Background: Chronic neuroinflammation is a hallmark of Parkinson’s disease (PD) pathophysiology, associated with increased levels of pro-inflammatory factors in PD brain tissues. The pro-inflammatory mediator and highly amyloidogenic protein S100A9 is involved in the amyloid-neuroinflammatory cascade in Alzheimer’s disease. This is the first report on the co-aggregation of α-synuclein (α-syn) and S100A9 both in vitro and ex vivo in PD brain. Methods: Single and sequential immunohistochemistry, immunofluorescence, scanning electron and atomic force (AFM) microscopies were used to analyze the ex vivo PD brain tissues for S100A9 and α-syn location and aggregation. In vitro studies revealing S100A9 and α-syn interaction and co-aggregation were conducted by NMR, circular dichroism, Thioflavin-T fluorescence, AFM, and surface plasmon resonance methods. Results: Co-localized and co-aggregated S100A9 and α-syn were found in 20% Lewy bodies and 77% neuronal cells in the substantia nigra; both proteins were also observed in Lewy bodies in PD frontal lobe (Braak stages 4–6). Lewy bodies were characterized by ca. 10–23 μm outer diameter, with S100A9 and α-syn being co-localized in the same lamellar structures. S100A9 was also detected in neurons and blood vessels of the aged patients without PD, but in much lesser extent. In vitro S100A9 and α-syn were shown to interact with each other via the α-syn C-terminus with an apparent dissociation constant of ca. 5 μM. Their co-aggregation occurred significantly faster and led to formation of larger amyloid aggregates than the self-assembly of individual proteins. S100A9 amyloid oligomers were more toxic than those of α-syn, while co-aggregation of both proteins mitigated the cytotoxicity of S100A9 oligomers. Conclusions: We suggest that sustained neuroinflammation promoting the spread of amyloidogenic S100A9 in the brain tissues may trigger the amyloid cascade involving α-syn and S100A9 and leading to PD, similar to the effect of S100A9 and Aβ co-aggregation in Alzheimer’s disease. The finding of S100A9 involvement in PD may open a new avenue for therapeutic interventions targeting S100A9 and preventing its amyloid self-assembly in affected brain tissues. Keywords: S100A9, α-Synuclein, Parkinson’s disease, Neuroinflammation, Amyloid, Cytotoxicity * Correspondence: email@example.com Istvan Horvath, Igor A. Iashchishyn and Roman A. Moskalenko contributed equally to this work. Department of Medical Biochemistry and Biophysics, Umeå University, 90187 Umeå, Sweden Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 2 of 16 Background microenvironment in the brain tissues. PD brains PD is the most common age-dependent neurodegenerative showed extensive microglial activation, infiltration of movement disorder affecting about 2% of the population blood-derived mononuclear phagocytes and lymphocytes, over 60 years old. The hallmark of PD is pathological and significant rise of pro-inflammatory cytokines—all self-assembly of the amyloidogenic protein α-synuclein deleterious responses, which can sustain inflammation (α-syn), which forms cytotoxic amyloid oligomers and ma- and exacerbate neurodegeneration [11–13]. Moreover, ture fibrils in PD brain tissues [1, 2]. Native α-syn exists as epidemiological studies have demonstrated that nonsteroi- an intrinsically disordered monomer. In the presence of dal anti-inflammatory drugs may lower the risk of neuro- lipid membranes, α-syn undergoes a conformational degenerative diseases, including both Alzheimer’s disease change to a folded α-helical secondary structure . Al- and PD [12, 14]. though the biological functions of α-syn are debated, it has The importance of co-aggregation of amyloidogenic been suggested to play role in maintaining a supply of syn- proteins in a number of neurodegenerative diseases aptic vesicles in presynaptic terminals by clustering synap- has recently been highlighted . It has been shown tic vesicles, andtobe involvedinregulating the release of that molecules designed to inhibit aggregation of one the neurotransmitter dopamine in controlling voluntary amyloidogenic protein may inhibit the aggregation of and involuntary movements [2, 4]. Despite numerous stud- others [16, 17] and thus potentially affect the whole ies, the critical factors triggering α-syn aberrant conversion amyloid cascade. into pathological β-sheet-rich amyloid aggregates and con- In this study, we have focused on the involvement in sequently initiation of PD remain unclear. The majority of PD pathology of the specific pro-inflammatory mediator PD incidents are sporadic, but inherited α-syn mutations S100A9, which we have found to play a critical role in leading to its amyloid formation at early onset are present connecting neuroinflammatory and amyloid pathologies in familial PD, which constitute ca. 10–15% of all PD cases. into the integrated amyloid-neuroinflammatory cascade When α-syn assembles into amyloid fibrils, these aggre- in Alzheimer’s disease . S100A9 belongs to the gates accumulate in the form of Lewy bodies and Lewy family of structurally homologous calcium-binding S100 neurite deposits in neuronal cells primarily in the substan- proteins, which are broadly involved in many inflamma- tia nigra. These deposits are also found in the frontal lobe, tory, cancer, and neurodegenerative conditions . We vagus dorsal motor nuclei, nucleus basalis of Meynert, and have previously shown that S100A9 is highly amyloido- locus coeruleus [5–7]. Amyloid oligomers of α-syn are genic and easily forms amyloid oligomers and fibrils considered to be a major cause of neuronal cell toxicity under in vitro conditions that mimic physiological and progressive neurodegeneration . The latter leads to conditions [18, 20]. In Alzheimer’s disease, S100A9 was PD pathology, which manifests in the loss of motor func- found to be abundant both in neuronal cells and in amyl- tion, bradykinesia, rigidity, instability, and tremor. Apart oid plaques, prompting co-aggregation with amyloid-β from the pathogenic neurotoxicity of oligomeric α-syn, the (Aβ), the major amyloidogenic peptide in Alzheimer’s depletion of α-syn monomers in their physiological loca- disease . It has therefore been suggested that tions due to their aggregation may also contribute to neu- pro-inflammatory S100A9, which possesses intrinsic amy- rodegeneration . An interesting idea of prion-like loidogenic properties as well as the ability to modulate Aβ transmission of α-syn pathology between neuronal cells aggregation, can serve as a link between the Alzheimer’s has recently been proposed , highlighting that α-syn in disease amyloid and neuroinflammatory cascades and as a an aberrantly folded, β-sheet-rich conformation can mi- prospective therapeutic target . Interestingly, the crit- grate from affected to unaffected neurons, and thus trigger ical role of S100A9 in Alzheimer’s disease development amyloid templating in the host cells. This indicates that was demonstrated in a mouse model, where S100A9 pro- once aggregation begins, it can effectively spread to the duction was induced by both the Aβ peptide and the surrounding tissues, leading to PD progression. C-terminal fragment of the Aβ precursor protein, while Growing evidence has been accumulated concerning S100A9 knockdown attenuated memory impairment and the importance of additional causal factors that can reduced amyloid plaque burden . Among the brain contribute to, or turn on, the pathological cascade of pathologies, a widespread expression of S100A9 has been α-syn amyloid aggregation. The most obvious common reported in cerebral malaria , cerebral ischemia , denominator in major neurodegenerative diseases, in- Alzheimer’s disease [18, 24], and traumatic brain injury cluding PD and Alzheimer’s disease, is neuroinflamma- , where it may initiate sustainable inflammatory tion [10, 11]. Previously, neuroinflammation was simply responses and perform mediator functions controlling regarded as a response to neurodegeneration in these inflammatory responses of other cells. An abundance of diseases. However, recent studies suggest that neuroin- S100A9 mRNA has recently been identified as a strong flammation could be the trigger and the key player in feature of aging in various mammalian tissues, including neurodegenerative diseases by creating a pathogenic the central nervous system, and a novel mechanism of Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 3 of 16 age-associated inflammation sustained by S100A9 has Table 1 Characteristics of PD and control subjects, including age, gender, and stage of PD  been suggested . The role of S100A9 in PD remains to be elucidated. Age Gender Stage There are clear indications that another protein from PD patients 83 Female Braak 5 the S100 protein family—S100B—plays a role in PD [27, 69 Male Braak 5 28]. Elevated S100B levels were found in the postmor- 83 Male Braak 5 tem substantia nigra of PD patients compared with con- 82 Male Braak 6 trol tissues, and S100B levels in the cerebrospinal fluid 75 Male Braak 4 were also higher in a large cohort of PD patients com- Control patients 83 Female – pared with controls . Autoimmune responses to S100B in the blood sera of PD patients were also signifi- 69 Male – cantly higher than in control subjects . Moreover, 83 Male – transgenic mice overexpressing S100B developed PD fea- 75 Male – tures, resulting in motor coordination impairment . Here, by using combined analysis of ex vivo PD-affected brain tissues and S100A9 co-aggregation with α-syn in sc-20173, 1 in 100), α-syn (mouse monoclonal, 3G282: vitro, we show that S100A9 indeed complements sc-69977, 1 in 100, raised against recombinant α-syn of α-syn amyloidaggregation andisinvolvedinthe human origin), and CD68 (mouse monoclonal, sc70761, 1 amyloid-neuroinflammatory cascade in PD pathology. in 100). In some control experiments, α-syn mouse mono- clonal, LB 509 (Novakemi AB), was also used, raised Methods against Lewy bodies purified from patients suffering de- Proteins mentia with Lewy bodies, and reactive with α-syn epi- S100A9 (13.2 kDa, 114 amino acid residues) and α-syn tope located in the region of amino acids 115-122. (14.4 kDa, 140 residues) were expressed and purified from Both α-syn antibodies showed consistent immuno- E. coli as described previously [30, 31]. Proteins were staining. Secondary antibodies from Vector Laborator- freeze-dried and used directly after dissolving them in ies were used: anti-mouse (MP-7402) and anti-rabbit phosphate buffered saline (PBS, 09-8912-100, Medicago). (MP-7401) IgG peroxidase reagent kits. 3,3′-Diamino- All samples were passed through a 0.22-μmfilter to elim- benzidine (DAB) and 3-amino-9-ethylcarbazole (AEC) inate spontaneously formed aggregates. N-labeled α-syn from Vector Laboratories were used as substrates pro- was purchased from AlexoTech AB. Protein concentra- ducing brown and reddish-brown staining of the sam- tions were determined by absorption at 280 nm with ex- ples, respectively. The tissues were scanned by a −1 −1 tinction coefficients of ε =0.53 (mg/mL) cm for Panoramic SCAN slide scanner 250 (3D Histech). −1 −1 S100A9 and 0.41 (mg/mL) cm for α-synuclein, Quantifications of S100A9-immunopositive cells in respectively. the substantia nigra and frontal lobe areas of five PD patients and four controls were conducted by selecting Tissue samples six random areas of 1 × 1 mm size in each tissue sec- The postmortem brain tissues from five PD patients and tion and calculating the mean value over all counts. four controls were examined (Table 1). Controls had no The percentage of co-localization of S100A9 and α-syn neurodegenerative diseases, in particular no Lewy bodies within cells in the substantia nigra was counted relative to or neurofibrillary tangles or anything else indicative of a α-syn-immunopositive cells. Lewy bodies were counted in neurodegenerative condition in the substantia nigra or the whole substantia nigra and frontal lobe regions of five frontal lobe as confirmed by the neuropathologist collab- PD patients, and co-localization of S100A9 and α-syn was orator at the Institute of Neurology, Medical University of calculated relative to α-syn-immunopositive inclusions. Vienna, Vienna, Austria, who provided the tissue samples for analysis. All tissue sections were from the frontal lobe Immunofluorescence and midbrain regions. They were paraffin-embedded and The brain tissues were cut into 5-μm-thick sections, depar- microtome-sectioned to 5-μm-thick slices. affinized in xylene, rehydrated in a graded series of alcohol, followed by antigen retrieval in citrate buffer, pH 6.0, and Immunohistochemistry washed in Tris-buffered saline Tween. The tissue sections Single and sequential immunohistochemistry with a series were blocked with 5% bovine serum albumin (BSA) in PBS of antibodies applied to the same tissue sections were per- for 30 min at 37 °C, and then incubated during 1 h with formed as described previously  with some modifica- primary antibodies diluted in 2.5% BSA. The following pri- tions . The following primary antibodies from Santa mary antibodies from Santa Cruz Biotechnology were used: Cruz Biotechnology were used: S100A9 (rabbit polyclonal, α-syn (mouse monoclonal, sc-69977, 1 in 200) and S100A9 Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 4 of 16 (rabbit polyclonal, sc-20173, 1 in 200). Then, the tissue intensity at the plateau level, t —midpoint of the growth samples were washed twice in PBS and incubated with phase (inflection point), τ—characteristic reaction time Alexa Fluor 555 goat anti-mouse IgG secondary antibodies equal to 1/K , where K is an apparent reaction rate app app (10 μg/mL, A28180, Thermo Fisher Scientific), interacting constant. Lag-phase time is defined as t =t −2τ. lag 0 with α-syn-specific primary antibodies, and observed by Fitting was conducted by using a Wolfram Mathematica orange fluorescence and with Alexa Fluor 488 goat 11 package. Each experimental amyloid formation kin- anti-rabbit IgG secondary antibodies (10 μg/mL, A-11034, etic curve is an average of five repeats, and each repeat Thermo Fisher Scientific), interacting with S100A9-specific was fitted by Eq. (1). The rates derived from all repeats primary antibodies, and observed by green fluorescence. were subjected to statistical analysis to determine if the The secondary antibodies were diluted in 2.5% mouse and difference between the group rates is statistically signifi- rabbit serum, respectively; they were applied for 30 min at cant. Level of statistical significance was set at 0.05. room temperature. Cell nuclei were stained with 4′,6-Dia- midine-2′-phenylindole dihydrochloride (DAPI, Roche). AFM Fluorescence microscopy was conducted on an Axio AFM imaging of the protein fibrils and tissue samples Imager A1 microscope (Carl Zeiss). was carried out by a BioScope Catalyst AFM (Bruker) in peak force mode in air, with resolution of 512 × 512 Congo red staining pixels. ScanAsyst air cantilevers (Bruker) were used. Pro- Deparaffinized and rehydrated tissue sections of 5 μm tein samples were diluted 50 times in deionized water thickness were stained with a saturated ethanol solution and incubated on the surface of freshly cleaved mica for of Congo red (Sigma) and sodium chloride and taken to 15 min, washed three times with 100 μl deionized water, pH 10.0 with 1% sodium hydroxide. Cell nuclei were and dried at room temperature. Instrumental set up in- stained with hematoxylin (Vector Laboratories). The cluding a BioScope Catalyst AFM combined with an stained samples were examined in an optical microscope inverted Nicon-Ti-S microscope equipped with tissue (Leica DM LB). slide holder was used for combining the immunohisto- chemical staining pattern with AFM imaging in analysis Amyloid formation kinetics of Lewy bodies in the substantia nigra brain tissues. To produce amyloid aggregates, S100A9 and α-syn pro- teins were incubated in PBS, pH 7.4 and 37 °C. The Lewy body dimensions amyloid formation kinetics was monitored as described Lewy body dimensions from a single patient were mea- previously . To monitor amyloid formation kinetics, sured by using AFM topographic imaging, which provides 20 μM thioflavin T (ThT) was initially added to native higher accuracy compared to immunohistochemical stain- protein samples. The samples were placed in 96-well ing. The measurements were performed in the AFM non-binding black plates (Corning), subjected to agita- cross-sections of topographic images as shown in Fig. 1i. tion at 300 rpm, and ThT fluorescence intensity was The estimates of outer and inner diameters of Lewy body monitored by a Tecan Infinite F200 plate reader. toroidal structures are shown by red lines, drawn through S100A9 and α-syn were cross-seeded with 5 and 10% the inflection points in their cross-sections, which were of pre-formed amyloids of their counter-parts (molar determined by first derivatives of the corresponding percentage). The α-syn/S100A9 amyloid cross-seeding cross-section profiles. experiments were performed at the above conditions under 200 rpm agitation and their ThT fluorescence was re- Scanning electron microscopy corded by a Fluostar Optima (BMG Labtech) plate reader. Scanning electron microscopy of the brain tissue sam- Fluorescence excitation wave length was set at 440 nm, and ples was performed by a Carl Zeiss Merlin field emission emission was registered at 480 nm every 10 min. Each scanning electron microscope using accelerating voltage sample was measured at least in eight replicates. of 4 kV. Prior to imaging process, the tissue slide was coated with carbon in a Quorum Q150T-ES specimen Amyloid kinetic fitting preparation station. Fitting of the amyloid kinetic curves was performed by using a sigmoidal function described in  Cellular toxicity SH-SY5Y neuroblastoma cells were cultured as de- max scribed previously . Cells were plated at a density of ItðÞ ¼ I þ ; ð1Þ min t−t ca. 100 cells/well in 96-well plates; the medium was 1 þ e changed after 24 h of incubation and before adding pro- where I(t) is normalized fluorescence intensity, I — tein samples. Initially, S100A9 and α-syn were incubated min fluorescence intensity at time 0, I —fluorescence at concentrations of 70 μM in PBS, pH 7.4, and 37 °C max Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 5 of 16 Fig. 1 Microscopy of α-syn Lewy bodies in the PD substantia nigra (an individual case). a–c Representative intracellular Lewy bodies immunostained with α-syn antibodies. Lewy bodies are shown in dark brown color and the host cells in lighter brown shade. Scale bars equal to 10 μm. d–f AFM height images of the corresponding Lewy bodies (from a–c). The surfaces of Lewy bodies and surrounding tissues are covered with DAB crystals used in immunohistochemical procedure to stain the tissue samples (shown in light color). Image sizes are 20 × 20 μm. g, h Scanning electron microscopy images of Lewy bodies shown in a, b. i AFM cross-section of Lewy body; its position is shown in f by white line. j, k Distribution of Lewy body mean outer and inner diameters, respectively, calculated by using BCa technique from AFM data. Mean diameters and their 95% CI are shown above the histograms. Probability density function (PDF) is shown along the y-axis. l Linear dependence between the inner and outer diameters of the Lewy bodies analyzed by AFM. Each point represents individual randomly selected Lewy body from the same patient and is shown in individual color during 10 and 60 h in order to produce the amyloid S100A9, α-syn, and their mixture were diluted in the samples populated with oligomers/protofilaments and fi- culture medium and added to SH-SY5Y cells at a final brils, respectively. The time points of sample collection concentration of 10 μM. Cell viability was measured by were selected in accord with the kinetics of amyloid for- WST-1 assay (Roche) after 24 h of co-incubation with mation and AFM imaging for all protein specimens. added protein samples. Absorbance at 450 nm was mea- Freshly dissolved and pre-incubated amyloid samples of sured in a plate reader (Tecan Infinite F200). Cell Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 6 of 16 viability was expressed as a percentage of the absorbance Results in wells containing control cells with added PBS. All ex- S100A9 and α-syn in Lewy bodies in the PD substantia periments were performed at least in triplicates, and nigra and frontal lobe regions each series was repeated three times. The amyloid sam- The tissue samples from five PD patients and four con- ples were incubated in sterile conditions prior to adding trol individuals (Table 1) were subjected to immunohis- to the cell culture media. tochemical analysis to examine the localization of α-syn and S100A9 antigens. Since Lewy body formation in the Circular dichroism (CD) substantia nigra is a hallmark of PD pathology , we Far UV CD spectra of both S100A9 and α-syn were re- have examined the prevalence of intracytoplasmic Lewy corded in 10 mM phosphate buffer, pH 7.4, and 20 °C bodies reactive with α-syn antibodies in the substantia with a Jasco J-710 spectropolarimeter using a 1-mm path nigra of five PD patients. A large number of Lewy bodies length quartz cuvette. distributed all over the substantia nigra were detected in all PD patients, and in one representative patient, they NMR spectroscopy were studied in more detail by combining immunohisto- A Bruker Avance 700 MHz NMR spectrometer equipped chemistry and AFM imaging. Lewy bodies were strongly with a triple-resonance cryoprobe was used to perform immunoreactive with α-syn antibodies as shown in the 1 15 NMR measurements. 2D H, N heteronuclear single representative images in Fig. 1a–c, displaying character- quantum coherence (HSQC) spectra of 77 μM istic pattern with a bright ring-shaped staining around isotope-labeled α-syn in 10 mM phosphate buffer, the pale central core. Most Lewy bodies were located pH 7.35, and 10 °C were recorded before and after within neuronal cells shown in lighter brown shade at addition of 2.5 mM S100A9. The spectra were refer- their background. Some neuronal cells contained two enced to the water signal, and the assignment of Lewy bodies (Fig. 1b, c), which is typical for PD path- α-syn amide cross-peaks was used from previous ology. This indicates that once the process of amyloid work . self-assembly has started within a cell, the developed amyloids can seed and propagate themselves. Surface plasmon resonance The topographic AFM images of the same Lewy bod- The interaction between α-syn and S100A9 were exam- ies in the substantia nigra tissues are shown in Fig. 1d–f, ined by using a Biacore X100 surface plasmon resonance the images were scanned by positioning the AFM canti- instrument (GE Healthcare). Monomeric α-syn was lever over the optical images of corresponding Lewy immobilized on a streptavidin-coated chip as described bodies. Since the Lewy bodies were initially localized previously . The binding was measured in HBS-P+ within the brain sections by immunostaining, the sur- buffer containing 10 mM HEPES, 150 mM NaCl, and faces of their sections were covered by DAB crystals 0.001% P20 detergent at pH 7.4 (GE Healthcare) and 25 ° used in immunohistochemical procedure. These surfaces C. Increasing concentrations of S100A9 were injected in a are higher and displayed in a light color in AFM images, single cycle sequence without regeneration steps between while the central parts not reactive with α-syn antibodies injections. At the end of the cycle, bound protein was are shown in darker color, respectively (Fig. 1d–f). It removed by injection of 50 mM NaOH. Experiments were was suggested that a granular core of Lewy bodies may repeated three times. Dissociation constant K was evalu- include a variety of nitrated, phosphorylated, and ubiqui- ated with the program provided by the instrument manu- tinated proteins surrounded by a filamentous halo con- facturer using the equation: R =(C*R )/(K + C), where taining α-syn amyloid fibrils . The same individual c max d R denotes binding level at concentration C and R is Lewy bodies were imaged also by using scanning elec- c max the extrapolated maximum binding capacity. tron microscopy as shown in Fig. 1g, h, where they dis- play the same morphology. Since the immunopositive Statistical analysis parts of Lewy bodies are visible as annuli, we measured The normality of all data sets was assessed by the their outer and inner diameters in the AFM Shapiro-Wilk test. Values of p ≤ 0.05 were considered cross-sections (Fig. 1f, i). By using corrected and acceler- statistically significant. The experimental data sets were ated bootstrap technique, we calculated the probability normally distributed and therefore analyzed by using density functions for means of both Lewy body diame- two-sample T test. These results are shown as mean ± ters and their respective 95% confidence intervals (Fig. 1j, standard deviation (SD). k). The mean value for outer diameters of all examined The means of outer and inner diameters of Lewy bod- Lewy bodies was 14.7 μm (CI 95% 13.0–16.7) and for ies (13 bodies) and their 95% confidence intervals (CI the inner diameters 7.9 μm (CI 95% 8.5–10.4), respect- 95%) were calculated by corrected and accelerated boot- ively. The dependence between the outer and inner di- strap (BCa) technique [38, 39]. ameters of Lewy bodies is linear with a slope of 0.99, Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 7 of 16 indicating that the thickness of the annuli is propor- fluorescence were noticed, indicating the deposition of tional to their diameters (Fig. 1l). The diameters of Lewy α-syn, S100A9, or their co-localization, characteristic for bodies were also measured by using scanning electron the initiation of amyloid pathology (Fig. 2i, j). The microscopy images (Fig. 1g, h), which resulted in the di- numbers of Lewy bodies in the PD frontal lobe areas mensions consistent with AFM measurements. were within a wider range of 80 to 240 per case (versus The substantia nigra tissue sections from five PD patients 300–350 per case in the substantia nigra) and up to 16 were also subjected to the sequential immunohistochemis- times lower per square millimeter of the tissue than in the try with pair of consecutively applied S100A9 and α-syn substantia nigra. In the frontal lobe, co-immunostaining antibodies, which revealed that some intracytoplasmic of the Lewy bodies with α-syn and S100A9 antibodies was Lewy bodies were clearly immunoreactive with both anti- observed in ca. 20% cases, similar to the substantia nigra. bodies as shown in two pairs of representative images (Fig. 2a–d). The host cells, containing these Lewy bodies, Neuronal S100A9 in the PD substantia nigra and frontal displayed typical neuronal morphology (Fig. 2a–d). Both lobe regions immunostaining patterns were overlapping, demonstrating The brain tissues of five PD patients compared to four the obvious co-localization of both S100A9 and α-syn controls were also examined for the intracellular presence within Lewy bodies. Particularly strong co-immunostaining, of both α-syn and S100A9 antigens, which may precede reflecting co-localization of these two antigens, was ob- the aggregation and Lewy body formation. Both proteins served at the outer layer region, shown as a bright ring are known to be expressed in neuronal cells, and the (Fig. 2c, d), though some Lewy bodies were more uniformly elevated levels of S100A9 have been reported in Alzhei- stained in the whole section (Fig. 2a, b). TheLewybodies mer’s disease and aging [18, 24, 26]; however, no informa- were also reactive with Congo red dye binding specifically tion is available for PD-affected brain tissues. Specifically, to amyloid inclusions as shown in Fig. 2e.The Lewy bodies the tissue sections from the PD-affected midbrain and in the substantia nigra were also observed by using frontal lobe areas have been analyzed, since both these immunofluorescence, i.e., intracellular Lewy body inclu- areas were reported to be dysfunctional in PD . sions reactive with α-syn antibodies were recognized by Numerous brightly stained S100A9-immunopositive cells orange fluorescence (Fig. 2f). Some small inclusions dis- with a characteristic neuronal morphology were observed played green fluorescence characteristic for S100A9-specific both in the substantia nigra and in the frontal lobe areas antibodies or yellow color, indicating the overlap of orange (Fig. 4a, e, i). Some of the neurons in the substantia nigra and green fluorescence and co-localization of both antigens contain also neuromelanin, which was reflected in their (Fig. 2f). The Lewy bodies immunoreactive with α-syn characteristic brown color compared to the antibodies were counted across all substantia nigra region antigen-specific reddish-brown staining (Fig. 4f, g). in all five PD patients and their amounts were within Sequential immunohistochemistry with antibodies to 300–350 per case. Lewy body inclusions immunoreactive S100A9, α-syn, and CD68 (specific for activated microglial with either α-syn or S100A9 antibodies were not observed cells and macrophages) was performed to examine if in the substantia nigra and frontal lobe tissues of control S100A9 and α-syn are produced also by microglial cells in individuals (Fig. 3). We have found that ca. 20% of the substantia nigra and two representative immunostain- α-syn-immunopositive Lewy bodies in the substantia nigra ing sequences are presented in Fig. 4a–d and Fig. 4e–h, were also immunopositive with S100A9 antibodies (Fig. 4l), respectively. The overlap of individual immunostaining indicating that S100A9 together with α-syn contributes to patterns for S100A9 and α-syn in the arbitrary colors Lewy body formation. In some neuronal cells in the indicates the co-localization of these proteins in neuronal substantia nigra, small granular structures immunopositive cells. The co-localization of the S100A9/α-syn and CD68 with both S100A9 and α-syn antibodies were also ob- immunostaining patterns in the substantia nigra was at served (as indicated by black arrows in Fig. 4e, g), suggest- the level of immunohistochemical detection, indicating ing initiation of PD pathology and potential Lewy body that S100A9 is accumulated primarily in neurons development. (Fig. 4d, h), though we cannot exclude its presence in Lewy body inclusions were found also in the frontal microglial cells. In the substantia nigra, α-syn, but not lobe areas of all five PD patients by using both immuno- S100A9, was also accumulated in Lewy neurites histochemistry and immunofluorescence (Fig. 2g–j). (Figs. 2b and 4c), which is a pathological hallmark of They were observed as round-shaped inclusions reactive PD. Substantial amount of S100A9-immunopositive with α-syn antibodies as shown by immunohistochemis- neurons were observed in the frontal lobe area of all try (Fig. 2g) or orange spheres as shown by immuno- five PD patients, as shown in representative image in fluorescence (Fig. 2j). Some inclusions were rather large Fig. 4i, and some neuronal cells were α-syn positive as but irregular in shape (Fig. 2h). In some neurons, the shown by immunofluorescence (Fig. 2i, j). The sequen- granular structures displaying orange, green, or yellow tial immunohistochemistry with S100A9 (Fig. 4i)and Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 8 of 16 Fig. 2 (See legend on next page.) Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 9 of 16 (See figure on previous page.) Fig. 2 Optical and fluorescence microscopy of Lewy bodies in the substantia nigra and frontal lobe regions of five PD patients. Representative sequential immunohistochemistry of Lewy bodies formed in neuronal cells in the PD substantia nigra tissues and conducted by using the pair of consecutive S100A9 (a, c) and α-syn (b, d) antibodies, respectively. e Congo red staining of intracellular Lewy body in the substantia nigra. f Immunofluorescence of the representative substantia nigra tissue with α-syn antibodies (shown in orange), S100A9 antibodies (in green), their superposition (in yellow), and DAPI staining of cell nuclei (in blue). The Lewy body in the center of image is shown as a well-defined round-shaped intracellular inclusion. The Lewy body in the lower right corner is shown in yellow indicating staining with both S100A9 and α-syn antibodies. g, h Immunohistochemistry of the representative frontal lobe tissues with α-syn antibodies. i, j Immunofluorescence of the representative frontal lobe tissues with α-syn and S100A9 antibodies and DAPI staining. Color representation as in f. In the images, the labelling notes for S100A9 and α-syn antibody staining is shown in green and orange, respectively. Scale bars are 50 μmin a–d, g,and h,20 μmin e, f, and 100 μmin i, j CD68 (Fig. 4j) antibodies, followed by the superposition interpret the occasional cytoplasmic staining as physio- of immunostaining patterns in the arbitrary colors logical expression of α-syn in neurons. This is clearly (Fig. 4k), indicated that there is no co-localization of different from the inclusion-body pathology (i.e., Lewy the staining patterns and therefore no detectable bodies) seen in the PD samples. S100A9 in microglial cells in the frontal lobe tissues. In both brain regions of all four control individuals, In the tissue samples from both the substantia nigra and the blood vessels were immunopositive for S100A9 anti- frontal lobe of all four control individuals, intra-neuronal bodies as shown in the representative staining in Fig. 3d, staining with α-syn antibodies was also observed shown in indicating that S100A9 is produced there by myeloid reddish-brown color (Fig. 3a, b), while brown coloring in cells. Some S100A9-immunopositive cells were also ob- the substantia nigra neurons corresponded to neuromela- served, but to a much lesser extent than in PD patient nin. However, the amount of cells immunoreactive with tissues (Fig. 3c, d). α-syn antibodies was much lower in the controls than in The amounts of S100A9-immunopositive cells were the PD patients. Since the α-syn antibody used in the counted in randomly selected areas of the substantia nigra study detects also the physiological form of α-syn and not and frontal lobe in each patient and averaged over all PD selective for the disease-associated conformation, we and control cases, respectively (Fig. 4l). Importantly, these ab cd Fig. 3 Immunohistochemical analysis of the brain tissues of four control cases. Representative immunostaining with α-syn (a, b)and S100A9 (c, d) antibodies of the substantia nigra (a, c) and frontal lobe (b, d) tissues, respectively. Neuromelanin is shown in brown color and immunostaining with corresponding antibody in reddish color. Scale bars are 100 μm Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 10 of 16 ab c d e fg h ij k l Fig. 4 Immunohistochemical analysis of neuronal S100A9 in the PD substantia nigra and frontal lobe tissues (five cases studied). a–c Sequential immunohistochemistry of the representative PD substantia nigra region with S100A9, CD68, and α-syn antibodies, respectively. d Superposition of the corresponding images from a–c shown in pseudo-colors: S100A9 immunostaining is shown in yellow, α-syn in blue, and CD68 in magenta. e–g Enlarged images of the sequential immunohistochemistry of the representative substantia nigra region collected from another PD patient and conducted by using S100A9, CD68, and α-syn antibodies, respectively. Black arrows mark the accumulation of α-syn and S100A9 in small granules. h Superposition of the corresponding images from e–g shown in pseudo-colors: color coding as in d. i, j Sequential immunohistochemistry of the representative PD frontal lobe tissues with S100A9 and CD68 antibodies, respectively. k Superposition of the corresponding images from i, j shown in pseudo-colors: color coding as in d. l Counts of S100A9-immunopositive neuronal cells in controls are shown by black bars and in PD patients by red bars; frontal lobe is labeled by FL and substantia nigra by SN. Percentage of the co-localization of S100A9 and α-syn within neuronal cells and Lewy bodies in the substantia nigra shown by blue bars. p ≤ 0.05 is indicated by *. Scale bars are 100 μmin a–d and 50 μmin e–k numbers significantly increased in PD patients compared ThT fluorescence assay, since ThT fluorescence increases to controls both in the substantia nigra and frontal lobe, when the dye binds specifically to amyloid structures i.e., by ca. sixfold. Sequential immunohistochemistry with (Fig. 5a). The S100A9 fibrillation was characterized by a S100A9 and α-syn antibodies revealed that 77% of the shorter lag phase compared to the α-syn amyloid assem- neuronal cells immunopositive with α-syn antibodies in bly, i.e., ca. 9 vs 16 h, respectively (Fig. 5a, b). The the substantia nigra were also immunopositive with growth phase of S100A9 fibrillation was slower than the S100A9 antibodies, indicating the co-localization of both α-syn amyloid growth with the midpoints at 36 vs 31 h, antigens (Fig. 4l). respectively (Fig. 5b). However, when both proteins were mixed at equimolar ratio, the lag phase of their joint Amyloid aggregation of α-syn and S100A9 amyloid aggregation became equal to that of S100A9 The amyloid formation kinetics of α-syn, S100A9, and alone, while the midpoint of the growth phase reduced their equimolar mixture, incubated under continuous to ca. 16 h (Fig. 5b). The described differences in the lag agitation in PBS, pH 7.4 and 37 °C, were monitored by times and midpoints for α-syn amyloid kinetics in the Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 11 of 16 presence and absence of S100A9 were statistically sig- AFM images in Fig. 5d–f. Amyloid fibrils of each protein nificant. In addition, the growth rate constant for α-syn and their mixture were collected after 60 h incubation, mixed with S100A9 was by ca. 40% faster than that of i.e., their time point of collection is indicated by red α-syn alone (Fig. 5c). The amyloid formation of α-syn arrow in the amyloid kinetic curves (Fig. 5a) and the was also observed in the presence of 5 and 10% S100A9 amyloid structures developed in each sample during this sample with preformed fibrils, characterized by smaller period of time are shown in the AFM images (Fig. 5g–i). shift of the lag phase, growth midpoint, and apparent rate The mature amyloid fibrils of α-syn, which are shown as constants to lower values compared to the fibrillation of micron length polymers in Fig. 5h, became significantly α-syn alone (Additional file 1). However, no deviations less toxic compared to all three oligomeric samples, re- from the original fibrillation curve of S100A9 (Fig. 5a) ducing cell viability to the level indistinguishable from were observed in the presence of 5 and 10% of α-syn pre- control (Fig. 5l). The amyloid fibrils of mixed proteins formed fibrils. These indicate that native S100A9, rather also became significantly less toxic than S100A9 oligo- than its fibrils, can potentiate α-syn amyloid formation. mers, reducing cell viability by ca. 30% compared to The amyloid formation of both proteins was moni- control. However, the change in cellular toxicity of short tored also by AFM imaging (Fig. 5d–i). After 10 h incu- amyloid fibrils of S100A9 was statistically insignificant bation, both S100A9 and α-syn developed round-shaped compared to those of all three oligomeric species (Fig. 5l, oligomers with ca. 3–4 nm and 5–6 nm heights in AFM Additional file 2). cross-sections, respectively (Fig. 5d, e). After the same incubation time of the equimolar mixture of both pro- Interactions of native S100A9 and α-syn teins, the elongated protofilaments were developed coex- The interactions of native S100A9 and α-syn were isting together with oligomers of ca. 3–10 nm heights analyzed by using solution NMR, CD spectroscopy, and 1 15 (Fig. 5f). After 60 h incubation, S100A9 self-assembled surface plasmon resonance. 2D H, N-HSQC spectra of into curly fibrils with ca. 4 nm heights in the AFM 77 μM α-syn were recorded before and after addition of cross-section (Fig. 5g, j), while α-syn formed mature S100A9 performed in titration steps with up to 2.5 mM fibrils of a few micron lengths and 8–10 nm heights in final concentration, to ensure the binding site saturation 1 15 AFM cross-sections (Fig. 5h, k). Interestingly, the (Fig. 6a). The 2D H, N-HSQC spectrum of free α-syn co-incubation for 60 h of their equimolar mixture resulted closely resembles previously assigned spectra of this pro- in the formation of large clustered aggregates coexisting tein under similar conditions , thus confirming that with smaller curly and round-shaped structures (Fig. 5i). it exists in an essentially unstructured conformation, characterized by transiently populated α-helices in the Cytotoxicity of S100A9 and α-syn amyloids N-terminal and central regions, but with the last 40 The cytotoxic effect on SH-SY5Y neuroblastoma cells residues of the C-terminus being more unfolded and produced by the native and amyloid species of S100A9, extended. Addition of S100A9 did not significantly α-syn, and their equimolar mixture was assessed by perturb the positions of α-syn HSQC amide cross-peaks using a WST 1 assay (Fig. 5l; statistical analysis is shown (Fig. 6a), but their intensities generally decreased to ca. in Additional file 2). Freshly dissolved S100A9 and its 80% compared to the original values. Interestingly, the mixture with α-syn added to SH-SY5Y cells caused decrease was more pronounced in the C-terminus, i.e., reduction of cell viability by ca. 23 and 18%, respectively, to ca. 70%, suggesting a possible binding site for S100A9 compared to control with added PBS, while freshly in the C-terminal part of monomeric α-syn (Fig. 6b). dissolved α-syn did not induce statistically significant The CD spectrum of α-syn in the far UV region also cytotoxic effect. The amyloid species of both α-syn, confirmed its largely unstructured conformation, display- S100A9, and their mixture formed after 10 h incubation ing the minimum of ellipticity at ca. 197 nm typical for appeared to be the most cell toxic, causing the reduction random coil (Fig. 6c). The CD spectrum of native S100A9 of cell viability by ca. 40, 70, and 45%, respectively; the is dominated by a signal corresponding to an α-helical difference between the cytotoxicity induced by α-syn conformation with characteristic strong minima at 220 and 1:1 mixed oligomeric samples was not statistically and 208 nm (Fig. 6c). Since the α-helical structure of significant. These samples were collected during the lag S100A9 is very stable, while the unfolded conformation of phase of the amyloid formation kinetics of α-syn and in α-syn can be easily perturbed upon interaction with the beginning of growth phase of the amyloid kinetics of S100A9, the differential CD spectrum of their equimolar S100A9 and 1:1 mixture as shown in Fig. 5a by green mixture was plotted, from which the contribution of the arrow. Therefore, the predominant amyloid species in S100A9 spectrum was subtracted. The ellipticity of the these samples were oligomers in the α-syn and S100A9 minimum at 197 nm did not change in this differential samples and oligomers together with short protofila- spectrum, but the amplitude of the peak with a minimum ments in the sample of mixed proteins as shown in the at 220 nm was increased, indicating that a certain change Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 12 of 16 bc de f gh i j l Fig. 5 Amyloid co-aggregation and cytotoxicity of S100A9 and α-syn. a Normalized kinetic curves of amyloid formation monitored by ThT fluorescence andfittedbysigmoidal function for 70 μM α-syn (in black), 70 μM S100A9 (blue), and both proteins taken at equimolar ratio (red). Experimental data points areshowningray. b Lag phase (dark bars) and midpoint of growth phase (light bars) of the amyloid formation kinetics derived from fitting. Protein samples are indicated under the x-axis and in the same color coding as in a. c Growth rate constant derived from fitting. Protein samples are indicated under the x-axis and in the same color coding as in a. Error bars represent SD. p ≤ 0.05 is indicated by *. d–i AFM height images of oligomeric/ protofilament species of d S100A9, e α-syn, and f their equimolar mixture formed after 10 h incubation; amyloid fibrils of g S100A9, h α-syn, and i their co-aggregates at equimolar ratio of both proteins formed after 60 h incubation. Scale bars equal to 500 nm in all images. j, k AFM cross-sections of amyloid fibrils of S100A9 and α-syn, as shown by white cross-sections in g, h, respectively. l Viability of SH-SY5Y cells measured by WST-1 assay after 24 h co-incubation with α-syn and S100A9 species. Viability of cells treated with PBS alone is shown by gray bar; cells treated with 10 μM α-syn, 10 μM S100A9, or their equimolar mixture are shown by black, blue, and red bars, respectively. The durations of amyloid sample aggregation prior adding to cell culture, i.e., 0, 10, and 60 h, are indicated under the x-axisbyblack,green,and red bars, respectively;the oligomeric and fibrillar sample time collections are marked by red and green arrows, respectively, in the amyloid formation kinetics in a. Error bars represent SD of at least nine measurements. All protein samples were incubated in PBS, pH 7.4 and 37 °C Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 13 of 16 a c b e 1 15 Fig. 6 NMR, CD, and surface plasmon resonance studies of α-syn and S100A9 interactions. a 2D H, N-HSQC spectra of 77 μM α-syn before (blue) and after (red) titration with up to 2.5 mM S100A9 in 10 mM phosphate buffer, pH 7.4 and at 10 °C. b Ratios of α-syn amide crosspeak intensities measured in the presence (I) and in the absence (I ) of S100A9, plotted as a function of α-syn amino acid residue numbers. c Far UV CD spectra of α- syn (black) and S100A9 (blue), 6 μM of each protein in 10 mM phosphate buffer, pH 7.4 and 20 °C. d Far UV CD spectrum of 6 μM α-syn alone before (black) and after addition of equimolar S100A9 (red), the signal of S100A9 was subtracted from the CD spectrum of the mixture. e Surface plasmon resonance response upon binding of injected S100A9 to surface-immobilized α-syn plotted as a function of S100A9 concentration in secondary structure is induced in α-syn upon S100A9 Discussion binding (Fig. 6d). Here, we have shown that S100A9 is abundant in the PD In order to quantify interactions between these two brain tissues both intracellularly and in Lewy body proteins, we have performed the titration of α-syn by amyloid deposits (Figs. 1, 2,and 4). Importantly, S100A9 increasing concentrations of S100A9 using surface plas- possesses dual properties as a pro-inflammatory mediator mon resonance technique. While α-syn was immobilized and amyloidogenic protein [18, 19]. Chronic neuroinflam- on the surface of the Biacore chip, increasing concentra- mation is one of the hallmarks of PD pathophysiology, as- tions of S100A9 were injected in a step-wise manner. sociated with increased pro-inflammatory mediator levels S100A9 binding to α-syn was detected by an increase in in the PD brain tissues [10–14]. Due to its inherent amy- the response signal, corresponding to the increased mass loidogenicity, S100A9 may play a more important role in of bound material. The dissociation constant for α-syn PD pathology than merely being one of the disease-related and S100A9 interaction calculated by fitting the titration pro-inflammatory mediators: it may contribute to the de- curve was ca. 5.0 ± 0.8 μM. velopment of the PD amyloid-neuroinflammatory cascade Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 14 of 16 in a similar way as shown previously in Alzheimer’sdis- ca. 5 μM as determined in a surface plasmon resonance ease . Immunohistochemical analysis showed that technique (Fig. 6e). The α-syn HSQC spectrum recorded S100A9 is present in 20% of all α-syn-immunopositive in the presence of S100A9 shows a decrease of cross-peak Lewy bodies in the PD substantia nigra and frontal lobe. intensities, particularly for residues corresponding to the Within those bodies, S100A9 is clearly co-localized with C-terminal region of α-syn, suggesting the location of the α-syn either within the same outer layer (Figs. 1 and 2)or S100A9 binding site (Fig. 6a, b). It has been shown distributed throughout the whole Lewy body section previously that transient interactions between the α-syn (Fig. 2). Lewy bodies were evenly spread throughout the C-terminal and N-terminal or central NAC regions are substantia nigra, which is typical for PD pathology. Lewy important in maintaining its natively unfolded structure, bodies were also observed in the frontal lobe areas of PD thereby preventing α-syn aggregation and fibrillation . patients (Braak stages 4–6), though they were in smaller Contacts between α-syn and S100A9, that perturb such quantities (Fig. 2g–j). Lewy bodies were not observed any- aggregation inhibiting self-interactions, could therefore where in the brain tissues of the controls (Fig. 3). enhance α-syn aggregation. AFM imaging of Lewy bodies in the substantia nigra Indeed, co-aggregation of α-syn and S100A9 occurs sig- enabled us to measure their dimensions more accurately. nificantly faster than amyloid formation of the individual Lewy bodies in the substantia nigra appeared to display proteins (Fig. 5a). Interestingly, native S100A9 is more very regular structures, characterized by rather tight prone to potentiate α-syn amyloid formation than its fi- margins for their outer and inner diameters and their an- brils. Previously, similar effect of native S100A6, which is nuli thicknesses proportional to the diameters (Fig. 1). This structurally homologous to S100A9, was observed on the signifies that Lewy bodies were formed via common mech- amyloid fibrillation of superoxide dismutase-1 (SOD-1) anisms. The fact that they have a dense core, as reported . In addition to the removal of the amyloid inhibiting previously , might be the underlying reason for the lack interactions within α-syn primary structure, the faster of immunostaining in their central area. The immunopo- nucleation process of S100A9, characterized by shorter lag sitive outer layer could be indicative of the lamellar phase than the lag phase of α-syn alone (Fig. 5a), may growth mechanism, which may be a diffusion-limited provide seeding effect on α-syn amyloid growth nuclei. In- process for the deposited protein and depend also on deed, the amyloid formation of S100A9 was well described the life span of the host cells, which provide the pro- by the Finke-Watzky model of nucleation-autocatalytic tein assembly environment. growth with dominant nucleation phase, while there was The immunohistochemical analysis revealed that S100A9 no evidence of the secondary nucleation pathways initiated and α-syn are present intracellularly both in the substantia on S100A9 fibrillar surfaces [18, 20], which may be the rea- nigra and frontal lobe areas of PD patients (Figs. 2 and 4). son for the small cross-seeding effect of its fibrils on α-syn S100A9 was found to be predominantly expressed in neur- fibrillation (i.e., inducing small but significant shortening of onal cells, which is similar to the previous observations in α-syn fibrillation lag phase and growth midpoint as shown the Alzheimer’s disease brain tissues . However, we in Additional file 2). At the same time, S100A9 fibrils are cannot exclude its presence in microglial cells. This indi- characterized by highly hydrophobic surfaces as shown cates that the pathogenic conditions associated with oxida- previously [18, 20] and therefore, they may sequestrate tive stress and amyloid self-assembly in amyloid diseases α-syn molecules from solution, effectively reducing its con- such as PD and Alzheimer’sdisease mayinduceexpression centration, which manifested in decreased α-syn fibrillation of S100A9 in neurons . Initially, S100A9 was discov- rate in their presence. ered in myeloid cells; however, its expression under cellular The co-aggregates of mixed S100A9 and α-syn, sampled stress conditions can be induced in various cell types at the oligomer and fibrillar stages, are significantly larger including epithelial cells, fibroblasts as well as neurons, in size than the corresponding amyloid structures of the indicating that S100A9 may exert its functions in a pleth- individual proteins (Fig. 5d–i), and this correlates with a ora of cell types [43–45]. In the substantia nigra, S100A9 reduction of S100A9 amyloid oligomer cytotoxicity and α-syn were found to be co-localized in up to 77% of (Fig. 5l). As the amyloid oligomers of S100A9 are more neuronal cells. At the same time, Lewy neurites were toxic than those of α-syn, S100A9 and α-syn immunoreactive only with α-syn, but not S100A9 anti- co-aggregation effectively rescues cells from S100A9 amyl- bodies (Fig. 2a, b), indicating that these proteins are not al- oid cytotoxicity. Interestingly, similar effect has been ob- ways present in the same cell regions. served previously upon the co-aggregation of S100A9 and Perturbation of the CD spectra of α-syn in the presence Aβ /Aβ , respectively, when interactions with the 1-40 1-42 of S100A9 indicated that S100A9 and α-syn interact with corresponding polypeptides also rescued S100A9 oligo- each other, thereby causing additional secondary structure mer cytotoxicity . Due to the abundance of formation of natively unfolded α-syn molecules (Fig. 6c, d). S100A9 in neuronal tissues, S100A9 oligomers may The apparent dissociation constant of their interactions is otherwise cause huge neuronal damage. This implies Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 15 of 16 also that the co-aggregation of amyloidogenic poly- Abbreviations AFM: Atomic force microscopy; CD: Circular dichroism; NMR: Nuclear peptides potentially could serve as a sink of danger- magnetic resonance;; PD: Parkinson’s disease; ThT: Tioflavin T; α-syn: ous amyloid species, removing them from active α-Synuclein circulation. Acknowledgements The abundance and inherent amyloidogenicity of We acknowledge Ranjeet Kumar at the Chalmers University of Technology for S100A9 raises the question if this protein can be a trig- providing us with α-syn used in the surface plasmon resonance experiments. ger of pathological amyloid cascade in PD. Indeed, We acknowledge the usage of Biochemical Imaging Center (BICU), Umeå University facilities, and Umeå Core Facility Electron Microscopy (UCEM) facilities S100A9 is present in the neuronal cells and Lewy bodies and technical assistance. in the substantia nigra, the region primarily affected by disease, and in the PD frontal lobe areas, which are af- Funding This study was funded by the ALF Västerbotten Läns Landsting (ALFVLL-369861 fected by the disease at its later stages (Figs. 1, 2, and 4). to L.A.M.-R.), Swedish Research Council (to A.G., 2014-3241 to L.A.M.-R.), FP-7 Moreover, S100A9 is present in neuronal cells and blood Marie Curie Action “Nano-Guard” (269138 to I.A.I., R.A.M. and L.A.M.-R.), vessels of aged patients without PD symptoms (Fig. 3). Insamlingsstiftelsen (FS 2.1.12–1605-14 to L.A.M.-R.), Brain Foundation (to A.G. and L.A.M.-R.), and Visby Programme, Swedish Institute (R.A.M.). We suggest here that sustained neuroinflammation pro- moting the spread of amyloidogenic S100A9 in the brain Availability of data and materials tissues can be a trigger of the amyloid cascade involving The datasets used and analyzed during the current study are available from the corresponding author on reasonable request. α-syn and leading to PD development , similar to the effect of S100A9 on Aβ aggregation in the Alzhei- Authors’ contributions mer’s disease . S100A9 could be a common denom- IH, IAI, RAM, and LAM-R designed the project. IH, IAI, RAM, ChW, SKTSW, and CeW performed the experiments. IH, IAI, RAM, SKTSW, CeW, AG, GGK, and inator in a broad variety of inflammation-dependent LAM-R analyzed the data. GGK collected the tissue samples. LAM-R wrote the amyloidoses, which may develop in various tissues and manuscript. IH, IAI, RAM, SKTSW, AG, GGK, and LAM-R reviewed and edited organs of the human body from the brain to the aging the manuscript. All authors read and approved the final manuscript. prostate, as we have shown previously [18, 48]. Ethics approval and consent to participate All experimental procedures with tissue samples were approved by the Medical Ethics Committees of the Institute of Neurology, Medical University Conclusions of Vienna, Austria, and the Umeå University Hospital, Sweden. This is the first report on the co-aggregation of α-syn and S100A9 both ex vivo and in vitro. Co-localized and Competing interests The authors declare that they have no competing interests. co-aggregated S100A9 and α-syn were found in Lewy bodies and neuronal cells in the PD substantia nigra Publisher’sNote and frontal lobe areas. In vitro, their co-aggregation Springer Nature remains neutral with regard to jurisdictional claims in occurred significantly faster than self-assembly of indi- published maps and institutional affiliations. vidual proteins, leading to the formation of larger Author details amyloid aggregates and mitigating the cytotoxicity of Department of Medical Biochemistry and Biophysics, Umeå University, S100A9 oligomers. The finding of S100A9 involvement 90187 Umeå, Sweden. Department of General Chemistry, Sumy State in PD may open a new avenue for therapeutic interven- University, Sumy 40007, Ukraine. Department of Pathology, Sumy State University, Sumy 40007, Ukraine. Department of Biochemistry and tions targeting S100A9 as a pro-inflammatory protein Biophysics, Stockholm University, 10691 Stockholm, Sweden. Institute of and preventing its amyloid self-assembly in the brain Neurology, Medical University of Vienna, Vienna, Austria. tissues. Received: 20 February 2018 Accepted: 20 May 2018 Additional files References 1. Breydo L, Wu JW, Uversky VN. alpha-Synuclein misfolding and Parkinson’s Additional file 1: Cross-seeding of α-syn amyloid formation by S100A9 disease. Biochim Biophys Acta. 2012;1822(2):261–85. amyloid fibrils. (A) Normalized kinetic curves of amyloid formation 2. Kim WS, Kagedal K, Halliday GM. Alpha-synuclein biology in Lewy body monitored by ThT fluorescence and fitted by sigmoidal function for diseases. Alzheimers Res Ther. 2014;6(5):73. 70 μM α-syn (shown in black), in the presence of 5% (magenta) and 3. Lashuel HA, Overk CR, Oueslati A, Masliah E. The many faces of alpha- 10% (cyan) of S100A9 fibrillar samples. Experimental data points are synuclein: from structure and toxicity to therapeutic target. Nat Rev shown in gray. (B) Lag phase (filled bars) and midpoint of growth phase Neurosci. 2013;14(1):38–48. (striped bars) of the amyloid formation kinetics derived from fitting. 4. Diao J, Burre J, Vivona S, Cipriano DJ, Sharma M, Kyoung M, Sudhof TC, Protein samples are indicated under x-axis and in the same color Brunger AT. Native alpha-synuclein induces clustering of synaptic-vesicle coding as in (A). (C) Growth rate constant derived from fitting. Protein mimics via binding to phospholipids and synaptobrevin-2/VAMP2. Elife. samples are indicated under x-axis and in the same color coding as in 2013;2:e00592. (A). Error bars represent SD. p ≤ 0.05 is indicated by *. (PDF 104 kb) 5. Spillantini MG, Schmidt ML, Lee VM, Trojanowski JQ, Jakes R, Goedert M. Additional file 2: Statistical analysis of the effects of S100A9, α-syn and Alpha-synuclein in Lewy bodies. Nature. 1997;388(6645):839–40. mixed specimens on SH-SY5Y cellular viability presented in Fig. 5l.Cell 6. Goedert M, Spillantini MG, Del Tredici K, Braak H. 100 years of Lewy viability values were compared pair-wise by using two-sample t-test, pathology. Nat Rev Neurol. 2013;9(1):13–24. n ≥ 9. NS denotes non-significant difference, * − p ≤ 0.05 and ** − p ≤ 0.01. 7. Jellinger KA. Significance of brain lesions in Parkinson disease dementia and (PDF 100 kb) Lewy body dementia. Front Neurol Neurosci. 2009;24:114–25. Horvath et al. Journal of Neuroinflammation (2018) 15:172 Page 16 of 16 8. Winner B, Jappelli R, Maji SK, Desplats PA, Boyer L, Aigner S, Hetzer C, Loher 31. Vogl T, Leukert N, Barczyk K, Strupat K, Roth J. Biophysical characterization of T, Vilar M, Campioni S, et al. In vivo demonstration that alpha-synuclein S100A8 and S100A9 in the absence and presence of bivalent cations. oligomers are toxic. Proc Natl Acad Sci U S A. 2011;108(10):4194–9. Biochim Biophys Acta. 2006;1763(11):1298–306. 9. Olanow CW, Brundin P. Parkinson’s disease and alpha synuclein: is 32. Glass G, Papin JA, Mandell JW. SIMPLE: a sequential immunoperoxidase Parkinson’s disease a prion-like disorder? Mov Disord. 2013;28(1):31–40. labeling and erasing method. J Histochem Cytochem. 2009;57(10):899–905. 33. Pirici D, Mogoanta L, Kumar-Singh S, Pirici I, Margaritescu C, Simionescu C, 10. Heneka MT, Kummer MP, Latz E. Innate immune activation in neurodegenerative Stanescu R. Antibody elution method for multiple immunohistochemistry disease. Nat Rev Immunol. 2014;14(7):463–77. on primary antibodies raised in the same species and of the same subtype. 11. Lim S, Chun Y, Lee JS, Lee SJ. Neuroinflammation in synucleinopathies. Brain J Histochem Cytochem. 2009;57(6):567–75. Pathol. 2016;26(3):404–9. 34. Uversky VN, Li J, Fink AL. Evidence for a partially folded intermediate in 12. Hirsch EC, Hunot S. Neuroinflammation in Parkinson’s disease: a target for alpha-synuclein fibril formation. J Biol Chem. 2001;276(14):10737–44. neuroprotection? Lancet Neurol. 2009;8(4):382–97. 35. Jia X, Gharibyan AL, Ohman A, Liu Y, Olofsson A, Morozova-Roche LA. 13. Ferrari CC, Tarelli R. Parkinson’s disease and systemic inflammation. Neuroprotective and nootropic drug Noopept rescues alpha-synuclein Parkinsons Dis. 2011;2011:436813. amyloid cytotoxicity. J Mol Biol. 2011;414(5):699–712. 14. Gao X, Chen H, Schwarzschild MA, Ascherio A. Use of ibuprofen and risk of 36. Eliezer D, Kutluay E, Bussell R Jr, Browne G. Conformational properties Parkinson disease. Neurology. 2011;76(10):863–9. of alpha-synuclein in its free and lipid-associated states. J Mol Biol. 15. Luo J, Warmlander SK, Graslund A, Abrahams JP. Cross-interactions 2001;307(4):1061–73. between the Alzheimer disease amyloid-beta peptide and other 37. Horvath I, Werner T, Kumar R, Wittung-Stafshede P. Copper chaperone amyloid proteins: a further aspect of the amyloid cascade hypothesis. J blocks amyloid formation via ternary complex. Q Rev Biophys. 2018;51 Biol Chem. 2016;291(32):16485–93. https://doi.org/10.1017/S0033583518000045. 16. Chemerovski-Glikman M, Rozentur-Shkop E, Richman M, Grupi A, Getler A, 38. DiCiccio TJ, Efron B. Bootstrap confidence intervals. Statist Sci. 1996; Cohen HY, Shaked H, Wallin C, Warmlander SK, Haas E, et al. Self-assembled 11(3):189–228. cyclic d,l-alpha-peptides as generic conformational inhibitors of the alpha- 39. Efron B, Tibshirani RJ. An introduction to the bootstrap. New York [etc.]: synuclein aggregation and toxicity: in vitro and mechanistic studies. Chapman & Hall; 1993. Chemistry. 2016;22(40):14236–46. 40. Wakabayashi K, Tanji K, Odagiri S, Miki Y, Mori F, Takahashi H. The Lewy 17. Richman M, Wilk S, Chemerovski M, Warmlander SK, Wahlstrom A, Graslund body in Parkinson’s disease and related neurodegenerative disorders. Mol A, Rahimipour S. In vitro and mechanistic studies of an antiamyloidogenic Neurobiol. 2013;47(2):495–508. self-assembled cyclic D,L-alpha-peptide architecture. J Am Chem Soc. 2013; 41. Ferrer I. Early involvement of the cerebral cortex in Parkinson’sdisease: 135(9):3474–84. convergence of multiple metabolic defects. Prog Neurobiol. 2009;88(2):89–103. 18. Wang C, Klechikov AG, Gharibyan AL, Warmlander SK, Jarvet J, Zhao L, Jia X, 42. Forno LS. Neuropathology of Parkinson’s disease. J Neuropathol Exp Neurol. Shankar SK, Olofsson A, Brannstrom T, et al. The role of pro-inflammatory 1996;55(3):259–72. S100A9 in Alzheimer’s disease amyloid-neuroinflammatory cascade. Acta 43. van Lent PL, Grevers L, Blom AB, Sloetjes A, Mort JS, Vogl T, Nacken W, van Neuropathol. 2014;127(4):507–22. den Berg WB, Roth J. Myeloid-related proteins S100A8/S100A9 regulate joint 19. Donato R, Cannon BR, Sorci G, Riuzzi F, Hsu K, Weber DJ, Geczy CL. inflammation and cartilage destruction during antigen-induced arthritis. Functions of S100 proteins. Curr Mol Med. 2013;13(1):24–57. Ann Rheum Dis. 2008;67(12):1750–8. 20. Iashchishyn IA, Sulskis D, Nguyen Ngoc M, Smirnovas V, Morozova-Roche 44. Averill MM, Kerkhoff C, Bornfeldt KE. S100A8 and S100A9 in cardiovascular LA. Finke-Watzky two-step nucleation-autocatalysis model of S100A9 biology and disease. Arterioscler Thromb Vasc Biol. 2012;32(2):223–9. amyloid formation: protein misfolding as “nucleation” event. ACS Chem 45. Goyette J, Geczy CL. Inflammation-associated S100 proteins: new Neurosci. 2017;8(10):2152–8. mechanisms that regulate function. Amino Acids. 2011;41(4):821–42. 21. Ha TY, Chang KA, Kim J, Kim HS, Kim S, Chong YH, Suh YH. S100a9 46. Hong DP, Xiong W, Chang JY, Jiang C. The role of the C-terminus of knockdown decreases the memory impairment and the neuropathology in human alpha-synuclein: intra-disulfide bonds between the C-terminus Tg2576 mice, AD animal model. PLoS One. 2010;5(1):e8840. and other regions stabilize non-fibrillar monomeric isomers. FEBS Lett. 22. Schluesener HJ, Kremsner PG, Meyermann R. Widespread expression of 2011;585(3):561–6. MRP8 and MRP14 in human cerebral malaria by microglial cells. Acta 47. Botelho HM, Leal SS, Cardoso I, Yanamandra K, Morozova-Roche LA, Fritz G, Neuropathol. 1998;96(6):575–80. Gomes CM. S100A6 amyloid fibril formation is calcium-modulated and 23. Postler E, Lehr A, Schluesener H, Meyermann R. Expression of the enhances superoxide dismutase-1 (SOD1) aggregation. J Biol Chem. 2012; S-100 proteins MRP-8 and -14 in ischemic brain lesions. Glia. 1997; 287(50):42233–42. 19(1):27–34. 48. Yanamandra K, Gruden MA, Casaite V, Meskys R, Forsgren L, Morozova- 24. Shepherd CE, Goyette J, Utter V, Rahimi F, Yang Z, Geczy CL, Halliday GM. Roche LA. α-Synuclein reactive antibodies as diagnostic biomarkers in blood Inflammatory S100A9 and S100A12 proteins in Alzheimer’s disease. sera of Parkinson’s disease patients. PLoS One. 2011;6(4):e18513. Neurobiol Aging. 2006;27(11):1554–63. 49. Braak H, Del Tredici K, Rub U, de Vos RA, Jansen Steur EN, Braak E. Staging 25. Engel S, Schluesener H, Mittelbronn M, Seid K, Adjodah D, Wehner HD, of brain pathology related to sporadic Parkinson’s disease. Neurobiol Aging. Meyermann R. Dynamics of microglial activation after human traumatic 2003;24(2):197–211. brain injury are revealed by delayed expression of macrophage-related proteins MRP8 and MRP14. Acta Neuropathol. 2000;100(3):313–22. 26. Swindell WR, Johnston A, Xing X, Little A, Robichaud P, Voorhees JJ, Fisher G, Gudjonsson JE. Robust shifts in S100a9 expression with aging: a novel mechanism for chronic inflammation. Sci Rep. 2013;3:1215. 27. Sathe K, Maetzler W, Lang JD, Mounsey RB, Fleckenstein C, Martin HL, Schulte C, Mustafa S, Synofzik M, Vukovic Z, et al. S100B is increased in Parkinson’s disease and ablation protects against MPTP-induced toxicity through the RAGE and TNF-alpha pathway. Brain. 2012;135(Pt 11):3336–47. 28. Wilhelm KR, Yanamandra K, Gruden MA, Zamotin V, Malisauskas M, Casaite V, Darinskas A, Forsgren L, Morozova-Roche LA. Immune reactivity towards insulin, its amyloid and protein S100B in blood sera of Parkinson’s disease patients. Eur J Neurol. 2007;14(3):327–34. 29. Liu J, Wang H, Zhang L, Xu Y, Deng W, Zhu H, Qin C. S100B transgenic mice develop features of Parkinson’s disease. Arch Med Res. 2011;42(1):1–7. 30. Sharma SK, Chorell E, Steneberg P, Vernersson-Lindahl E, Edlund H, Wittung- Stafshede P. Insulin-degrading enzyme prevents alpha-synuclein fibril formation in a nonproteolytical manner. Sci Rep. 2015;5:12531.
Journal of Neuroinflammation – Springer Journals
Published: Jun 4, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera