NECAB2/MitoNEET represents an alternative activity-triggered mitochondrial quality control system
Date issued
Authors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
License
Abstract
In recent years, more and more mitochondrial quality control (mtQC) pathways,
sometimes overlapping, have been discovered. Neurons cannot proliferate and, at
the same time, have high energy demands, requiring carefully orchestrated mtQC to
ensure mitochondrial health. The mtQC pathways can, to some degree, compensate
for each other (Hammerling et al., 2017); the impact of each one remains to be
measured. This may help explain why most neurodegenerative diseases, whether
sporadic or caused by gene mutations, usually manifest at an advanced age. In this
thesis I propose a novel activity-triggered mitochondrial quality control (mtQC) mechanism actuated by neuronal calcium-binding protein 2 (NECAB2) and MitoNEET
(mNT). NECAB2 is a protein primarily found in the striatum and was first described by
Bernier et al. in 2001, while mNT was first described as a target of the type 2 diabetes
drug pioglitazone by Colca et al. (2004).
In my experiments, I was able to demonstrate that NECAB2, which was originally
described as a cytosolic protein (Canela et al., 2009), is present at mitochondria and
involved in mtQC. My mNT and NECAB2 colocalization studies were inconclusive,
but mNT expression was increased in cells expressing NECAB2 compared to cells
with no NECAB2 expression. I hence here propose a model for NECAB2-mediated
mitochondrial quality control: high calcium levels caused by damaged mitochondria
combined with low mitochondrial membrane potential (Δψm) represent a mitophagy-inducing “double-hit”. As speculated by Dey et al. (2021), high calcium levels would
cause NECAB2’s localization to defective mitochondria. I hypothesize that NECAB2
then binds to phosphorylated or ubiquitinated mNT. Here, thanks to its ABM domain,
NECAB2 degrades mNT’s 2Fe-2S cluster, thus tagging mitochondria for degradation.
Concurrently, Karmi et al. (2017) proposed mNT plays a role in traditional autophagy.
Furthermore, mNT is a Parkin ubiquitination substrate (Lazarou et al., 2013) and
lowers the mitochondrial membrane potential (Δψm) (Kusminski et al., 2016). The
ABM domain’s function in NECAB2 has not been uncovered yet, but analogous
monooxygenases in prokaryotes degrade heme to ferrous iron (Lojek et al., 2017,
Lyles and Eichenbaum, 2018).
mNT KO in mice results in a Parkinson’s disease (PD) phenotype (Geldenhuys et al.,
2017). NECAB2 KO in mice results in motor symptoms (Dey et al., 2021) and the
expression of NECAB2 is upregulated in patients with sporadic PD and genetic PD mutations (Fernandez-Santiago et al., 2015, Schondorf et al., 2014). Along with Dey
et al. (2021), I speculate NECAB2 could thus be involved in the etiology of PD, but
also in amyotrophic lateral syndrome (ALS). Congruently, in human iPSCs-derived
dopaminergic neurons, I was able to demonstrate NECAB2 and mNT upregulation in
two different PINK1 KOs compared to their isogenic control.
Given early endosomes’ continuous presence in the cell, Hammerling et al. (2017)
propose they are a first-line defense against malfunctioning mitochondria. If the damage
exceeds endosomal capacity, canonical autophagy is initiated. The results from
my experiments lead me to conclude that NECAB2/mNT-dependent mtQC is also a
first-line defense mechanism and therefore tags mitochondria for endolysosomal
degradation, rather than traditional autophagy. Indeed, mitochondrial and endosomal
colocalization in primary striatal neurons from NECAB2 WT and KO mice was about
the same at baseline and dramatically increased in KO after treatment with the mitochondrial uncoupler FCCP. This is probably because mitochondrial distress caused
by FCCP and KO (over-) activates other mtQC pathways involving Rab5 (an
endosomal marker). The rest of my experiments to study Rab5’s implication in the
NECAB2/mNT mtQC mechanism were inconclusive. However, in healthy human iPSCs-
derived dopaminergic neurons, NECAB2 and Rab5 colocalization was significantly
increased compared to the two PINK1 KOs. This has two potential explanations:
if the NECAB2/mNT pathway is PINK1/Parkin-dependent, NECAB2 and mNT
would not be degraded in PINK1 KO, explaining their higher levels in Western blot.
Higher colocalization between Rab5 and NECAB2 in healthy cells could be a product
of lower NECAB2 levels or it could be an indicator for better mtQC orchestration. On
the other hand, if NECAB2/mNT can be phosphorylated by a kinase other than
PINK1/Parkin, it could activate mitophagy in PINK1 KO cells while somewhat compensating PINK1/Parkin loss. In this case, however, I would have expected higher
Rab5/NECAB2 colocalization in the PINK1 KO cells. To settle which of the two hypotheses is correct, more experiments need to be performed comparing single NECAB2 and PINK1 KO to double KO. In any case, NECAB2’s involvement in mtQC
and upregulation in cells derived from PD patients make it an interesting potential
therapeutic target in PD, but further experimental work is required.