Title: Mitochondrial Function in Cognitively Unimpaired Older Adults: Associations with Sex, Depression, Cytokines, and Aβ

Authors: Chelsea Reichert Plaska, Brandi Quintanilla, Sang Han Lee, Ashutosh Tripathi, Bruno P. Imbimbo, Henrik Zetterberg, Anilkumar Pillai, Nunzio Pomara

Keywords: Mitochondria, Amyloid-β, Plasma Biomarkers, Depression, Alzheimer’s disease

 

Introduction

• Late-life major depressive disorder (MDD) is a leading cause of disability, contributes to accelerated cognitive decline in older adults and is a risk factor for Alzheimer’s disease (AD).
• Mitochondrial dysfunction is increasingly recognized as key contributor to the pathophysiology of late-life depression and AD.
• Mitochondria regulate energy production, stress responses, and apoptosis in neurons and glia, while their dysfunction can trigger pro-inflammatory signaling, cellular senescence, and interfere with amyloid-beta -beta (Aβ) clearance.
• Circulating mitochondrial DNA (mtDNA), as cell-free fragments (ccf-mtDNA) or encapsulated within extracellular vesicles (EV-mtDNA), has emerged as a non-invasive biomarker of mitochondrial stress.
• However, the relationship between plasma mtDNA, depression status, biomarkers of cerebral amyloidosis, APOE genotype and sex remain largely unexplored.
• We investigated whether depression, sex and APOE genotype influence plasma mtDNA levels and their longitudinal associations with the plasma Aβ42/Aβ40 ratio, a marker of cerebral amyloidosis.

 

Methods

• This study included a longitudinal cohort of cognitively unimpaired older adults with and without MDD (N = 112; see demographics table below).

• All participants underwent annual clinical evaluation for three years (N=80 completed all visits).
• Baseline ccf-mtDNA and EV-mtDNA levels were determined using quantitative polymerase chain reaction.
• Plasma AD biomarkers and cytokines were determined as previously described.
• A multi-factor ANOVA examined whether age, sex, APOE status, and mtDNA level (baseline relative quantity [RQ]) were associated with plasma Aβ42/40 at year 3, and whether the mtDNA biomarker–outcome association differed by sex.
• Pearson’s correlations with cytokines were run.

 

Results

For ccf-ND1, there were significant main effects of APOE Status (p=0.023), sex (p=0.049), and ccf-ND1 (p=0.0260), but not age (p>.05). The Sex×mtDNA biomarker interaction was significant, F(1,74)=4.17, p=0.045, indicating a longitudinal association between baseline mtDNA biomarker levels and plasma Aβ42/40 at Year 3 (Fig. 1) which differed by sex (Fig 2). For ccf-ND4 there were similar main effect results. Importantly, the Sex×mtDNA biomarker interaction was significant for ccf-ND4 as well, F(1,74)=4.34, p=0.041. As for ccf-ND1, ccf-ND4 showed a negative association with Year 3 plasma Aβ42/Aβ40 ratio in females but not in males (Fig. 3). Additionally, in females only, we found a significant negative moderately-strong association between baseline ccf-ND1 and IL-10 (rho=-0.405, p=0.002)  and ccf-ND4 and baseline IL-10 (rho=-0.404, p=0.002).

 

 

Conclusions

•  In cognitively unimpaired older women, higher baseline mtDNA levels—suggestive of increased mitochondrial stress—were associated with lower baseline levels of the anti-inflammatory cytokine IL-10 and with lower Year 3 plasma Aβ42/Aβ40 ratios, consistent with reduced anti-inflammatory signaling and greater cerebral amyloid deposition over time.
•  These findings suggest a sex-specific relationship linking mitochondrial dysfunction, reduced anti-inflammatory signaling, and the development of AD-related amyloid pathology.
•  One potential interpretation is that mitochondrial stress may promote a shift toward a more pro-inflammatory immune milieu, thereby impairing mechanisms involved in amyloid clearance and increasing vulnerability to amyloid accumulation.
•  Future studies are needed to elucidate the cellular and molecular mechanisms underlying these associations, including the potential contributions of neural cell-derived extracellular vesicle mtDNA (EV-mtDNA), microglial function, and sex-related differences in neuroimmune regulation.

Limitations

• We did not have direct measures of brain amyloid, like PET.
• The extent to which blood plasma brain-derived EV-mtDNA and ccf-mtDNA reflects a neural origin remains to be determined

 

1) Ownby RL, et al. Arch Gen Psychiatry. 2006 May;63(5):530-8. doi: 10.1001/archpsyc.63.5.530. PMID: 16651510; PMCID: PMC3530614.

2) Allen J, et al. Front Neurosci. 2018 Jun 6;12:386. doi: 10.3389/fnins.2018.00386. PMID: 29928190; PMCID: PMC5997778.

3) Wang W, et al. Mol Neurodegener. 2020 May 29;15(1):30. doi: 10.1186/s13024-020-00376-6. PMID: 32471464; PMCID: PMC7257174.

4) Rangaraju V, et al. Journal of Neuroscience 2019; 39(42): 8200-8208. doi: 10.1523/JNEUROSCI.1157-19.2019

5) Trumpff C, et al. Mitochondrion 2021; 59: 225-245. doi: 10.1016/j.mito.2021.04.002

6) De Gaetano A, et al. Cells 2021; 10(11): 2898. doi: 10.3390/cells10112898