Page 53 - Read Online

Novati et al. Ageing Neur Dis 2022;2:17 https://dx.doi.org/10.20517/and.2022.19 Page 7 of 29

to the degradation of damaged mitochondria. PINK1 acts as a sensor for depolarization of mitochondrial
membrane potential [85,90-93] , recruiting the E3 ubiquitin ligase Parkin which ubiquitinates substrates on the
[94]
outer mitochondrial membrane, thus eliciting a vicious cycle resulting in mitophagy . Protein deglycase
DJ-1 is a stress-dependent chaperone localized in mitochondria, which plays an essential role in ATP
production and complex I activity [95,96] . Interestingly, it has been observed that Lewy bodies can be absent in
[97]
PD patients with either PARKIN, PINK1, or DJ-1 mutation (see review ). In comparison, mitochondrial
pathology and neuronal loss in animal models of autosomal recessive PD are expected as important
pathological phenotypes. PINK1 knockout (KO) rats show decreased complex I level and increased proton
leak in the electron transport chain, indicating a mitochondrial respiration defect, as well as a reduced
number of TH-positive neurons and proteinase K resistant α-synuclein aggregates [47,98] . By contrast, no
evidence reflecting neurodegeneration was found in PINK1 KO mice , not even in the triple knockout
[99]
mice with deficiency of Parkin/PINK1/DJ-1, all known gene deficiencies related to autosomal recessive PD
forms . Similarly, DJ-1 KO rats show significantly progressive neuronal loss with approximately 50%
[100]
dopaminergic cell loss at 8 months of age in the substantia nigra, combined with altered mitochondrial
respiration [101,102] . In contrast to rat models, no dopaminergic neuron loss-related event or mitochondrial
dysfunction has been observed in all existing DJ-1 KO mouse models, while one DJ-1 KO mouse model only
shows increased sensitivity to the neurotoxin MPTP [103-105] . Notably, PARKIN KO mice and rats also have
been generated, while PARKIN KO mice exhibited increased striatal extracellular DA concentration, which
[106]
is opposite as expected , PARKIN KO rats did not show any neuropathological differences compared to
[101]
wild-type controls . Whether these results can be explained by the genetic and biological differences
between human and rodent remains unaddressed. Nevertheless, both PINK1 and DJ-1 monogenic KO rat
models are valuable for investigating mitochondrial pathology in autosomal recessive PD, whereas the
comparable mouse models lack disease-related neuropathological phenotypes.

Neuropathological phenotypes in genetic rat models of Huntington’s disease: tgHD and BACHD rats
To date, two genetic rat models have been generated and well characterized for HD research. One carries
the whole genomic sequence and regulatory elements of human HTT with 97 mixed CAG-CAA repeats in a
bacterial artificial chromosome construct (BACHD rats), thereby bearing the mutation in its appropriate
[107]
genomic context as in HD patients . The interruption in CAG repeats avoids somatic instability of polyQ
size and variation in repeat length within the animal colony. The other rat model carries N-terminal rat Htt
cDNA fragments under the rat Htt promoter, with 51 CAG repeats (tgHD rats) . In humans, the CAG
[108]
length present in the tgHD construct would lead to an adult-onset of disease, whereas 97 CAGs, as in the
BACHD rats, would result in the juvenile form of the disorder. Both BACHD and tgHD rat models have a
wide expression pattern of transgene HTT/Htt throughout the brain that, to some extent, resembles the
human condition. BACHD rats have a 4.5-fold higher expression level of transgenic HTT as the endogenous
Htt, while tgHD rats show a strongly reduced transgene expression level compared to the endogene [107,108] .
Both rat models show subtle evidence for neurodegeneration, including structural changes in white
matter [109,110] , reduced brain volume in BACHD rats , and age-dependent enlarged ventricles in tgHD rats.
[111]

Although neuronal loss in HD patients is most prominent in the striatum, mHTT aggregates have been
more frequently detected in the cerebral cortex. Subcellular localization studies revealed a prevalent
neuropil localization of mHTT aggregates, while smaller amounts of mHTT inclusion bodies were found in
the nucleus [31-33] . One of these studies reported that in all 12 investigated HD brains, only 1%-4 % of striatal
neurons had nuclear inclusion bodies, while a large number of mHTT aggregates were detected in the
cortex with prominent subcellular localization in neuropil and perikarya. Although juvenile HD patients
show an increased number of nuclear inclusion bodies compared to patients with adult-onset, neuropil
aggregates were still predominantly distributed in the striatum and cortex . Consistent with these
[32]