The pathologic findings in the brain, in particular neuronal loss and LBs in the SN, are consistent with the clinical diagnosis of PD [14
]. In addition to PD, there were iatrogenic processes, including glial scar tissue in the ventrolateral thalamus that is likely the result of DBS and a neuroglial lesion in the basal ganglia and amygdala that is the result of a fetal tissue transplant.
As fetal mesencephalic tissue developed α-synuclein pathology at least 10 years after transplantation, premature aging of the graft might be a contributing factor [6
]. Although no surrogate marker for aging was found in a study reporting 9-14 year old grafts without α-synuclein pathology, other studies showed the presence of neuromelanin and age-related ubiquitin in the grafts [3
]. In our patient, there were a few neurons with neuromelanin, lipofuscin and ubiquitin pathologies (, ). Corpora amylacea were sparse in the body of the graft, but more numerous at the periphery, especially near the amygdala (). Although the graft in our case was only 14 years old, these findings suggest that the graft had little evidence in support of accelerated aging; myelination was poor, age-related ubiquitin pathology and lipofuscin (autofluorescent) pigment was minimal and corpora amylacea were sparse. In addition, phosphorylation of neurofilament, a characteristic feature of neurofilaments in mature axons was lacking in the neuron-rich portions of the graft (, , ). The distribution of corpora amylacea, a surrogate marker for gliosis since they are found in cytoplasmic processes of astrocytes, was dissimilar to that of α-synuclein pathology (, ).
As α-synuclein is a synaptic protein, it was not unexpected that the distribution of SNAP-25 was partly similar to that of α-synuclein (). In addition to some discrepancies between the distribution of SNAP25 and α-synuclein, there was no significant difference in α-synuclein pathologies between SNAP25-rich and SNAP-25-poor regions in the amygdala portion of the graft, while α-synuclein pathology was infrequent in the SNAP25-poor region in the putaminal portion of the graft (). Therefore, synaptic maturation was not enough to explain the distribution of α-synuclein pathology.
In our patient, the density of activated microglia also showed a gradient, with less in the nodular portion of the graft in the putamen and denser in the neuroglial portion of the graft in the amygdala. This pattern was similar to the distribution of LBs (, ). Thus, in view of the close relationship between activated microglia and α-synuclein pathology in the literature, microglial activation could play a role in the generation of α-synuclein pathologies [15
]. However, the widespread distribution of LNs or α-synuclein immunoreactive dots in the graft could not be readily explained by the limited distribution of activated microglia.
It should be noted that our patient exhibited α-synuclein inclusions in glia as well as neurons. This is of interest, as there is an ongoing debate about the possibility of α-synuclein production by astrocytes [16
]. An alternative hypothesis to explain glial α-synuclein inclusion is that astrocytes might take up abnormal α-synuclein secreted from the neurons [17
]. However, α-synuclein pathologies were prominent even where neurons were sparse in the graft (). Although only a few glial inclusions were found in the graft, more active involvement of astrocytes in the production of α-synuclein could not be excluded ().
Recent studies suggested a direct neuron-to-neuron propagation of α-synuclein in experimental models [7
]. Neurons containing abnormal α-synuclein can secrete α-synuclein possibly through exocytosis, which could then be engulfed by nearby neurons [7
]. The hypothetical prion-like behavior of α-synuclein might place α-synuclein pathology primarily in the periphery of the graft. However, the distribution pattern of α-synuclein pathology contradicted the expectation in our case. Moreover, α-synuclein pathology was more prominent in the neuroglial portion of the graft in the vicinity of the amygdala. The differential sensitivity of the graft to α-synuclein pathology might argue against the direct propagation model in our case.
Interestingly, in many neurodegenerative diseases, the putamen and the amygdala showed different susceptibility to pathologic processes. That is, the amygdala can be more significantly affected than the putamen by various abnormal protein aggregations, such as LBs, neurofibrillary tangles, TDP-43 neuronal cytoplasmic inclusions and argyrophilic grains. In contrast, the amygdala is unusually sensitive to abnormal aggregates, where lesions are found in early disease stages and disproportionate to other brain regions, including PD. Thus, the amygdala appears to be particularly vulnerable to neuronal protein aggregation abnormalities. In our study, there was a predominance of activated microglia and α-synuclein pathology in the ventral portion of the graft near the amygdala compared to the dorsal portion of the graft embedded in the putamen.
The mechanism of the preponderance of pathology in the vicinity of the amygdala remains speculative. Inflammatory signals from abundant reactive glia in the adjacent amygdala might play an important role [24
]. Alternatively, unknown tissue factors contributing to the selective vulnerability of the amygdala to neurodegenerative disease could be transferred to the nearby graft. Interestingly, tissue-specific susceptibility has also been demonstrated in fetal transplantation for the patients with Huntington disease (HD), in which the grafts in the caudate nucleus could not survive, while grafts in the putamen were viable [25
]. Although pathologic inclusions were not found in the fetal graft, as the caudate nucleus usually displayed most severe pathology in HD, the different pathologic outcome could be related to host tissue factors. The nature of putative transmissible tissue factors is currently unknown, it is still plausible that the fate and pathology of graft is under critical influences from neighboring tissues.
Although there was no long-term pathologic study on α-synuclein pathologies in other types of transplantation therapy, such as retinal pigment epithelial cells in PD, the “spreading” of the disease from the host could be a potential limitation of restorative therapy [10
]. Further studies are needed to investigate possible tissue factors responsible for apparent spreading of pathologic processes from host to graft, apart from the emerging concept of neuron-to-neuron propagation of α-synuclein.