Access Type

Open Access Dissertation

Date of Award

January 2020

Degree Type


Degree Name




First Advisor

Thomas H. Sanderson

Second Advisor

Karin Przyklenk


Ischemic brain injury caused by cardiac arrest or stroke continue to be leading caused of death and disability in the U.S. While restoration of blood flow is necessary to salvage ischemic tissue, reperfusion paradoxically exacerbates injury via the production of reactive oxygen species, which, damage mitochondria and induce cell death. Therefore, it is critical to have stringent quality control mechanisms to ensure a healthy mitochondrial network. Mitochondrial fragmentation has been well characterized in the progression of ischemia/reperfusion and its association with cell death. Conversely, the role of mitophagy has been controversial regarding whether upregulation of mitophagy serves as a restorative process for the turnover of dysfunctional mitochondria, or have deleterious effects through energy imbalance and induction of apoptosis. Furthermore, the interplay between mitochondrial dynamics and mitophagy in ischemia/reperfusion remains unknown.

Utilizing an in vitro model of ischemia/reperfusion in primary cortical neurons, we tested four hypotheses:

i) Mitochondrial morphology in primary cortical neurons can be classified into 4 distinct groups utilizing a machine-learning based classification sytem.

ii) Mitochondria undergo significant remodeling during oxygen-glucose deprivation and reoxygenation, resulting in extensive fragmentation and swelling.

iii) Mitophagic flux is significantly increased via the PINK1/Parkin pathway of mitophagy and are sequestered by Rab5 endosomes for transport to the lysosome. Furthermore, mitochondria inside lysosomes are primarily punctate mitochondria.

iv) Knockout of Drp1 stabilizes mitochondrial architecture, but has no effect on mitophagic flux during OGD/R.

In support of hypothesis I, we demonstrate the development of a highly sensitive

and specific machine-learning based classification system of 4 distinct mitochondrial morphologies: network, unbranched, swollen, and punctate. Using the Random Forest algorithm, our high-throughput classification system demonstrated mitochondrial morphologies can be predicted with accuracy. In agreement with hypothesis II, mitochondrial fragmentation was quantitatively measured utilizing our machine-learning classification system during oxygen-glucose deprivation ad reoxygenation. This was demonstrated by decreases in networks and unbranched mitochondria with corresponding increases in punctate and swollen mitochondria. In support of hypothesis III, we demonstrated that mitophagic flux was increased during oxygen-glucose deprivation, by the presence of mCherry puncta in primary cortical neurons isolated from mitoQC reporter mice. Additionally, mitophagy was activated by the PINK1/Parkin pathway of mitophagy and that punctate mitochondria were sequestered by Rab5 endosomes for transport to the lysosome. Lastly, in support of hypothesis IV, we demonstrate the KO of Drp1 prevents mitochondrial fragmentation and swelling during oxygen-glucose deprivation in primary cortical neurons, but has no effect on mitophagic flux. In conclusion, our results illustrate that in the context of ischemia/reperfusion injury, Drp1 shifts the balance from physiological fission and healthy mitochondrial turnover, to pathological fragmentation and swelling inducing cell death.

Included in

Physiology Commons