Phosphorylated alpha‐synuclein in Iba1‐positive macrophages in the skin of patients with Parkinson's disease

Abstract Background Increasing evidence suggests that alpha‐synuclein (αSyn) accumulation in cholinergic and adrenergic fibers in the skin is a useful biomarker to diagnose idiopathic Parkinson's disease (IPD). It has been widely reported that phosphorylated αSyn (p‐αSyn) deposits in autonomic fibers in IPD are a biomarker in the skin, but other tissue localizations have not been fully investigated. Objective It has been previously suggested that αSyn aggregates activate peripheral macrophages and that peripheral macrophages ingest pathological αsyn aggregates in aged rats or IPD patients. However, it remains to be elucidated whether peripheral macrophages in the skin of IPD patients accumulate αSyn. We evaluated whether (1) p‐αSyn deposits in dermal macrophages might represent a useful biomarker for IPD and (2) dermal macrophages play a role in the underlying pathogenesis of IPD. Methods We performed an immunohistological analysis of skin biopsy specimens from IPD patients and controls. Results We found that (1) p‐αSyn accumulation is present in dermal macrophages in skin biopsy specimens from patients with IPD, (2) not only dermal adrenergic fibers with p‐αSyn deposits but also dermal macrophages with p‐αSyn deposits are useful biomarkers for IPD patients and (3) the number of macrophages was significantly positively correlated with the number of macrophages with p‐αSyn deposits in the dermis of IPD patients. Interpretation Our results suggest that dermal macrophages, which are innate immune cells, play an important role in IPD patients and are a novel biomarker for IPD.


Introduction
Idiopathic Parkinson's disease (IPD) is a common secondary neurodegenerative disorder that affects more than 1% of the population over 65 years of age. 1 The routine diagnosis of IPD depends on clinical criteria mainly focused on motor symptoms. 2 However, the clinical criteria are not sufficient to precisely differentiate IPD from other parkinsonisms. 2,3 The histopathological features of Parkinson's disease (PD) are the loss of dopaminergic (DA) neurons in the substantia nigra (SN) and the presence of cytoplasmic protein aggregates known as Lewy bodies (LBs). 2 The major components of LBs are b-sheet fibrillar aggregates of alpha-synuclein (aSyn) associated with synaptic vesicles in presynaptic nerve terminals. 4 Intracellular aSyn forms aggregates within neurons but also spreads out into extracellular spaces. 5,6 Notably, Braak, Del Tredici 7 postulated that a neurotoxic pathogen, possibly a viral toxin, may enter the brain via two routes: (1) A nasal route, with anterograde progression into the limbic system, and (2) a gastric route, after swallowing nasal secretions in saliva. [7][8][9] The latter nasal secretions might contain a neurotoxic pathogen that, after penetrating the epithelial lining, could enter the axons of Meissner's plexus and reach the preganglionic parasympathetic motor neurons of the vagus nerve via transsynaptic transmission. These processes are hypothesized to be followed by retrograde transport into the medulla, pons, and midbrain, and finally, the SN might be affected. [7][8][9] Thus, stereotypical cell-to-cell spreading of aSyn is believed to contribute to disease progression.
Several lines of evidence show that microglia are associated with the removal of extracellular aSyn aggregates and the response to neurodegeneration in the early stages of IPD pathogenesis. 5,6,10,11,12 Although a correlation between aSyn aggregates and microglia is speculated in IPD pathogenesis, the phagocytosis of aSyn aggregates by microglia has not been demonstrated in human brain tissues. In this study, we examined possible relationships among neurodegeneration, microglia, and aSyn aggregates in brain autopsy specimens from IPD patients.
The dermal macrophages in the skin as well as the microglia in the brain are phagocytic cells. Dermal macrophages are involved in peripheral inflammation in innate immunity, autoinflammation, and acquired immunity and are involved in cutaneous tissue remodeling. 13 Furthermore, it has been suggested that aSyn aggregates activate peripheral macrophages and that peripheral macrophages ingest pathological asyn aggregates in aged rats or PD patients. 14,15 Although phosphorylated aSyn (p-aSyn) deposits have been observed in sympathetic adrenergic and cholinergic fibers in the skin of IPD patients [16][17][18] and cholinergic fibers have been observed in the rectal submucosa of IPD patients, 19 whether p-aSyn deposits are observed in macrophages in the skin has not been previously reported. In this report, we demonstrated that dermal macrophages in skin biopsy specimens from IPD patients contain p-aSyn deposits.

Clinical information of the patients for skin biopsy
We studied 10 patients with IPD who met the clinical diagnostic criteria of the UK PD Society Brain Bank. 2 The ages of the patients ranged from 56 to 77 years, with an average of 67.8 years. The 10 patients included four female and six male patients (Table 1). All patients had late-onset disorders (older than 45 years), and none of them had a family history of PD. All patients were treated with L-dopa alone or with other antiparkinsonian medications and maintained good control (CN) of their motor symptoms. We examined four consecutive autopsied decedents located at the Tokyo Metropolitan Geriatric Hospital as healthy CNs. The ages of the CNs ranged from 65 to 70 years, with an average of 67.8 years. The four CNs included two female and two male decedents (Table 1).

Brain tissue from autopsied decedents
We examined eight consecutive autopsied decedents located at the National Hospital Organization Sendai Medical Center. Four of the decedents were diagnosed with IPD, and four of the decedents were used as CNs (Table 2). CNs were defined as follows: clinical symptoms were not observed, and abnormal p-aSyn, amyloid b or phosphorylated tau aggregates were not observed in brain specimens from autopsied individuals. The ages of the decedents ranged from 62 to 87 years, with an average of 74.8 years. The entire brain tissue was fixed in 20% buffered formalin (Wako, Osaka, Japan) for 5-11 days. Serial coronal sections were dehydrated in a graded ethanol series, cleared in xylene, and embedded in paraffin using an automated tissue processor.

Biopsied skin tissue
According to a previous report suggesting that postmortem skin specimens from patients with Braak stage III PD demonstrated the highest incidence of dermal p-aSyn deposits, 16 we recruited cognitively normal to mildly impaired IPD patients in the relatively early stage of IPD (Table 1). Skin specimens were collected from the cervical C8 paravertebral area, that is, close to the spinal ganglia, 17 and from an axilla in IPD patients. Skin specimens were collected from an upper arm in CNs. Skin biopsies were performed using a 3-mm disposable punch biopsy kit under sterile conditions after topical anesthesia with lidocaine. In IPD and CNs, skin biopsy specimens were fixed in 4% paraformaldehyde for 24-48 h.

Quantitative analysis of IF staining
To examine the number of macrophages and macrophages containing p-aSyn deposits in skin sections, we measured the number of p-aSyn-immunoreactive (ir) positive deposits, Iba1-ir positive macrophages, and Iba1-ir and p-aSyn double-positive macrophages under 20-fold magnified fields in the dermis in the region of interest of one section.
To examine the numbers of adrenergic fibers and adrenergic fibers containing p-aSyn deposits in skin sections, we measured the numbers of p-aSyn-ir positive deposits, DBH-ir adrenergic fibers, and DBH-ir and p-aSyn doublepositive adrenergic fibers under 40-fold magnified fields in the dermis in the region of interest of one section. To examine microglial cells, p-aSyn aggregates and microglial cells containing p-aSyn aggregates in brain autopsy specimens, we identified Iba1-ir positive microglial cells, aSyn-ir positive aggregates, and p-aSyn-ir and Iba1-ir doublepositive microglial cells under 20-fold magnified fields in the region of interest. For the analysis of the entorhinal cortex (EC) and amygdala, the gray matter area in the same region was defined as the region of interest.

Quantitative analysis of neuronal loss with hematoxylin and eosin staining
To evaluate neuronal loss in brain autopsy specimens, we counted the number of nuclei in neurons by hematoxylin and eosin (H&E) staining under 20-fold magnified fields. The nuclei of neurons were identified morphologically by hematoxylin staining.

Statistical evaluation
All values are expressed as the median (interquartile range). Significant differences between groups were identified using Wilcoxon rank sum test. Spearman's rank correlation coefficient was used to correlate the number of macrophages and the number macrophages with of p-aSyn deposits. The data were analyzed with a JMP13 computer software system (SAS Institute, Tokyo, Japan). A p < 0.05 denoted a statistically significant difference.

Results
Microglia with p-aSyn aggregates in brain autopsy specimens To identify microglia with LB-related aSyn aggregates, a double-labeling IF study was performed by incubating sections of brain autopsy specimens with anti-p-aSyn mouse monoclonal antibody and anti-Iba1 rabbit polyclonal antibodies. Iba1 is a 17-kDa EF-hand protein that is specifically expressed in microglia and macrophages and upregulated during immune activation. 21  [p < 0.05]) were classified as the severe neuronal loss regions (SNR) because of the reduced number of neurons in IPD patients compared to CNs in these regions (Fig. 1E). Iba1-positive microglial cells with p-aSyn aggregates in brain autopsy specimens from IPD patients and CNs were analyzed. Interestingly, Iba1-positive microglia with p-aSyn aggregates were observed ( Fig. 1A and B) in the EC of brain autopsy specimens from IPD patients but not CNs (Fig. 1C).
Next, LB-related Iba1-positive microglial cells and p-aSyn aggregates in brain autopsy specimens from four IPD patients were analyzed in the SNR and MNR. LB- [p < 0.001]) of brain autopsy specimens from four IPD patients (Fig. 1G). In addition, Iba1-positive microglial cells with p-aSyn aggregates were observed in the MNR but not SNR (the number of Iba1positive microglial cells with p-aSyn aggregates, 6.5 [5-8.75] vs. 0; [p < 0.001]) of brain autopsy specimens from four IPD patients (Fig. 1H). Thus, in the brains of patients with IPD, the results suggest that Iba1-positive microglia containing LB-related aSyn aggregates are mainly localized in brain regions where neuronal cell death is not prominent. Our observation is in accordance with a previous report suggesting that aSyn aggregaterelated synaptic dysfunction may trigger later neurodegeneration in IPD. 22  To investigate the distribution of abnormal aSyn deposits in skin specimens from IPD patients, a single-label IF study was performed by incubating skin sections from IPD patients with anti-p-aSyn (phospo-Ser129, mouse monoclonal). p-aSyn deposits were observed in both the posterior cervical C8 paravertebral and axillary skin of all IPD patients. We identified Iba1-ir microglial cells with p-aSyn aggregates in the central nervous system (CNS) of PD patients. This suggests that microglial cells are directly involved in the degradation of aSyn aggregates. Thus, we hypothesized that the presence of p-aSyn deposits in dermal Iba1-positive macrophages, which are analogous to microglia, could be a biomarker based on immunological mechanisms in IPD. Based on this hypothesis, we examined whether macrophages in the axillary skin from IPD patients contained p-aSyn deposits. The double-labeling IF study with anti-p-aSyn mouse monoclonal antibody and anti-Iba1 rabbit polyclonal antibodies revealed macrophages with p-aSyn deposits in the dermis of IPD patients but not CNs ( Fig. 2A). Iba1-ir-positive macrophages with p-aSyn deposits were found diffusely in the dermis of IPD patients but not CNs (Fig. 2C). The double-labeling IF study with anti-p-aSyn mouse monoclonal antibody and anti-CD163, another marker of macrophages, rabbit polyclonal antibodies also revealed macrophages with p-aSyn deposits in the dermis of IPD patients but not CNs (Fig. 2D). p-aSyn deposits were colocalized with p62/SQSTM1, a cytoplasmic protein associated with autophagy, in the dermis of IPD patients but not CNs (Fig. 2E). The number of p-aSyn-ir and Iba1-ir double-positive macrophages in the dermis were observed in 10 IPD patients but not four CNs (the number of p-aSyn-ir and Iba1-ir double-positive cells, 10 [5.75-20] vs. 0; [p < 0.01]) (Fig. 3A). There was no difference in macrophages with p-aSyn deposits staining pattern between the cervical C8 paravertebral and axillary skin in all IPD patients.

Adrenergic nerve fibers with p-aSyn deposits in skin specimens
In the skin, adrenergic DBH-ir fibers mainly exist around arterioles. It is reported that p-aSyn deposits in adrenergic fibers in the cervical C8 paravertebral skin from IPD patients. 23 Thus, in this study double-labeling IF studies were performed by incubating sections with anti-p-aSyn mouse monoclonal antibody and anti-DBH rabbit polyclonal antibodies in the cervical C8 paravertebral skin sections obtained from IPD patients and CNs. The doublelabeling IF studies revealed the localization of p-aSyn deposits in DBH-ir adrenergic vasomotor fibers in the dermis of IPD patients but not CNs (Fig. 2B). p-aSyn deposits in DBH-ir adrenergic vasomotor fibers were mainly observed around blood vessels in the dermis of IPD patients. The % p-aSyn-ir and DBH-ir doublepositive adrenergic fibers among DBH-ir adrenergic fibers were observed in the dermis of 10 IPD patients but not four CNs (the % p-aSyn-ir and DBH-ir double-positive adrenergic fibers among DBH-ir adrenergic fibers, 73.2 [57.5-80] vs. 0; [p < 0.01]) (Fig. 3B).

Correlation between p-aSyn deposits and macrophages in skin specimens from IPD patients
Spearman's rank correlation coefficient was used to correlate the number of Iba1-ir macrophages and the number of Iba1-ir macrophages with p-aSyn deposits in the dermis of IPD patients. Interestingly, the number of Iba1-ir macrophages was significantly positively correlated with the number of Iba1-ir macrophages with p-aSyn deposits in the dermis of IPD patients. (q = 0.9693, p < 0.0001) (Fig. 3C). These results suggest that the increase in aSyn deposition in dermal macrophages of IPD patients may be directly related to an increase in macrophages.   (Fig. 3D). These results suggest that p-Syn deposits were observed not only in adrenergic fibers but also in macrophages in the dermis.

Discussion
Microglia with p-aSyn aggregates may indicate an early stage of neurodegeneration in IPD patients The microglial cells containing p-aSyn aggregates were mainly localized in brain regions where neuronal cell death was not remarkable. Activated microglia within the CNS are detectable using macrophage markers, including major histocompatibility complex II and Iba1. 13,21 Depending on the inflammatory environment, microglia exhibit both phenotypic and functional changes from resting to activating states. Activated microglia are amoeboid (larger cell bodies, shorter, and thicker processes) and are morphologically similar to the macrophages in peripheral tissues. A correlation between microglial activation and DA terminal loss in the affected nigrostriatal pathway in a PD mouse model also supports the hypothesis that neuroinflammatory responses induced by activated microglia contribute to the early neurodegeneration process. 24,25 Other previous reports also suggest that microglial activation is associated with the early stage of neurodegeneration in PD models. 5,10 In addition, microglia are capable of clearing extracellular aSyn aggregates by internalization and degradation in PD models. 5,10 Although microglial activation in PD models has been suggested, the presence of aSyn aggregates in the microglia of human PD brains has not been reported. In this study, p-aSyn-positive round structures within Iba1positive cells were observed in the MNR of PD patients but not CNs. Thus, for the first time, we demonstrated that LB-related aSyn aggregates were observed in Iba1-ir microglial cells of autopsy brain specimens from IPD patients. In addition, p-aSyn-positive round structures within Iba1-negative cells were more observed in the SNR compared to MNR of IPD patients. On the other hand, p-aSyn-positive round structures within Iba1-positive microglial cells were observed in the MNR but not SNR of IPD patients. Thus, microglia with aSyn aggregates may indicate an early stage of neurodegeneration in IPD patients. We also examined whether amyloid b or phosphorylated tau aggregates were observed in the MNR of IPD patients by using immunohistochemical analysis. We could not observe abnormal amyloid b or phosphorylated tau aggregates in the MNR of PD patients (data not shown). Thus, in this study, we suggested that there was no influence of concomitant amyloid b and tau pathologies on microglial activation in the MNR. In summary, our results suggest that the phagocytosis of p-aSyn aggregates by microglia is a good marker for early-stage synucleinopathies.
p-aSyn deposits as a potential useful biomarker for IPD In this study, Iba1-ir microglial cells with p-aSyn aggregates in the CNS of IPD patients suggests that microglia are directly involved in the degradation of aSyn aggregates. Thus, we hypothesized that the presence of p-aSyn deposits in dermal Iba1-positive macrophages, which are analogous to microglia, could be a biomarker-based on immunological mechanisms in IPD. Although aSyn accumulation in human induced pluripotent stem cell macrophages from PD patients has been observed, 15 macrophages with p-aSyn deposits in human skin tissues have not been reported. Therefore, we examined whether macrophages in the dermis of IPD contain p-aSyn deposits. In this report, we first demonstrated dermal macrophages with p-aSyn deposits in IPD. Moreover, we propose that not only dermal nerve terminals containing p-aSyn deposits but also dermal macrophages with p-aSyn deposits are good markers to differentiate IPD patients from CNs. We also demonstrated that the number of Iba1-ir macrophages significantly correlated with the number of Iba1-ir macrophages with p-aSyn deposits in the dermis in IPD patients. Increasing evidence suggests that the spread of aSyn is the underlying mechanism for LBrelated neurodegeneration. 5,6 In addition, it has been suggested that microglia in the CNS, which are analogous to macrophages, may clean extracellular aSyn and prevent the spread of aSyn-related pathology. 26 Phillips et al. showed that macrophages adjacent to dystrophic nerve terminals were swollen and contained insoluble aSyn in the myenteric plexus of aged rats. 14 Haenseler et al. studied the removal role in aSyn using human induced pluripotent stem cell macrophages from PD patients. In this study, macrophages have the ability to remove fibrillary aSyn by phagocytosis, but high levels of exogenous or endogenous aSyn may reduce this ability and cause PD. 15 These results suggest that macrophages play an active removal role in aSyn. Thus, our results indicated that Iba1-ir peripheral macrophages were directly related to p-aSyn deposits in the dermis in IPD patients in the present study.
In the present study, we demonstrated that p-aSyn deposits in biopsied skin tissues of IPD patients were coexpressed with the cytoplasmic autophagy-related protein p62. aSyn aggregates are also preferred targets for p62-dependent autophagy in IPD. 27,28 Recently, aSyn aggregates with p62 accumulation were shown to be induced by autophagy dysfunction in DA neuron-specific autophagy-deficient mice. 20 Thus, our results suggested that p-aSyn deposits with p62 accumulation in biopsied skin tissues of IPD patients might be degraded by the autophagy pathway.
Precise IPD diagnosis is important for predicting prognosis, selecting therapeutic options, and conducting appropriate clinical studies. To improve the accuracy of IPD diagnosis, several ongoing studies aim to identify biomarkers. Possible biomarkers include olfactory function, cerebral spinal fluid tests, and cardiac scintigraphy. [29][30][31][32][33][34] Since specific biomarkers have not yet been identified to distinguish IPD from other neurodegenerative parkinsonisms, a novel biomarker with high specificity and sensitivity is required. In the present study, dermal p-aSyn deposits were observed in 10 (100%) of 10 IPD patients in the posterior cervical paravertebral skin. Several studies demonstrated that the incidence of dermal p-aSyn staining in distal skin 18,35 was less than that in proximal skin in IPD patients. 17,23,36,37 Interestingly, dermal p-aSyn deposits were observed in 80-100% of IPD patients when proximal cervical skin was used as the skin biopsy site. 17,23,36,37 Our results are in good accordance with previous studies demonstrating that cervical skin might be a promising biopsy site. We also observed dermal p-aSyn deposits in the axillary skin of 10 (100%) of 10 IPD patients. Thus, we suggest that p-aSyn deposits in axillary skin and posterior cervical skin might be a good target biomarker for IPD, and the former might be a more cosmetically suitable site for skin biopsy.
Here, we identified p-aSyn deposits in the DBH-ir adrenergic nerve vasomotor fibers of IPD patients. Our results were in accordance with several studies that observed dermal p-aSyn deposits in postganglionic adrenergic fibers 16,23,35 in IPD patients. Thus, the detection of p-aSyn deposits in the skin autonomic fiber terminal may be a useful biomarker for IPD diagnosis.

Limitations of this research
There are several limitations of this research. First, the number of samples was relatively small. Second, crosssectional longitudinal studies need to be performed.
Third, and possibly the major limitation of this study, the IPD patients from whom skin biopsy specimens were taken were not genetically and pathologically diagnosed. Fourth, autopsy skin specimens of autopsy confirmed PD cases were not available, so IF study of these specimens could not be performed. Fifth, the skin biopsy samples are from different areas between IPD patients (C8 and axilla) and CNs (upper arm). We chose C8 and axilla to obtain highest incidence of dermal p-aSyn deposits in IPD. Collaborators have reported that 149 of 1241 consecutive autopsy cases in the elderly were incidental LB. 38 We considered that skin specimens with autopsy confirmed CNs were necessary for this study. Therefore, although skin biopsy samples are from different areas, the autopsy confirmed CNs obtained from the collaborator was included in this analysis. Sixth, this study was analyzed by IF staining only. Electron microscopic analysis was not performed. Thus, electron microscopy should be performed in future studies.

Summary of the results
The main results of our study are as follows: (1) p-aSyn deposits were detected in DBH-ir adrenergic nerve vasomotor fibers and in Iba1-positive macrophages in skin specimens; and (2) the number of macrophages was significantly positively correlated with the number of macrophages with p-aSyn deposits in the dermis of IPD patients. These results suggest that p-aSyn deposits in dermal macrophages, as well as adrenergic fibers are useful biomarkers to differentiate IPD patients from CNs and that peripheral macrophages are directly related to p-aSyn deposits in the dermis in IPD. These results provide new insights into our understanding of pathophysiological processes in synucleinopathies.