Cellular senescence, characterised by the development of a toxic secretory phenotype in
response to persistent DNA damage, has been studied in proliferating cells but is not well
understood in post-mitotic neurones. Recent evidence showed a neuronal senescent-like
state in response to persistent DNA damage in vivo, which could contribute to neuronal
dysfunction in aging and neurodegeneration. The current study hypothesised that
oxidative stress, a hallmark of neurodegeneration, activates a persistent DNA damage
response and promotes neuronal senescence. To investigate this, activation of senescence
in response to oxidative DNA damage was studied in human post-mitotic neurones in
culture, as well as in the brains of control and ALS/MND donors.
For the in vitro study, post-mitotic LUHMES were stressed with a double dose of 50 µM
H2O2, which caused a persistent DNA damage in the form of double-strand breaks that
was detectable 96 hours’ post-stress. Expression of the “classical” senescence marker SA
β-gal and formation of senescence-associated heterochromatin foci (SAHF) were
evaluated at the 96 hours-timepoint, using cytochemical methods. A co-culture system of
double stressed LUHMES and healthy LUHMES was developed to study DNA damage
propagation; gene expression profiling was used to investigate changes in known
senescence pathways in 96 hours-double stressed LUHMES. Results from this study
revealed a highly variable SA-β-gal activity in healthy and double stressed LUHMES;
SAHF were not present in these cells and propagation of DNA damage was not seen in
the co-culture system. Transcriptomic analysis of double stressed LUHMES revealed
dysregulation of the APC/C:Cdh1 cell cycle regulatory pathway, ATR signalling and
II
mitochondrial complex I activity, which could be related to oxidative stress but not to
senescence.
The current work also investigated the relevance of neuronal senescence in
neurodegeneration. “Classical” senescence (SA-β-gal, p16 and p21) and oxidative DNA
damage markers (8-OHdG and γH2AX) were investigated in the motor cortex, spinal
cord and frontal association cortex of ALS/MND and control donors using
immunohistochemistry. Transcriptome analysis of LCM neurones obtained from the
frontal cortex of control and ALS/MND donors was used to investigate early changes in
gene expression that could be linked to a senescent-like state.
Transcriptomic analysis suggested dysregulation of DDR, cell cycle and oxidative
phosphorylation pathways as a consequence of the persistent oxidative DNA damage, but
no of “classical” senescence was found in the in vitro neuronal model. In vivo, p21
expression was found in neurones and glia, whereas p16 was exclusively expressed in
glial cells. A significantly higher percentage of p21+ neurones was detected in the frontal
association cortex of ALS/MND donors. Transcriptome analysis showed alteration of
DDR pathways and mitochondrial function in these neurones, but did not reveal
dysregulation of “classical” senescence pathways.
Neurones may activate a senescent-like state in response to persistent DNA damage but
signalling pathways involved in this mechanism may differ from the ones described in
mitotic cells. Thus, “classical” senescence markers should be used cautiously when
studying neuronal senescence, as they could reflect induction of related but different
mechanisms in these cells. In order to study neuronal senescence, it is necessary to
understand first the cell cycle regulatory mechanisms that occur in neurones as part of the
DDR.
