Additional abbreviations: G6P: Glucose-6-phosphate; F6P: Fructose-6-phosphate; F1,6 P: Fructose-1,6-phosphate; F2,6BP: Fructose-2,6-bis-phosphate; NA: nicotinic acid; NAMN: nicotinate mononucleotide; NADD: nicotinic acid adenine dinucleotide; NAM: Nicotinamide; NMN: nicotinamide mononucleotide; NAPRT: Nicotinate Phosphoribosyltransferase; NADS: NAD synthetase.