Neurodegenerative Disease-Related Proteins within the Epidermal Layer of the Human Skin

Abstract

There is increasing evidence suggesting that amyloidogenic proteins might form deposits in non-neuronal tissues in neurodegenerative disorders such as Alzheimer’s or Parkinson’s diseases. However, the detection of these aggregation-prone proteins within the human skin has been controversial. Using immunohistochemistry (IHC) and mass spectrometry tissue imaging (MALDI-MSI), fresh frozen human skin samples were analyzed for the expression and localization of neurodegenerative disease-related proteins. While α-synuclein was detected throughout the epidermal layer of the auricular samples (IHC and MALDI-MSI), tau and Aβ₃₄ were also localized to the epidermal layer (IHC). In addition to Aβ peptides of varying length (e.g., Aβ₄₀, Aβ₄₂, Aβ₃₄), we also were able to detect inflammatory markers within the same sample sets (e.g., thymosin β-4, psoriasin). While previous literature has described α-synuclein in the nucleus of neurons (e.g., Parkinson’s disease), our current detection of α-synuclein in the nucleus of skin cells is novel. Imaging of α-synuclein or tau revealed that their presence was similar between the young and old samples in our present study. Future work may reveal differences relevant for diagnosis between these proteins at the molecular level (e.g., age-dependent post-translational modifications). Our novel detection of Aβ₃₄ in human skin suggests that, just like in the brain, it may represent a stable intermediate of the Aβ₄₀ and Aβ₄₂ degradation pathway.

Keywords

Aging α-synuclein amyloid-β human skin neurodegenerative proteins tau

INTRODUCTION

Proteinopathic neurodegenerative diseases, such as Alzheimer’s (AD) and Parkinson’s (PD) diseases, are associated with the misfolding of proteins characterized by aggregate formation, toxicity, and deposition. For example, the oligomerization of amyloid-β (Aβ) peptides and aggregation of tubulin-associated unit (tau) protein contribute to the observed pathology of AD [1], whereas α-synuclein is linked genetically and neuropathologically to PD [2]. Central to AD onset and pathogenesis, the sequential cleavage of the amyloid-β precursor protein (AβPP) by β- and γ-secretases generates Aβ peptides ranging from 30 to 51 amino acid residues in length [3]. Possessing unique aggregation and toxic properties, Aβ₄₀ and Aβ₄₂ are the most prevalent species detected in the brains of cognitively impaired patients. While Aβ₄₀ is typically less toxic, Aβ₄₂ is prone to aggregation, rendering it highly toxic. As semi-stable intermediates of both the Aβ₄₀ and Aβ₄₂ product lines [4], shorter peptides like Aβ₃₄ [5] may represent putative markers of amyloid degradation.

Equally critical to AD pathology is tau, a normally soluble cytoplasmic protein that possesses multiple phosphorylation sites, binds tubulin, and participates in microtubule formation [6]. Altering the native phosphorylation state of tau (i.e., increased phosphorylation by kinases and/or diminished dephosphorylation by phosphatases) results in the atypical ‘neurofibrillary tangles’ found inside neurons. In PD patients, atypical α-synuclein inclusion bodies have been detected in several regions of the brain and, interestingly, in the skin [7].

A major risk factor for many neurodegenerative diseases such as AD or PD is aging. Since skin is the largest organ in the body and has the same embryonic origin as brain tissue during development, it was hypothesized that the significant AD- and PD-associated protein accumulations that manifest in the brain may also be prevalent in the skin [8]. During the 1990s, a limited number of groups tested for the presence of Aβ in human skin as a potential diagnostic marker for AD and other neurological diseases [9, 10]. Notably, Wen and co-workers demonstrated that AβPP and Aβ were present in the dermal-epidermal junction (DEJ) and in blood vessels of skin biopsies from AD and Down syndrome patients.

More recent literature suggests that in addition to production, inefficient clearance mechanisms may contribute to the abnormal protein deposits in age-associated neurodegenerative disorders such as AD and PD [11]. Indicative of the growing parallel between brain and skin tissues [12, 13], recent work has detected α-synuclein deposits in skin biopsies from PD patients [14]. UVB-induced photo-oxidative stress has also been shown to induce the fibrillization of crystallin protein mixtures [15], a mechanism that can lead to the formation of highly toxic fibril-incompetent oligomers of α-synuclein [16]. This implies that the aggregation and accumulation of amyloidogenic proteins associated with AD (i.e. Aβ and tau) and PD (i.e., α-synuclein) may be sensitive to oxidative stress-induced protein modifications during the initial stages of pathogenesis [17, 18].

Our current study has used complementary techniques, IHC combined with leading-edge matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI), to test for neurodegenerative disease-related targets in human skin samples (auricular region) from young versus aged donors. The innovative antibody-independent MALDI-MSI approach, commonly performed as low as 20 or 50 μm resolution, was used as an initial target screening approach to test for the detection of amyloidogenic proteins within the skin samples. Interesting targets were then confirmed by IHC staining and analyzed on a sub-cellular level. Overall, our current study provides important new insights to the presence of neurodegenerative disease-related targets in human skin.

MATERIALS AND METHODS

Antibodies

The primary antibodies acquired from commercial sources included: anti-synuclein (Novus Biologicals #NBP2-15365, diluted 1 : 100), anti-tau (HT7; ThermoFisher #MN1000, diluted 1 : 50), anti-phosphorylated tau (PHF-tau; ThermoFisher #MN1040, diluted 1 : 50), anti-human thymosin β4 (ImmunDiagnostik #A9522.1, diluted 1 : 50). In-house primary antibodies from the Multhaup laboratory included: anti-AβPP / Aβ residues 4–10 (mAb W02, diluted 1 : 50), anti-Aβ₃₄ (mAb 226, diluted 1 : 100), anti-Aβ₄₀ (mAb G2–10, diluted 1 : 100), anti-Aβ₄₂ (mAb G2–13, diluted 1 : 1000). The secondary antibodies were acquired from Life Technologies: goat anti-rabbit IgG cross adsorbed Alexa Fluor 647 (ThermoFisher catalog #A-21245; diluted 1 : 10,000) or goat anti-mouse IgG cross adsorbed Alexa Fluor 647 (ThermoFisher catalog #A-21235; diluted 1 : 10,000).

Human skin samples

This study was approved by the Ethical Committee of the Hospital of Poitiers (Poitiers Cedex, France) and informed written consent was obtained for each sample by the donors themselves or by their legal guardian for samples obtained from minors. The samples selected from the BIOalternatives SAS repository (Gençay, France) were surgical waste and were not collected as part of a clinical study. Fresh frozen human skin samples were from the auricular region of healthy subjects undergoing an otoplasty (young group) or a lifting surgery (aged group). The “young” sample group included: #864, female, age 8; #963, female, age 9; #991, male, age 9; #990, male, age 10; #498, male, age 10; #965, female, age 11; #966, female, age 11. The “aged” sample group included: #1407, female, age 56; #796, female, age 60; #1042, female, age 62; #848, female, age 65; #920, female, age 65; #1293, female, age 75; #813, female, age 79. The aged group was only composed of females reflecting the fact that very few males undergo lifting surgery.

Sample sectioning

The fresh frozen skin samples were mounted onto pre-chilled blocks, using Optimal Cutting Temperature (OCT) compound, prior to microtome sectioning by the GCRC Histology Facility (McGill Life Sciences Complex). Serial sections (10 μm-thick) for IHC were mounted onto positively-charged microscope slides (Assure+microscope slides, Epic Scientific™, Oregon, USA). Serial sections (10 μm-thick) for MALDI-MSI were specially thaw-mounted onto conductive, indium tin oxide (ITO)-coated microscope slides (Bruker Daltonics #8237001).

H&E staining

For each skin sample, a single serial section was stained using hematoxylin and eosin (H&E). The H&E-stained sections were imaged at the Advanced BioImaging Facility (ABIF; McGill Life Sciences Complex) using an upright microscope (Axioskop 2 plus) with a 10x objective lens (Zeiss).

Matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI)

At the SPR-MS Facility (McGill Life Sciences Complex), the mounted ITO slides were stored under vacuum in a silica-containing desiccator to dry overnight at room temperature. The dried slides were sequentially washed in 75% and 90% isopropanol (1 min each to reduce background interferences from lipids) before the mounted sections were co-crystallized with super-DHB matrix (Sigma #50862) using an ImagePrep sprayer (Bruker Daltonics; 10 mg/mL super-DHB in a 50 : 50 mix of acetonitrile and 0.1% (v/v) trifluoroacetic acid). MSI spectra were acquired using a Bruker UltrafleXtreme MALDI-TOF/TOF system in linear positive mode (FlexControl v3.4 software; 3,000–50,000 m/z and 100 μM laser diameter). Spectra were processed using FlexImaging v4.1 software and normalized to total ion current (TIC).

Immunohistochemistry (IHC)

The mounted IHC sections were fixed in fresh ice-cold 4% (v/v) paraformaldehyde then washed with cold PBS buffer. The IHC slides were then permeabilized and blocked for 1 h at room temperature (in PBS containing 3% (v/v) horse serum, 0.3% (v/v) Triton X-100, and 0.05% (w/v) sodium azide) before overnight incubation at 4C with primary antibody. The next day, the IHC slides were washed and then incubated with secondary antibody, followed by another round of washing before the nuclei were stained with Hoechst.

Competition assays to determine antibody specificity

Briefly, ten-fold excess of either recombinant α-synuclein (generously provided by Dr. Edward Fon, Montreal Neurological Institute) or synthetic Aβ₃₄ (human sequence from Peptide Speciality Laboratories GmbH, Germany) was co-incubated with the primary antibody in the blocking buffer (PBS containing 3% (v/v) horse serum, 0.3% (v/v) Triton X-100, and 0.05% (w/v) sodium azide) overnight at 4C. The next day, the IHC slides were processed as described in the IHC section.

Confocal microscopy and image analysis

Single- or double-immunolabeled (Alexa Fluor-568 or -647) samples were analyzed at the Imaging & Molecular Biology Platform (IMBP; McGill Life Sciences Complex) using a TCS SP8 multi-photon confocal microscope (Leica) with either HC PL APO CS2 40x/1.30 or 63x/1.40 oil-immersion objectives (Leica, Wetzlar, Germany). Collagen structures were visualized using second harmonic generation (SHG). The fluorochromes were excited using a 552 nm and 638 nm solid state laser line, and the emission detected by photomultiplier tube photodetectors. The tissue was also excited with Coherent Chameleon Vision II multiphoton at 730 nm (2660 mW) for Hoechst imaging and at 900 nm (2370 mW) for SHG-collagen imaging. For each sample, 10–12 z-stack images at 1024× 1024 resolution and 0.35 μm thickness were acquired using the same laser intensity settings for quantification. Z-stack images were processed using the IMARIS Image Analysis Software (Bitplane (Oxford Instruments), MA, USA). Z-stacks were reconstructed in 3-dimension and subsequently quantified using the surface tool in IMARIS.

RESULTS

Preliminary staining with hematoxylin and eosin (H&E) allowed for the general anatomy of each tissue sample to be examined (Supplementary Figure 2). Overall, the H&E staining showed that while some samples showed thick epidermis, others showed a rather thin epidermis and the dermal integrity of the tissues did vary between tissues. This variation can be expected as the epidermal thickness of the skin naturally decreases with aging [19]. Moreover, the epidermal layer thickness of the skin is proportional to the body mass [20]. The dermal structures will evidently be very different from one tissue to another, depending on the sections used. In addition, the histological processing of the tissue might also affect the dermal integrity (for an overview of skin physiology, see Supplementary Figure 1).

MALDI-MSI is an innovative, label-free technique that allows for the simultaneous detection of multiple targets within a single tissue section [21 –23]. To detect peptides and low molecular weight proteins (<20 kDa) in fresh frozen samples, serial tissue sections were mounted onto conductive ITO-coated glass slides, washed briefly with a solvent gradient (e.g., 75% and 90% isopropanol to remove contaminant lipids), and sprayed with an ionization matrix [24, 25]. The MALDI mass spectrometer then scans back and forth across the co-crystallized tissue sections (resolution typically set between 20 and 200 μm), and intact peptides and proteins are detected based on their unique mass-to-charge ratios (m/z). Using the appropriate m/z filters, MSI imaging software then creates the color-coded “heat maps” to indicate the location(s) and relative intensity of each candidate peptide or protein within the tissue section.

MALDI-MSI has been elegantly used to reveal the distribution of Aβ peptides in the brains of aged people as well as in AD or cerebral amyloid angiopathy brains [26]. While sinapinic acid matrix is very efficient for detection of Aβ peptides and numerous other targets in brain tissue, we found that the use of this matrix on skin tissue did not yield good ionization. We developed a standardized MSI protocol in which super-DHB matrix was optimal for the detection of our targets of interest in the fresh frozen auricular samples. Acquired in linear positive mode (i.e., intact masses), the MALDI spectra were analyzed with the appropriate mass filters (m/z) to create “heat maps” for the candidate peptides and proteins within each tissue section. Under similar assay conditions, intact α-synuclein (14,460 m/z) was detectable in all auricular samples tested (Fig. 1A).

Fig.1

MALDI-MSI of auricular samples. Detection of predicted intact mass for α-synuclein (14,460 m/z) in young (ID # 498, 864, 963, 965, 966, 990, 991) and old samples (ID # 796, 813, 848, 920, 1042, 1293, 1407). Coloured max/min scale bar indicates the relative intensity of the α-synuclein signals within each heat map for the different sections.

Knowing that the ratio between the Aβ₄₀ and Aβ₄₂ peptides is often increased to reflect a higher percentage of Aβ₄₂ in brains of AD patients [27], it was interesting to note that Aβ₄₀ (4,330 m/z) was detectable in most auricular samples (Fig. 2B) whereas Aβ₄₂ (4,512 m/z) was only detectable in the aged samples (Fig. 2C). Since Aβ₄₀ and Aβ₄₂ have been previously shown to be degraded through two distinct pathways that ultimately produce the same Aβ₃₄ intermediate [5], we detected candidate peptides for the Aβ₃₄ (3,785 m/z) degradation product in most auricular samples (Fig. 2A).

Fig.2

MALDI-MSI of auricular samples, both young (ID # 498, 864, 963, 965, 966, 990, 991) and old (ID # 796, 813, 848, 920, 1042, 1293, 1407). Detection of predicted intact masses for A) Aβ₃₄ (3,785 m/z), B) Aβ₄₀ (4,330 m/z), and C) Aβ₄₂ (4,512 m/z). Coloured max/min scale bars indicate the relative intensity of the corresponding Aβ species within each set of heat maps for the different sections.

In addition to the candidate targets of interest (e.g., α-synuclein, Aβ₃₄, Aβ₄₀, Aβ₄₂), our MALDI-MSI protocol was also able to detect inflammation-related proteins (e.g., thymosin β-4, psoriasin; Supplementary Figure 3) and other ubiquitous proteins (e.g., fatty acid binding protein, ubiquitin; Supplementary Figure 4) in the same auricular sections. Notably, the overall signal intensities for thymosin β-4 and psoriasin appeared to be similar between the young and aged samples. Across the 14 different serial sections, the variable detection of some peptides/proteins might be linked to their abundance in the tissue sections (e.g., Aβ₄₀ in some but not all auricular samples), the specific sampling region, and/or technical concerns (i.e., small, intact proteins like the 14 kDa α-synuclein are generally easier to ionize by MALDI compared to large, intact proteins like 46 kDa tau).

In parallel, we developed optimized IHC protocols to successfully cross-validate the MALDI-MSI results at higher, sub-cellular resolution using classical confocal microscopy. For localization and image analysis purposes, Hoechst stain was used to visualize the nuclei of all cells, and second harmonic generation (SHG) was used to visualize collagen structures in a label-free fashion. Both unstained control sections and secondary antibody controls for Alexa 647 did not exhibit any fluorescence signal (Supplementary Figures 7 and 8), as compared to the corresponding secondary antibody controls for Alexa 568 which exhibited a faint background. We have therefore chosen Alexa 647 secondary antibody for all subsequent staining. α-synuclein, tau and Aβ₃₄ yielded reliable signals within the epidermal layer (Figs. 3–5). We have also performed competition assays using recombinant α-synuclein or synthetic Aβ₃₄ peptide (Supplementary Figure 6). While this report focuses on the epidermal layer, it is important to note that these targets yielded staining within ductal regions of the dermis.

Fig.3

Analysis of α-synuclein by IHC in auricular samples. Shown here are representative images of either old (A) (ID # 813, 848, 920, 1293), or young (B) (ID# 864, 963, 990, 991) samples; nuclei were visualized by Hoechst. Z-stack images were taken using the 40x objective on the Leica TCS SP8 with 1.0 optical zoom, and maximum projections created using the IMARIS software (Bitplane). Scale bar = 20 μm. C) The nuclear and extra-nuclear surfaces generated through image analysis by IMARIS software was used to calculate a fluorescence ration of nuclear synuclein to extra-nuclear synuclein.

Fig.4

Analysis of tau by IHC in auricular samples. A) Shown here are representative images of either young (ID #498, 864) and old (ID #813, 1293) samples; nuclei were visualized by Hoechst. Z-stack images were taken using the 40x objective on the Leica TCS SP8 with 1.0 optical zoom, and maximum projections created using the IMARIS software (Bitplane). Scale bar = 20 μm. B) The nuclear and extra-nuclear surfaces generated through image analysis by IMARIS software was used to calculate a fluorescence ration of nuclear tau to extra-nuclear tau (two-tailed Student’s t-test, p-value 0.1465).

Fig.5

Analysis of Aβ₃₄ by IHC in auricular samples. Shown here are representative images of either (A) old (ID #1293, 1407, 1042) or (B) young (ID #965, 966, 992); nuclei were visualized by Hoechst. Z-stack images were taken using the 40x objective on the Leica TCS SP8 with 1.0 optical zoom, and maximum projections created using the IMARIS software (Bitplane). Scale bar = 20 μm.

Immunohistochemical staining demonstrated robust α-synuclein staining in auricular samples within the epidermal surface (Fig. 3), and localized staining within the dermis (i.e., ductal regions). Interestingly, closer inspection demonstrated possible nuclear staining. Indeed, analysis of z-stack confocal images using the IMARIS software, suggested a nuclear presence of this amyloidogenic protein. Briefly, a nuclear surface was generated using Hoechst staining and this surface was then used to exclude any voxels that are non-nuclear (Fig. 3 and Supplementary Figure 5). Using this approach, we could determine the approximate ratio of fluorescence signal in the nucleus of epidermal cells compared to the extra-nuclear fluorescence signal within the epidermis. The quantification supports our assumption that both young and old samples could have α-synuclein within the nucleus as well as extra-nuclear space.

The relatively high molecular weight of the intact tau protein could not be identified by MALDI-MSI in linear positive mode. Hence, we investigated the tissue distribution as well as subcellular localization of tau protein by IHC, as the expression of the protein in skin tissue has been confirmed by multiple studies in the past [28 –31]. Immunoreactivity against tau protein within skin tissue is again robust at the epidermal region (Fig. 4A) with only localized staining within the dermis (i.e., ductal regions). Analysis of the ratio of nuclear to extra-nuclear tau fluorescence intensity suggested that samples from young and aged donors showed nuclear and extra-nuclear tau fluorescence.

Since the Aβ₃₄ peptide, a common intermediate product of both the Aβ₄₀ and Aβ₄₂ degradation pathways [5], was detected by MALDI-MSI analyses, we performed complementary Aβ₃₄ localizations by IHC using a C-terminal neo-epitope specific antibody. Interestingly, IHC (Fig. 5) revealed that Aβ₃₄ seemed to localize exclusively to the stratum corneum as well as the stratum spinosum of the epidermal layer in samples #1407 and #1042, while the signal was predominant across all layers of the epidermis in other samples, such as in #965 and #992. Additional samples (i.e., #1293 and #966) did not show DEJ signal.

Despite the very different nature of the two techniques, the data obtained by MALDI-MSI and IHC on serial auricular tissue sections showed some correlation (Fig. 6). The representative MALDI-MSI data in Figure 6A, tissue sections analyzed using a mass filter for intact α-synuclein (14,460 m/z), exhibited both epidermal and dermal signals. By IHC, adjacent tissue sections stained with α-synuclein antibody also exhibited epidermal and dermal signals. Similarly, for the representative serial sections in Figure 6B, MALDI-MSI (mass filter selected for intact TYB4, 5053 Da) and IHC (stained with anti-TYB4 antibody) contained both epidermal and dermal signals. While most of the tissue samples analyzed by the two techniques showed a good correlation, some differences could be observed which likely reflects morphological differences between the serial sections (spaced 10 μm apart in the cryostat). The IHC findings of α-synuclein, tau, and Aβ₃₄ localization in epidermal and dermal structure analyzed within this investigation have been summarized in Figure 7.

Fig.6

Correlation between MALDI-MSI and IHC data. A) Representative MALDI-MSI and IHC images of synuclein in the auricular sample ID #848 and #990. B) Representative MALDI-MSI and IHC images of TYB4 in the auricular sample ID #963 and #991. C) Representative MALDI-MSI and IHC images of Aβ₃₄ in the auricular sample ID #1042 and #1407.

Fig.7

Schematic diagram of the human skin with the localization of α-synuclein (purple), tau (light blue), and Aβ₃₄ (orange) found in this study.

DISCUSSION

Oxidative stress (upon UV exposure) has been shown to induce aggregation and accumulation of the amyloidogenic proteins that are typically associated with the neurodegenerative disorders AD (i.e., Aβ and tau) and PD (i.e., α-synuclein). Our present study has combined innovative MALDI-MSI and classical IHC to advance our biological understanding of disease-associated, misfolded proteins in the skin. We anticipate that combining these distinct, yet complementary techniques will provide new opportunities to study the cellular metabolism of neurodegenerative disease-related targets in organs other than the brain.

MALDI-MSI is advantageous because it is a label-free method that can detect multiple targets simultaneously. In addition to α-synuclein and different Aβ peptides, we were able to detect other skin-related targets within the same sample sets such as thymosin-β4 (TYB4) and psoriasin, which are key markers of inflammation [32, 33]. TYB4 is the most abundant and biologically active member of the thymosin family in most mammalian cells, accounting for about 80% of total β-thymosin [34]. Aside from the well-characterized cytoskeletal functions of TYB4, it is also thought to be involved in angiogenesis, wound healing and apoptosis. This candidate protein detected by MALDI-MSI (5053 Da in literature) was cross-validated by classical IHC using an anti-TYB4 antibody. Both imaging approaches showed that TYB4 was readily detectable in all the auricular samples and localized to the epidermis and ductal regions of the dermis. On the other hand, psoriasin is an abundant protein in the permeability barrier, localized in the stratum corneum, serving as a mechanical protective hurdle against bacterial infection [32, 35]. The expression of this protein, which represents one of the first stages of antimicrobial defense, is not only altered in the presence of infections but also changed during localized stress such as in tumorigenesis [33]. Psoriasin’s intact mass was detectable in most of the samples by MALDI-MSI (Supplementary Figure 3).

Using both MALDI-MSI and IHC, we were able to detect α-synuclein in all samples analyzed. Seemingly, data from our study supports the localization of α-synuclein to the nucleus of epidermal cells. Upon exposure to stress, α-synuclein has been shown to localize to the nucleus of neurons [36]. Furthermore, this nuclear relocalization was accompanied by a colocalization with histones in the nuclei of nigral neurons in animals exposed to paraquat-induced oxidative stress. This suggests that α-synuclein might be important for DNA protection upon exposure to cellular stress. As such, in Saccharomyces cerevisiae, α-synuclein plays a role in the nucleus to protect against hydroxyurea-induced replication stress [37]. However, the exact role(s) of α-synuclein in this organelle remains elusive as it has also been proposed to act as a transcriptional modulator of the master mitochondrial transcription activator PGC1α in both in vitro and in vivo models of PD, where downregulation results in impaired mitochondrial morphology as well as activity [38]. Nevertheless, to date, there is not really evidence of nuclear α-synuclein in human skin tissue. One study investigating the presence of α-synuclein inclusions in the skins of Parkinson and parkinsonism patients demonstrated that PD skin tissue contained juxtanuclear α-synuclein staining while tissues from healthy controls did not contain α-synuclein inclusions [7].

As analyzed by IHC, similar to α-synuclein localization, we found that tau predominantly localized to the epidermis of the auricular region skin tissues analyzed. This is not surprising giving the physiological structure of the skin tissue with epidermal layer made of 90% keratinocytes while dermal layer being mainly composed of extracellular-matrix synthesized by scattered fibroblasts. It is well established that tau is expressed in human skin tissue. Over twenty years ago, a study revealed that fibroblasts isolated from skin biopsies of control individuals and those carrying the Swedish AD mutation (AβPP KM670/671NL) both expressed tau protein [39]. Just two years ago, a comprehensive study investigating the presence of tau species in human peripheral tissues found tau and phosphorylated-tau (pT231 or pS396 and pS404) in skin biopsies from both non-demented individuals and AD patients [40]. Some isoforms of tau protein have long been described in the nucleus of neuronal cells [41, 42]. Nuclear tau has then been shown to interact with DNA and play a role in nucleolar organization [43]. Recent studies have provided insight to the role of nuclear tau where it is suggested to be involved in DNA protection and repair in response to cellular stress [44, 45]. Whether tau plays a similar role outside of the central nervous system (CNS) remains to be elucidated. Interestingly, in human fibroblasts from frontotemporal dementia patients, tau has been shown to localize to the nucleus [46]. In a different study, primary cultures of human fibroblasts exposed to oxidative stress showed increased phospho-tau in the nuclei [47]. To date, few studies have shown a nuclear localization of tau outside the CNS, highlighting the novelty of our findings. Moreover, we show by IHC staining and nuclear to extra-nuclear tau immunofluorescence analysis of z stack confocal images, a nuclear localization of tau. It would be of interest to further investigate in a larger cohort with skin tissue exposed and unexposed to photo-oxidative stress. Likewise, we have previously demonstrated that nuclear Aβ could be detected in human neuroblastoma cells as well as in a transgenic AD mouse model, and this nuclear Aβ played a role in gene regulation [48 –50]. One could envision that in skin tissue from AD patients, Aβ might also localize to the nuclei of epidermal cells and participate in gene regulation.

We have shown using MALDI-MSI that intact masses for Aβ₃₄, Aβ₄₀, and Aβ₄₂ could be detected in aged human skin samples that we currently tested. The presence of Aβ₃₄, an Aβ species that has never been investigated in skin tissue before, was also confirmed by IHC using a neo-epitope specific antibody. In 1989, Joachim and co-workers were the first to suggest that Aβ deposited in skin tissue as well as intestines of healthy elderly control and AD patients [8]. Soininen and co-workers then showed that both non-demented and AD individuals had detectable levels of Aβ immunoreactivity in the corresponding skin tissue samples [51]. Findings from another group demonstrated that vessels within skin tissue were labeled by antisera against purified native cerebral Aβ peptide, and there was a significant difference between the control and AD groups [52]. Interestingly, it has been shown that patients with amyotrophic lateral sclerosis have increased levels of amyloid peptides in their skin [53]. Upon this observation, it is tempting to hypothesize that in the skin there is a metabolism similar to that of the brain tissues, that also reflects in the skin with aging. Aβ peptides have been reported to be produced outside the CNS (i.e., neuromuscular junctions) [54]. The amyloid-β protein precursor, AβPP, has a quite ubiquitous expression pattern with particularly high expression in brain tissue, followed by lung tissue. AβPP has also been reported to be expressed and processed in platelets [55]. However, the metabolism of Aβ peptides generated outside the CNS is not well understood today [56]. As both skin and brain tissue originate from embryonic ectodermis during development, it is not surprising that skin tissue also shows AβPP expression [57, 58]. This expression is thought to be localized predominantly in basal keratinocytes within the epidermal layer [59]. Due to the same origin during the development and a potential similar metabolism in both tissues, it is conceivable that epithelial cells could show comparable sensitivities to amyloidogenic proteins in a diseased state. However, one important difference between epidermal cells and neurons is the fact that epidermal cells are continuously renewed as opposed to neuronal cells. It is perhaps an evolutionary selected strategy to escape deleterious effects of amyloidogenic proteins as the epidermal layer of the human skin is directly exposed to environmental stress. Future studies with the inclusion of skin tissue from AD patients will shed light to possible metabolic parallels between these two organs in health and disease and could result in the development of novel diagnostic approaches.

Footnotes

ACKNOWLEDGMENTS

This work was supported from a L’Oréal Research and Innovation grant.

We thank L’Oréal Research and Innovation for their expertise (including Drs. Jessica Langer and Amin Samiul) and financial support. GM holds a Tier 1 CRC. Shared research resources at the McGill Life Sciences Complex (Advanced BioImaging Facility, ABIF; GCRC Histology Facility; Imaging & Molecular Biology Platform, IMBP; SPR-MS Facility) thank the Canada Foundation for Innovation (CFI) for infrastructure support.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Masters

, Bateman

, Blennow

, Rowe

, Sperling

, Cummings

(2015) Alzheimer’s disease. Nat Rev Dis Primers 1, 15056.

Lashuel

, Overk

, Oueslati

, Masliah

(2013) The many faces of alpha-synucle: From structure and toxicity to theraeutic target. Nat Rev Neurosci 14, 38–48.

Glabe

, Kayed

(2006) Common structure and toxic function of amyloid oligomers implies a common mechanism of pathogenesis. Neurology 66, S74–S78.

Munter

, Voigt

, Harmeier

, Kaden

, Gottschalk

, Weise

, Pipkorn

, Schaefer

, Langosch

, Multhaup

(2007) GxxxG motifs within the amyloid precursor protein transmembrane sequence are critical for the etiology of Abeta42. EMBO J 26, 1702–1712.

Olsson

, Schmidt

, Althoff

, Munter

, Jin

, Rosqvist

, Lendahl

, Multhaup

, Lundkvist

(2014) Characterization of intermediate steps in amyloid beta (Aβ) production under near-native conditions. J Biol Chem 289, 1540–1550.

Ballatore

, Lee

, Trojanowski

(2007) Tau-mediated neurodegeneration in Alzheimer’s disease and related disorders. Nat Rev Neurosci 8, 663–672.

Rodríguez-Leyva

, Calderón-Garcidueñas

, Jiménez-Capdeville

, Rentería-Palomo

, Hernandez-Rodriguez

, Valdés-Rodríguez

, Fuentes-Ahumada

, Torres-álvarez

, Sepúlveda-Saavedra

, Soto-Domínguez

, Santoyo

, Rodriguez-Moreno

, Castanedo-Cázares

(2014) α-Synuclein inclusions in the skin of Parkinson’s disease and parkinsonism. Ann Clin Transl Neurol 1, 471–478.

Joachim

, Mori

, Selkoe

(1989) Amyloid beta-protein deposition in tissues other than brain in Alzheimer’s disease. Nature 341, 226–230.

Heinonen

, Soininen

, Syrjänen

, Neittaanmäki

, Paljärvi

, Kosunen

, Syrjänen

, Riekkinen

(1994) beta-Amyloid protein immunoreactivity in skin is not a reliable marker of Alzheimer’s disease. An autopsy-controlled study. Arch Neurol 51, 799–804.

10.

Wen

, Wisniewski

, Blondal

, Benedikz

, Frey

, Pirttila

, Rudelli

, Kim

(1994) Presence of non-fibrillar amyloid beta protein in skin biopsies of Alzheimer’s disease (AD), Down’s syndrome and non-AD normal persons. Acta Neuropathol 88, 201–206.

11.

Deleidi

, Maetzler

(2012) Protein clearance mechanisms of alpha-synuclein and amyloid-Beta in lewy body disorders. Int J Alzheimers Dis 2012, 391438.

12.

Clos

, Kayed

, Lasagna-Reeves

(2012) Association of skin with the pathogenesis and treatment of neurodegenerative amyloidosis. Front Neurol 3, 5.

13.

Foster

(2012) The “brain-skin connection” in protein misfolding and amyloid deposits: Embryological, pathophysiological, and therapeutic common grounds? Front Neurol 3, 56.

14.

Rodriguez-Leyva

, Chi-Ahumada

, Carrizales

, Rodriguez-Violante

, Velazquez-Osuna

, Medina-Mier

, Martel-Gallegos

, Zarazua

, Enriquez-Macias

, Castro

, Calderon-Garciduenas

, Jimenez-Capdeville

(2016) Parkinson disease and progressive supranuclear palsy: Protein expression in skin. Ann Clin Transl Neurol 3, 191–199.

15.

Cetinel

, Semenchenko

, Cho

, Sharaf

, Damji

, Unsworth

, Montemagno

(2017) UV-B induced fibrillization of crystallin protein mixtures. PLoS One 12, e0177991.

16.

Carmo-Goncalves

, Pinheiro

, Romao

, Cortines

, Follmer

(2014) UV-induced selective oxidation of Met5 to Met-sulfoxide leads to the formation of neurotoxic fibril-incompetent alpha-synuclein oligomers. Amyloid 21, 163–174.

17.

Bisht

, Sharma

, Tremblay

(2018) Chronic stress as a risk factor for Alzheimer’s disease: Roles of microglia-mediated synaptic remodeling, inflammation, and oxidative stress. Neurobiol Stress 9, 9–21.

18.

Mule

, Singh

(2018) Diabetes mellitus to neurodegenerative disorders: Is oxidative stress fueling the flame? CNS Neurol Disord Drug Targets.

19.

Branchet

, Boisnic

, Frances

, Robert

(1990) Skin thickness changes in normal aging skin. Gerontology 36, 28–35.

20.

Wei

JCJ

, Edwards

, Martin

, Huang

, Crichton

, Kendall

MAF

(2017) Allometric scaling of skin thickness, elasticity, viscoelasticity to mass for micro-medical device translation: From mice, rats, rabbits, pigs to humans. Sci Rep 7, 15885.

21.

Chaurand

(2012) Imaging mass spectrometry of thin tissue sections: A decade of collective efforts. J Proteomics 75, 4883–4892.

22.

Fujimura

, Miura

(2014) MALDI mass spectrometry imaging for visualizing in situ metabolism of endogenous metabolites and dietary phytochemicals. Metabolites 4, 319–346.

23.

, Buck

, Balluff

, Sun

, Gorzolka

, Feuchtinger

, Janssen

, Kuppen

, van de Velde

, Weirich

, Erlmeier

, Langer

, Aubele

, Zitzelsberger

, McDonnell

, Aichler

, Walch

(2016) High-mass-resolution MALDI mass spectrometry imaging of metabolites from formalin-fixed paraffin-embedded tissue. Nat Protoc 11, 1428–1443.

24.

Seeley

, Oppenheimer

, Mi

, Chaurand

, Caprioli

(2008) Enhancement of protein sensitivity for MALDI imaging mass spectrometry after chemical treatment of tissue sections. J Am Soc Mass Spectrom 19, 1069–1077.

25.

Thomas

, Chaurand

(2014) Advances in tissue section preparation for MALDI imaging MS. Bioanalysis 6, 967–982.

26.

Kakuda

, Miyasaka

, Iwasaki

, Nirasawa

, Wada-Kakuda

, Takahashi-Fujigasaki

, Murayama

, Ihara

, Ikegawa

(2017) Distinct deposition of amyloid-beta species in brains with Alzheimer’s disease pathology visualized with MALDI imaging mass spectrometry. Acta Neuropathol Commun 5, 73.

27.

Patterson

, Elbert

, Mawuenyega

, Kasten

, Ovod

, Ma

, Xiong

, Chott

, Yarasheski

, Sigurdson

, Zhang

, Goate

, Benzinger

, Morris

, Holtzman

, Bateman

(2015) Age and amyloid effects on human central nervous system amyloid-beta kinetics. Ann Neurol 78, 439–453.

28.

Dugger

, Whiteside

, Maarouf

, Walker

, Beach

, Sue

, Garcia

, Dunckley

, Meechoovet

, Reiman

, Roher

(2016) The presence of select tau species in human peripheral tissues and their relation to Alzheimer’s disease. J Alzheimers Dis 51, 345–356.

29.

Makrantonaki

, Brink

, Zampeli

, Elewa

, Mlody

, Hossini

, Hermes

, Krause

, Knolle

, Abdallah

, Adjaye

, Zouboulis

(2012) Identification of biomarkers of human skin ageing in both genders. Wnt signalling - a label of skin ageing? PLoS One 7, e50393.

30.

Rodríguez-Leyva

(2015) Presence of phosphorylated tau protein in the skin of Alzheimer's disease patients. J Mol Biomark Diagn s6, 6–11.

31.

Jong

, Ford

, Seehra

, Malave

, Baenziger

(2003) Alzheimer’s disease skin fibroblasts selectively express a bradykinin signaling pathway mediating tau protein Ser phosphorylation. FASEB J 17, 2319–2321.

32.

Glaser

, Meyer-Hoffert

, Harder

, Cordes

, Wittersheim

, Kobliakova

, Folster-Holst

, Proksch

, Schroder

, Schwarz

(2009) The antimicrobial protein psoriasin (S100A7) is upregulated in atopic dermatitis and after experimental skin barrier disruption. J Invest Dermatol 129, 641–649.

33.

Alowami

, Qing

, Emberley

, Snell

, Watson

(2003) Psoriasin (S100A7) expression is altered during skin tumorigenesis. BMC Dermatol 3, 1.

34.

Hannappel

, Huff

(2003) The thymosins. Prothymosin alpha, parathymosin, and beta-thymosins: Structure and function. Vitam Horm 66, 257–296.

35.

Wolf

, Ruzicka

, Yuspa

(2011) Novel S100A7 (psoriasin)/S100A15 (koebnerisin) subfamily: Highly homologous but distinct in regulation and function. Amino Acids 41, 789–796.

36.

Goers

, Manning-Bog

, McCormack

, Millett

, Doniach

, Di

Monte DA

, Uversky

, Fink

(2003) Nuclear localization of alpha-synuclein and its interaction with histones. Biochemistry 42, 8465–8471.

37.

Liu

, Lee

, Liou

, Ren

, Zhang

, Wang

, Witt

(2011) Alpha-synuclein functions in the nucleus to protect against hydroxyurea-induced replication stress in yeast. Hum Mol Genet 20, 3401–3414.

38.

Siddiqui

, Chinta

, Mallajosyula

, Rajagopolan

, Hanson

, Rane

, Melov

, Andersen

(2012) Selective binding of nuclear alpha-synuclein to the PGC1alpha promoter under conditions of oxidative stress may contribute to losses in mitochondrial function: Implications for Parkinson’s disease. Free Radic Biol Med 53, 993–1003.

39.

Ingelson

, Vanmechelen

, Lannfelt

(1996) Microtubule-associated protein tau in human fibroblasts with the Sweedish Alzheimer mutation. Neurosci Lett 220, 9–12.

40.

Dugger

, Whiteside

, Maarouf

, Walker

, Beach

, Sue

, Garcia

, Dunckley

, Meechoovet

, Reiman

, Roher

(2016) The presence of select tau species in human peripheral tissues and their relation to Alzheimer’s disease. J Alzheimers Dis 51, 345–356.

41.

Loomis

, Howard

, Castleberry

, Binder

(1990) Identification of nuclear tau isoforms in human neuroblastoma cells. Proc Natl Acad Sci U S A 87, 8422–8426.

42.

Greenwood

, Johnson

(1995) Localization and in situ phosphorylation state of nuclear tau. Exp Cell Res 220, 332–337.

43.

Sjoberg

, Shestakova

, Mansuroglu

, Maccioni

, Bonnefoy

(2006) Tau protein binds to pericentromeric DNA: A putative role for nuclear tau in nucleolar organization. J Cell Sci 119, 2025–2034.

44.

Sultan

, Nesslany

, Violet

, Bégard

, Loyens

, Talahari

, Mansuroglu

, Marzin

, Sergeant

, Humez

, Colin

, Bonnefoy

, Buée

, Galas

(2011) Nuclear tau, a key player in neuronal DNA protection. J Biol Chem 286, 4566–4575.

45.

Mansuroglu

, Benhelli-Mokrani

, Marcato

, Sultan

, Violet

, Chauderlier

, Delattre

, Loyens

, Talahari

, Begard

, Nesslany

, Colin

, Soues

, Lefebvre

, Buee

, Galas

, Bonnefoy

(2016) Loss of Tau protein affects the structure, transcription and repair of neuronal pericentromeric heterochromatin. Sci Rep 6, 33047.

46.

Rossi

, Dalpra

, Crosti

, Lissoni

, Sciacca

, Catania

, Di

Fede G

, Mangieri

, Giaccone

, Croci

, Tagliavini

(2008) A new function of microtubule-associated protein tau: Involvement in chromosome stability. Cell Cycle 7, 1788–1794.

47.

Ibanez-Salazar

, Banuelos-Hernandez

, Rodriguez-Leyva

, Chi-Ahumada

, Monreal-Escalante

, Jimenez-Capdeville

, Rosales-Mendoza

(2017) Oxidative stress modifies the levels and phosphorylation state of tau protein in human fibroblasts. Front Neurosci 11, 495.

48.

Barucker

, Harmeier

, Weiske

, Fauler

, Albring

, Prokop

, Hildebrand

, Lurz

, Heppner

, Huber

, Multhaup

(2014) Nuclear translocation uncovers the amyloid peptide Aβ42 as a regulator of gene transcription. J Biol Chem 289, 20182–20191.

49.

Barucker

, Sommer

, Beckmann

, Eravci

, Harmeier

(2015) Alzheimer amyloid peptide Aβ42 regulates gene expression of transcription and growth factors. J Alzheimers Dis 44, 613–624.

50.

Multhaup

, Huber

, Buée

, Galas

M-C

(2015) Amyloid precursor protein (APP) metabolites APP intracellular fragment (AICD), Aβ42, and tau in nuclear roles. J Biol Chem 290, 23515–23522.

51.

Soininen

, Syrjänen

, Heinonen

, Neittaanmäki

, Miettinen

, Paljärvi

, Syrjänen

, Beyreuther

, Riekkinen

(1992) Amyloid beta-protein deposition in skin of patients with dementia. Lancet 339, 245.

52.

Ikeda

, Shoji

, Yamaguchi

, Matsubara

, Amari

, Ishiguro

, Kawarabayashi

, Harigaya

, Hirai

(1993) Diagnostic significance of skin immunolabelling with antibody against native cerebral amyloid in Alzheimer’s disease. Neurosci Lett 150, 159–161.

53.

Tamaoka

, Matsuno

, Ono

, Shimizu

, Shoji

(2000) Increased amyloid beta protein in the skin of patients with amyotrophic lateral sclerosis. J Neurol 247, 633–635.

54.

Askanas

, Engel

, Alvarez

(1992) Light and electron microscopic localization of beta-amyloid protein in muscle biopsies of patients with inclusion-body myositis. Am J Pathol 141, 31–36.

55.

Meyer GR

, De

Cleen DM

, Cooper

, Knaapen

, Jans

, Martinet

, Herman

, Bult

, Kockx

(2002) Platelet phagocytosis and processing of beta-amyloid precursor protein as a mechanism of macrophage activation in atherosclerosis. Circ Res 90, 1197–1204.

56.

Kokjohn

, Van Vickle

, Maarouf

, Kalback

, Hunter

, Daugs

, Luehrs

, Lopez

, Brune

, Sue

, Beach

, Castano

, Roher

(2011) Chemical characterization of pro-inflammatory amyloid-beta peptides in human atherosclerotic lesions and platelets. Biochim Biophys Acta 1812, 1508–1514.

57.

Puig

, Combs

(2013) Expression and function of APP and its metabolites outside the central nervous system. Exp Gerontol 48, 608–611.

58.

Kummer

, Wehner

, Quast

, Werner

, Herzog

(2002) Expression and potential function of beta-amyloid precursor proteins during cutaneous wound repair. Exp Cell Res 280, 222–232.

59.

Herzog

, Kirfel

, Siemes

, Schmitz

(2004) Biological roles of APP in the epidermis. Eur J Cell Biol 83, 613–624.