Neuroinflammation is increased in the parietal cortex of atypical Alzheimer’s disease

Neuroinflammation is increased in the parietal cortex of atypical Alzheimer’s disease Background: While most patients with Alzheimer’s disease (AD) present with memory complaints, 30% of patients with early disease onset present with non-amnestic symptoms. This atypical presentation is thought to be caused by a different spreading of neurofibrillary tangles (NFT) than originally proposed by Braak and Braak. Recent studies suggest a prominent role for neuroinflammation in the spreading of tau pathology. Methods: We aimed to explore whether an atypical spreading of pathology in AD is associated with an atypical distribution of neuroinflammation. Typical and atypical AD cases were selected based on both NFT distribution and amnestic or non-amnestic clinical presentation. Immunohistochemistry was performed on the temporal pole and superior parietal lobe of 10 typical and 9 atypical AD cases. The presence of amyloid-beta (N-terminal; IC16), pTau (AT8), reactive astrocytes (GFAP), microglia (Iba1, CD68, and HLA-DP/DQ/DR), and complement factors (C1q, C3d, C4b, and C5b-9) was quantified by image analysis. Differences in lobar distribution patterns of immunoreactivity were statistically assessed using a linear mixed model. Results: We found a temporal dominant distribution for amyloid-beta, GFAP, and Iba1 in both typical and atypical AD. Distribution of pTau, CD68, HLA-DP/DQ/DR, C3d, and C4b differed between AD variants. Typical AD cases showed a temporal dominant distribution of these markers, whereas atypical AD cases showed a parietal dominant distribution. Interestingly, when quantifying for the number of amyloid-beta plaques instead of stained surface area, atypical AD cases differed in distribution pattern from typical AD cases. Remarkably, plaque morphology and localization of neuroinflammation within the plaques was different between the two phenotypes. Conclusions: Our data show a different localization of neuroinflammatory markers and amyloid-beta plaques between AD phenotypes. In addition, these markers reflect the atypical distribution of tau pathology in atypical AD, suggesting that neuroinflammation might be a crucial link between amyloid-beta deposits, tau pathology, and clinical symptoms. Keywords: Atypical Alzheimer’s disease, Microglia, Complement, Human brain tissue, Neuroinflammation, Amyloid-beta plaque * Correspondence: b.boon@vumc.nl Department of Neurology, Alzheimer Center, Amsterdam Neuroscience, VU University Medical Center, Amsterdam, The Netherlands Department of Pathology, Amsterdam Neuroscience, VU University Medical Center, Amsterdam, The Netherlands © 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. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 2 of 16 Background Evidence for the correlation of pTau, neuroinflamma- Patients with Alzheimer’s disease (AD) typically tion, and microglia in AD subtypes is lacking. In this present with episodic memory impairment followed study, we aimed to explore whether an atypical spread- by deterioration of executive functioning, praxis, and ing of NFT pathology in non-amnestic AD is associated visuospatial skills. However, AD patients may also with an atypical distribution of neuroinflammation. In a present with an atypical phenotype [1, 2]. An atypical well-defined cohort of typical and atypical AD, we presentation is seen in 10% of the late-onset AD assessed and compared the presence of pTau, amyloid- (LOAD) patients (≥ 65 years of age) and up to 30% of beta, (activated) glial cells, and complement proteins in the early onset (< 65 years) AD (EOAD) patients [3]. temporal and parietal cortical areas. So far, three variants of atypical AD have been de- scribed: the posterior cortical atrophy (PCA) variant Methods characterized by visuoperceptual problems [4], the Post-mortem brain tissue logopenic variant characterized by aphasia [5], and Post-mortem brain tissue was obtained from the the frontal variant associated with behavioral changes Netherlands Brain Bank (NBB; Amsterdam, The [1, 2]. In addition to clinical differences, these differ- Netherlands). Donors signed informed consent for brain ent AD variants show syndrome-specific atrophy pat- autopsy, and the use of brain tissue and medical records terns on MRI [6]. for research purposes. Neuropathological diagnosis was AD is characterized by the deposition of amyloid-beta based on histochemical stainings including hematoxylin plaques and the formation of neurofibrillary tangles and eosin, congo red staining, Bodian or Gallyas and (NFT) in the brain. During disease progression, both methenamine silver stainings, and immunohistochemical plaques and NFTs are assumed to spread through the stainings for amyloid-beta, pTau, alpha-synuclein, and brain in a fixed order [7, 8]. However, the typical NFT p62. These stainings were performed on formalin-fixed distribution as originally described by Braak and Braak paraffin-embedded (FFPE) brain tissue of multiple brain [9] does not seem to hold for all AD cases. Clinicopatho- regions including the frontal cortex, temporal pole, logical studies indicated that AD patients with an atyp- superior parietal lobe, occipital pole, amygdala, and the ical phenotype have an atypical NFT distribution [9, 10]. hippocampus. Neuropathological diagnosis of AD was Furthermore, this atypical NFT distribution was demon- based on Braak stages for NFT and amyloid [7], Thal strated in living AD patients using the tau tracer phases for amyloid-beta [8], and CERAD criteria for [ F]AV1451 [11]. While the atypical distribution of NFTs neuritic plaques [26]. corresponds with the observed clinical phenotype, the cause of this difference in NFT spreading between AD Selection of typical and atypical AD cases variants remains elusive. Between 1996 and 2014, 352 AD cases came to autopsy There is accumulating evidence that inflammation and were semi-quantitatively scored by two neuropathol- plays a prominent role in the pathogenesis of AD. Re- ogists (WK, AR) for the NFT load using Bodian or cently, genome-wide association studies have identified Gallyas staining in the temporal pole, the frontal, super- several genes involved in inflammation, especially those ior parietal, and occipital cortex as previously described engaged in microglia function, as risk factors for devel- by Hoogendijk et al. [27]. The NFT load was scored in a oping AD [12–15]. The AD brain shows an increased 0.4-mm area as being absent (0), sparse (1), mild (2; 2 presence of activated microglia, reactive astrocytes, pro- to 3 NFTs), or severe (3; > 3 NFTs) for each brain region inflammatory cytokines, acute phase proteins, and acti- separately. From this cohort, we selected cases with an vated complement proteins compared to controls [16]. NFT score of 3 in either the temporal or parietal section, Complement proteins co-localize with NFTs [17, 18], as or in both sections, resulting in 296 cases (for flowchart, well as with amyloid-beta deposits [19], and are actively see Fig. 1). For 142 cases, the NFT score was higher in involved in the formation of these pathological struc- the temporal section than the parietal section. These tures. Clusters of activated microglia are found in amyl- cases were referred to as having a typical NFT oid plaques, and the presence of activated microglia distribution [7]. In 126 of 296 cases, an NFT score of 3 increases with disease severity [20, 21]. Recent disease wasfound in the temporal aswell as the parietal section. A models suggest that microglia are actively involved in higher NFT score in the parietal compared to the temporal the spreading of phosphorylated Tau (pTau) [22–24]. section was observed in 28 cases and was defined as a Tau pathology is heavily reduced in disease-modeled parietal dominant and thus atypical NFT distribution. mice that are depleted for microglia compared to To study the distribution of neuroinflammation in typ- their microglia-positive peers [23, 24]. In the human ical and atypical AD, we further refined our cohort to brain, the presence of activated microglia correlates include only cases with a concordance between clinical with Braak staging for NFTs [25]. presentation and NFT distribution. The clinical Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 3 of 16 Fig. 1 Flowchart of neuropathologically assessed and semi-quantitatively scored AD cohort. Between 1996 and 2014, 352 AD cases came to autopsy and were semi-quantitatively scored for NFTs as described by Hoogendijk et al. [27]. In 296 cases, an NFT score of ≥ 3 in the temporal and/or parietal cortex was observed. Typical NFT distribution was defined as a higher NFT score in the temporal compared to the parietal cortex. Atypical NFT distribution was defined as a higher NFT score in the parietal compared to the temporal section. From the cases with typical and atypical NFT distribution, 18 cases per group were selected for which the clinical phenotype was stratified as either amnestic or non-amnestic (results shown in Table 2). From these clinical phenotyped cases, 9 cases with an atypical NFT as well as non-amnestic clinical presentation were compared with 10 cases with an amnestic presentation and typical NFT distri- bution using immunohistochemistry. Mean ± SD is shown for age at death and disease duration in years. AD Alzheimer’s disease, NFT neurofibrillary tangle phenotype of 36 cases with atypical and typical NFT randomly chose 18 cases with both typical NFT distribu- pathology was retrospectively assessed. In 18 out of 28 tion and sufficient clinical information for clinical phe- cases with an atypical NFT distribution, the available notyping. Clinical assessment was performed clinical information was sufficient to come to a retro- retrospectively and independently by 2 cognitive neurol- spective clinical diagnosis (see Table 1 for demograph- ogists (YP and FB) using the NIA-AA criteria [2]. Both ics). To have an equal group for comparison, we clinicians were blinded to the pathological stratification Table 1 Demographics of 36 cases with typical and atypical NFT distribution for which extensive retrospective clinical assessment was performed Typical NFT distribution Atypical NFT distribution n =18 n =18 Phenotype Amnestic (n = 16) Non-amnestic (n = 2) Amnestic (n = 5) Non-amnestic (n = 13) Male, n (%) 6 (37) 0 3 (60) 8 (62) Age at death 82 (± 7) 88 (± 5) 71 (± 11) 67(± 7) Disease duration 8 (± 5) 7 (± 4) 11 (± 5) 8 (± 4) NFT stage [7] n per stage 4/5/6 2/10/4 1/1/0 0/2/3 0/7/6 Amyloid stage [7] n per stage O/A/B/C 0/0/16 0/0/2 0/0/5 0/1/12 Data are mean ± SD. Age at death and disease duration shown in years NFT neurofibrillary tangle Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 4 of 16 when assessing the clinical phenotype. The clinical Immunohistochemistry (IHC) stratification was based on (collateral) history and cogni- IHC was performed to detect pTau (AT8); amyloid-beta tive examination documented by the clinical neurologist. (N-terminal; IC16); reactive astrocytes (GFAP); microglia Cases with first-degree relatives affected by EOAD were (Iba1); activated microglia (CD68 and HLA-DP/DQ/DR); excluded to minimize the risk of genetic AD. Other ex- and complement proteins C1q, C3d, C4b, and C5b-9 clusion criteria were sepsis, other neurodegenerative or (Table 3). FFPE sections (5-μmthick) from the temporal psychiatric diseases, significant cerebrovascular disease, pole and superior parietal lobe of the right hemisphere post-mortem interval > 12 h, and prior known genetic were used. mutations. Due to our exclusion criteria and availability IHC for pTau, amyloid-beta, GFAP, and Iba1 was per- of archived brain tissue samples, our inclusion was lim- formed using the Ventana BenchMark ULTRA staining ited to 9 atypical AD cases and 10 typical AD cases of system (Roche, Basel, Switzerland). Tissue sections were which the temporal pole and superior parietal lobe were mounted on TOMO adhesive glass slides (Matsunami, assessed by immunohistochemistry (see Fig. 1 for inclu- Osaka, Japan) and deparaffinized. After blocking for sion flowchart, Table 2 for patient details, and Table 4 endogenous peroxidase, antigen retrieval was performed for demographics in the “Results” section). Cases were by heating sections at 100 °C in Cell Conditioning 1 not intentionally matched for disease duration, brain solution (pH 8.5) (Roche) for different durations per weight, ApoE status, or post-mortem interval. antibody (see Table 3). For detection of primary Table 2 Clinical and neuropathological characteristics of typical and atypical AD cases Case Phenotype Symptoms at clinical presentation Sex Age at Disease NFT Amyloid Brain Cause of death PMI ApoE death duration stage [7] stage [7] weight (grams) 1 Typical AD Memory F 92 8 5 C 933 Heart failure 7:00 43 2 Typical AD Memory, disorientation F 84 4 4 C 908 Cardiogenic 4:15 32 shock 3 Typical AD Memory, disorientation F 84 9 5 C 827 Cachexia 6:40 43 4 Typical AD Memory F 89 5 5 C 962 Pneumonia 6:28 43 5 Typical AD Memory F 83 6 6 C 1100 Dehydration 6:17 43 6 Typical AD Memory, behavior F 91 3 4 C 1026 Cachexia 6:25 33 7 Typical AD Memory, F 77 2 5 C 999 Pneumonia 6:05 33 8 Typical AD Memory, behavior M 70 2 6 C 1261 Metastasized 6:20 43 colon carcinoma 9 Typical AD Memory F 76 12 5 C 1223 Unknown 10:45 44 10 Typical AD Memory M 60 2 6 C 1191 Cachexia 6:15 43 11 Atypical AD Aphasia, dyscalculia, agraphia, F 65 6 6 C 975 Pneumonia 5:40 33 left-right agnosia, visuoconstruction problems 12 Atypical AD Aphasia, dyslexia, apraxia, M 65 2 6 C 1057 Cardiac 8:50 44 visuoconstruction problems insufficiency 13 Atypical AD Aphasia, acalculia, M 64 7 5 C 1135 Pneumonia 4:45 42 fingeragnosia, apraxia 14 Atypical AD Parkinsonism, falling, alien F 67 3 5 C 817 Epileptic insult 7:35 33 hand syndrome 15 Atypical AD Aphasia, apathy, agitation M 59 6 6 C 1300 Cachexia 5:05 44 16 Atypical AD Aphasia, dyscalculia, dyslexia, M 62 3 6 C 1116 Malign 4:15 43 disorientation neuroleptic syndrome 17 Atypical AD Aphasia, dyslexia, apathy, apraxia M 65 1 5 C 1150 Euthanasia 6:50 43 18 Atypical AD Aphasia, dyslexia, apraxia, M 62 6 5 B 1153 Cachexia 4:40 33 visuospatial problems, behavior 19 Atypical AD Visual hallucinations, psychosis M 61 6 6 C 1355 Pneumonia 5:00 43 Age at death and disease duration in years; post-mortem interval in hours to minutes. Typical AD defined as more neurofibrillary tangles in the temporal compared to the parietal cortex by semi-quantitative scoring as described by Hoogendijk et al. [27] and an amnestic presentation during life. Atypical AD defined as more neurofibrillary tangles assessed by semi-quantitative scoring in the parietal compared to the temporal cortex and a non-amnestic presentation AD Alzheimer’s disease, F female, M male, NFT neurofibrillary tangle, PMI post-mortem interval Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 5 of 16 Table 3 Characteristics of primary antibodies and staining details Antibody Antigen Species Origin details Dilution Incubation Antigen retrieval Detection time method pTau, clone AT8 Tau phosphorylated at Ser202 Mouse ThermoFisher, Pittsburgh, USA 1:10000 32 min Heat-induced (pH 8.5) Optiview and Thr205 IgG1 at 36 °C for 24 min Amyloid-beta, N-terminus of amyloid-beta Mouse Dr. Carsten Korth, University 1:25 32 min Heat-induced (pH 8.5) Optiview clone IC-16 (aa 1-16) IgG2a of Dusseldorf, Germany at 36 °C for 24 min GFAP, clone Glial fibrillary acidic protein Mouse Roche, Basel, Switzerland 1: 2 8 min Heat-induced (pH 8.5) Optiview EP672Y at 37 °C for 32 min Iba1 C-terminus of Iba1 Rabbit Wako Pure Chemical 1:4000 32 min Heat-induced (pH 8.5) Optiview Industries, Osaka, Japan at 36 °C for 16 min CD68, clone CD68 Mouse Dako, Glostrup, Denmark 1:1200 Overnight Heat-induced (pH 6.0) EnVision KP1 IgG1 at 4°C by autoclave HLA-DP/DQ/DR, Alpha and beta-chains of all Mouse Dako 1:800 Overnight Heat-induced (pH 6.0) EnVision clone CR3/43 products of the DP, DQ, and IgG1 at 4°C by autoclave DR subregions C1q C1q Rabbit Dako 1:25600 Overnight Heat-induced (pH 6.0) EnVision at 4°C by autoclave C3d C3d Rabbit Dako 1:3200 Overnight Heat-induced (pH 6.0) EnVision at 4°C by autoclave C4b C4b Rabbit Abcam, Cambridge, United 1:1600 Overnight Heat-induced (pH 6.0) EnVision Kingdom at 4°C by autoclave C5b-9, clone Neoepitope on C9 in the Mouse Hycult Biotech, Plymouth 1:400 Overnight Heat-induced (pH 6.0) EnVision WU13-15 membrane attack complex meeting, USA at 4°C by autoclave antibodies with 3,3′-diaminobenzidine tetrahydrochlo- perpendicular to the cortical surface of the cortex were ride (DAB), Optiview DAB IHC detection kit (Roche) photographed. Total surface, depending on the width of was used. Finally, the sections were mounted with Cov- the cortex, could vary for each ROI and contained at erslipping film (Sakura Tissue-Tek, Leiden, The least 2 columns. Images were taken using a × 10 object- Netherlands). ive on an Olympus BX 41 photomicroscope with a Leica IHC for CD68, HLA-DP/DQ/DR, C1q, C3d, C4b, and MC 170 HD digital camera. The presence of DAB stain- C5b-9 was performed manually. The sections were ing was quantified with ImageJ (NIH) using the color mounted on SuperFrost Plus glass slides (Menzel-Gläser, threshold plugin. Our outcome measurement was the Braunschweig, Germany) and deparaffinized. Subse- percentage of DAB-stained area per marker, also referred quently, the sections were blocked for endogenous per- to as immunoreactivity. In addition to immunoreactivity, oxidase using 0.3% hydrogen peroxide in phosphate we quantified the number of amyloid-beta and C4b pla- buffer saline (PBS; pH 7.4). The sections were immersed ques. For the amyloid-beta plaques, diffuse deposits were in sodium citrate buffer (10 mM sodium citrate, 5 M not taken into account and only dense plaques were NaOH, dH O, pH 6.0) and heated to 120 °C in an auto- quantified, defined as particles with an immunoreactive clave for antigen retrieval. Primary antibodies were surface area of 100 μm or more [29, 30]. C4b-positive diluted in normal antibody diluent (ImmunoLogic, Dui- deposits of the same surface area were quantified as a ven, The Netherlands) and incubated overnight at 4 °C. measurement of the atypical appearing plaques as de- Primary antibodies were detected using EnVision (Dako, scribed in the “Results” section. Glostrup, Denmark). Between steps, the sections were washed in PBS. Subsequently, antibodies were visualized Fluorescent triple stainings with DAB (Dako). After counterstaining with Co-localization of C4b and CD68 or HLA-DP/DQ/DR hematoxylin, the sections were mounted with Entellan with amyloid-beta and thioflavine S was visualized in the (Merck, Darmstadt, Germany). parietal section of 4 atypical AD cases and the temporal section of 2 typical AD cases. The typical AD cases Image analyses and quantitative assessment of served as positive controls and reference since immunostainings localization of complement and microglia in classical- For quantitative assessment, 2 regions of interest (ROI) cored plaques is widely described in literature (for CD68 were randomly selected within non-curved areas of each [20]/for complement [16, 19]). section containing all 6 cortical layers [28]. Within each After deparaffinization, the sections were submerged ROI, contiguous microscopic fields arranged in columns in sodium citrate buffer and heated to 120 °C in an Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 6 of 16 autoclave. Subsequently, the sections were incubated model. Statistical significance was set at p <.05 for compari- using different combinations of primary antibodies: son of baseline characteristics. mouse IgG2a-anti-amyloid-beta (1:200), rabbit-anti-C4b (1:200), and mouse IgG1-anti-HLA-DP/DQ/DR (1:25) or Results mouse IgG1-anti-CD68 (1:300). Antibodies were diluted Atypical AD cases are younger than typical AD cases in normal antibody diluent (ImmunoLogic) and incu- The post-mortem cohort used for IHC consisted of 9 bated overnight at 4 °C. Subsequently, the sections were atypical AD cases and 10 typical AD cases. For a sum- incubated with the following secondary antibodies: goat- mary of initial clinical symptoms at presentation for each anti-mouse IgG2a Alexa Fluor dye 594, goat-anti-mouse case, see Table 2. Most atypical AD cases presented with IgG1 Alexa Fluor dye 647, and donkey-anti-rabbit Alexa symptoms of aphasia, consisting of word-finding difficul- Fluor dye 647 (1:250 dilution, ThermoFisher, Waltham, ties and spelling mistakes, combined with apraxia. None USA). For visualization of amyloid structures, the sec- of our atypical cases retrospectively met the criteria for tions were counterstained with thioflavine S (1% in an isolated primary progressive aphasia [5]. One case dH O) and subsequently rinsed in 70% ethanol. Autoflu- presented with Parkinsonism and an alien hand syn- orescence was blocked with 0.1% Sudan black in 70% drome, fitting a corticobasal syndrome during life. All 10 ethanol for 5 min. Between steps, the sections were typical AD cases presented with memory complaints as rinsed with PBS. Finally, the sections were enclosed with most prominent initial symptom. Demographic character- 80% glycerol/20% tris buffered saline. Representative pic- istics of both AD phenotypes used for immunohistochem- tures were taken with a Leica DMi8 inverted fluorescent ical analysis are shown in Table 4. Similar to the large microscope equipped with a Leica DFC300 G camera. cohort of 296 AD subjects (see Fig. 1), atypical AD pa- tients selected for IHC analysis were younger at age of Statistical analysis death and more often male. The disease duration, brain Demographics of the typical and atypical AD groups were weight, post-mortem interval, disease severity, and ApoE compared using Fisher’s exact test for categorical and genotype did not differ between groups. Mann-Whitney U test for numerical and not normally dis- tributed data. Outcome measures were compared between Distribution of pTau and amyloid-beta in typical and the 2 AD groups by using linear mixed model analysis. Lin- atypical AD ear mixed model analysis was used to adjust for the nested Immunohistochemistry for pTau showed neuronal inclu- observations within cases. In the linear mixed model ana- sions as well as neuritic threads (Fig. 2a). Typical AD lyses, the group variable (typical versus atypical AD), the re- cases showed more pTau immunoreactivity in the tem- gion (temporal versus parietal), and the interaction between poral compared to the parietal section (Fig. 2c). This group and region were added. Correcting for age and sex was contrary to the pTau distribution in atypical AD made the model less stable and was therefore not per- cases, in which the parietal section showed more formed. An assumption to apply a linear mixed model is that residuals of outcome measurements are normally dis- tributed. To meet this assumption, all outcome variables Table 4 Demographic characteristics of the AD cases used for (pTau, amyloid-beta, GFAP, Iba1, HLA-DP/DQ/DR, CD68, immunohistochemical analysis C1q, C3d, C4b, number of amyloid-beta plaques, and num- Typical AD Atypical AD p-value ber of C4b plaques) were transformed by taking the natural (n = 10) (n =9) log of the (variable + 1). The covariance structure was set Male, n 27<.05 to unstructured. Using the linear mixed model, we an- Age at death 81 (± 10) 63 (± 3) < .01 swered if the difference in outcome measurement over the Disease duration 5 (± 3) 4 (± 2) .78 2 regions was different between the 2 AD phenotypes (re- Brain weight (grams) 1043 (± 146) 1117 (± 161) .32 gion × phenotype), also referred to as interaction effect. Both phenotypes showed a similar distribution over the 2 PMI (h:min) 6:19 (± 1:48) 5:51 (± 1:33) .66 regions if no interaction effect was found. Statistical analysis NFT stage [7] 2/5/3 0/4/5 .46 n per stage 4/5/6 was performed in IBM SPSS statistics version 22.0 (IBM SPSS Statistics, Armonk, NY, USA). Bonferroni correction Amyloid stage [7] 0/0/10 0/1/8 .47 n per stage O/A/B/C was used to correct for multiple testing. Since we tested 11 outcome measurements (amyloid-beta, pTau, GFAP, Iba1, ApoE genotype 1/2/0/6/1 0/3/1/3/2 .48 n per category CD68, HLA-DP/DQ/DR, C1q, C3d, C4b, # amyloid-beta 32/33/42/43/44 plaques, and # C4b plaques), statistical significance was set Data in mean (± SD). Age at death and disease duration in years. Mann-Whitney at p < .0045 (p < .05/11 outcome measurements) for each U test for continuous data. Fisher’s exact test for categorical data effect (region × phenotype and region) of the linear mixed AD Alzheimer’s disease, NFT neurofibrillary tangle, PMI post-mortem interval Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 7 of 16 Fig. 2 pTau and amyloid-beta distribution in typical AD and atypical AD. a In typical AD, the temporal cortex (blue border) shows more immunoreactivity for pTau than the parietal cortex (burgundy border). This distribution is inversed in atypical AD (boxplot in c). b Although both typical and atypical AD show more overall amyloid-beta immunoreactivity in the temporal cortex compared to the parietal region, the atypical AD group shows increased number of amyloid-beta plaques in the parietal compared to temporal section (boxplot in e). Bar represents 100 μm. c, d,and e Boxplots showing pTau immunore- active area (%), amyloid-beta immunoreactive area (%), and the number of amyloid-beta plaques, respectively, in the temporal and parietal section ofboth AD phenotypes. Data shown as median (bar), 1st and 3rd quartile (box boundaries), and min to max (error bars). A difference in distribution over the two regions between the two AD phenotypes is indicated by # (Table 5), * p < .0045 immunoreactivity compared to the temporal section. In the atypical AD group. Besides a difference in plaque addition, the distribution of pTau over the 2 regions dif- number, we observed a contrast in plaque morphology fered significantly between the 2 phenotypes (Table 5). between the two groups, which will be addressed below. Amyloid-beta immunoreactivity was present in the form of diffuse deposits, dense plaques, and classical Glial activation is increased in atypical AD cored plaques (Fig. 2b). Both the typical and atypical AD Staining for GFAP-positive astrocytes showed variably group showed more immunoreactivity for amyloid-beta sized star-like GFAP-positive structures in all AD cases in the temporal than parietal section (Fig. 2d), and no (Fig. 3a). Both phenotypes showed higher levels of difference in distribution was observed (Table 5). While immunoreactivity for GFAP in the temporal section total immunoreactivity levels did not show differences compared to the parietal section. Atypical AD cases between the 2 phenotypes, the number of dense plaques showed relatively more GFAP immunoreactivity in the with an immunoreactive surface area of 100 μm or parietal cortex compared to typical AD cases (Fig. 3c; more did show a significant difference in distribution Table 5). over the 2 regions between the 2 phenotypes (Fig. 2e) Iba1 immunostaining showed positivity in both the (Table 5). In contrast to the typical AD group, more cell soma as well as the processes of the microglia, dense plaques were observed in the parietal section of mostly in the form of ramified microglia (Fig. 3b). Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 8 of 16 Table 5 Results for the linear mixed model of transformed immunohistochemistry variables Transformed variable Fixed Effect Phenotype Results for linear mixed model Beta-coefficient 95% CI pTau Region x phenotype 1.53* [1.33 – 1.72] Region Typical AD - 0.82* [- 0.95 – - 0.68] Atypical AD 0.71* [0.57 – 0.85] Amyloid-beta Region x phenotype 0.04 [- 0.06 – 0.15] Region Both - 0.46* [- 0.54 – - 0.39] GFAP Region x phenotype 1.25* [1.02 – 1.48] Region Typical AD - 1.71* [- 1.87 – - 1.56] Atypical AD - 0.46* [- 0.63 – - 0.29] Iba1 Region x phenotype 0.10 [0.002 – 0.19] Region Both - 0.47* [- 0.53 – - 0.41] CD68 Region x phenotype - 0.43* [0.38 – 0.48] Region Typical AD - 0.13* [- 0.17 – - 0.10] Atypical AD 0.29* [0.26 – 0.33] HLA-DP/DQ/DR Region x phenotype 0.97* [0.88 – 1.06] Region Typical AD - 0.08* [- 0.14 – - 0.03] Atypical AD 0.89* [0.82 – 0.96] C1q Region x phenotype - 0.05 [- 0.15 – 0.06] Region Both 0.03 [- 0.05 – 0.10] C3d Region x phenotype 0.74* [0.62 – 0.86] Region Typical AD - 0.39* [ - 0.47 – - 0.31] Atypical AD 0.36* [0.26 – 0.45] C4b Region x phenotype 0.95* [0.84 – 1.07] Region Typical AD - 0.24* [- 0.32 – - 0.17] Atypical AD 0.71* [0.62 – 0.80] # of amyloid-beta plaques Region x phenotype 0.58* [0.41 – 0.75] Region Typical AD - 0.33* [- 0.44 – - 0.21] Atypical AD 0.25* [0.13 – 0.37] # of C4b plaques Region x phenotype 1.84* [1.58 – 2.11] Region Typical AD - 0.39* [- 0.56 – - 0.22] Atypical AD 1.46* [1.25 – 1.66] Results of the linear mixed model for analyzed immunohistochemistry variables are shown. All variables were transformed: ln(variable + 1). We tested if the distribution over the 2 regions was different between the 2 AD phenotypes, defined as the interaction effect: region × phenotype. When an interaction effect was found, the beta-coefficient is shown per phenotype. To correct for multiple testing, a p value < .0045 was considered significant (p < .05/11 outcome measurements) and indicated with * AD Alzheimer’s disease, CI confidence interval # Number Both AD groups showed a similar temporal domi- CD68 and HLA-DP/DQ/DR in the parietal compared to nancy for Iba1. the temporal section, which was in contrast to the typ- Activated microglia were stained using CD68 and ical AD cases. In addition, the levels of CD68 and HLA- HLA-DP/DQ/DR. Whereas CD68 positivity was mostly DP/DQ/DR immunoreactivity were relatively high in the found in the soma of microglia, HLA-DP/DQ/DR parietal section of atypical AD compared to the temporal showed a prominent staining in the processes of micro- section of typical AD. glia (Fig. 4a, b). Both markers showed a different distribution over the Increased presence of complement proteins in atypical AD 2 regions between the 2 AD phenotypes (Fig. 4c, d;Table 5). To visualize different parts of the complement cascade, we Atypical AD cases showed more immunoreactivity for stained for C1q, C3d, and C4b, representing the start of the Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 9 of 16 Fig. 3 GFAP and Iba1 immunoreactivity is temporal dominant in both AD phenotypes. a, b The temporal cortex (blue border) shows more GFAP and Iba1 immunoreactivity respectively than the parietal cortex (burgundy border) in typical and atypical AD. Bar represents 100 μm. c, d Boxplot showing GFAP and Iba1 immunoreactive area (%) in the temporal and parietal section, respectively, of both AD phenotypes. Atypical AD shows relatively more GFAP immunoreactivity in the parietal cortex than typical AD (Table 5). Iba1 distribution is not different between the two phenotypes. Data shown as median (bar), 1st and 3rd quartile (box boundaries), and min to max (error bars). *p <.0045 cascade, as well as for C5b-9, defining the end-stage of the the parietal cortex (Fig. 5e, f; Table 5). In contrast, atyp- cascade and forming the membrane attack complex. ical AD showed higher immunoreactivity levels for both C1q immunoreactivity was observed as a weak diffuse markers in the parietal section compared to the tem- staining in the form of plaque-like structures, as punctu- poral section. ate staining of the neuropil, and sometimes in neurons Analysis of C5b-9 exposed very low to no immunore- (Fig. 5a). While the plaque-like structures were more activity in both regions of both phenotypes (data not often observed in the parietal section, the punctate stain- shown). For this reason, no quantification or statistical ing of the neuropil was more prominent in the temporal analysis was performed for C5b-9. The few structures section of both AD groups. No difference in C1q immu- that were C5b-9 positive included parenchymal and noreactivity was observed between regions or AD phe- meningeal vessels. The serum within these vessels also notypes (Fig. 5d; Table 5). stained positive for C5b-9. Compared to C1q, staining for C3d and C4b showed Our data show a different distribution in temporal an intense plaque-like staining, which had a morphology and parietal regions between both AD phenotypes for resembling that of compact and classical cored plaques complement factors C3d and C4b. Both complement (Fig. 5b, c). Levels of immunoreactivity for C3d and C4b factors were mostly abundant in the parietal section in typical AD were higher in the temporal compared to of atypical AD. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 10 of 16 Fig. 4 CD68 and HLA-DP/DQ/DR immunoreactivity show a parietal dominant distribution in atypical AD. a, b In typical AD, the temporal section (blue border) shows more CD68 and HLA-DP/DQ/DR immunoreactivity than the parietal section (burgundy border). In contrast, atypical AD shows more CD68 and HLA-DP/DQ/DR immunoreactivity in the parietal than the temporal section. Note the relative difference in immunoreactivity in the parietal section of atypical AD compared to the temporal section in typical AD. Bar represents 100 μm. c, d Boxplot showing CD68 and HLA- DP/DQ/DR immunoreactive area (%) in the temporal and parietal section, respectively, of both AD phenotypes. Linear mixed model shows a different distribution over the 2 regions between phenotypes (#) (Table 5). *p < .0045 Different plaque appearance in typical and atypical AD This staining pattern was compared with that of Looking at the morphology of plaques stained by classical-cored plaques observed in typical AD. In amyloid-beta and C4b, we observed a difference in classical-cored plaques, C4b, amyloid-beta, and thiofla- appearance between the 2 AD phenotypes. Amyloid-beta vine S co-localized in the core of the plaque, while the and C4b immunoreactive plaques in the parietal section corona only stained for amyloid-beta. Compared to of atypical AD cases showed a more granular compos- cored plaques, coarse-grained plaques showed a fibrillar, ition compared to deposits in the temporal section of less organized morphology with co-localization of C4b, this phenotype or compared to deposits in both regions amyloid-beta, and thioflavine S all over the plaque sur- of the typical AD cases (Fig. 6). The surface area positive face. Since an increased presence of activated microglia for C4b of these coarse-grained plaques was larger (> in atypical AD was observed, the localization of CD68 100 μm ) than that of typical plaques. Atypical AD cases and HLA-DP/DQ/DR with cored and coarse-grained had more of these C4b plaques in the parietal compared plaques was compared (Fig. 7). Like cored plaques, to the temporal cortex, which was contrary to typical coarse-grained plaques were associated with clusters of AD cases (Fig. 6i; Table 5). The coarse-grained plaques CD68 and HLA-DP/DQ/DR-positive microglia. In cored observed in the parietal cortex of atypical AD triple- plaques, activated microglia were located between the stained for C4b, amyloid-beta, and thioflavine S (Fig. 7). core and corona of the plaque. In coarse-grained Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 11 of 16 Fig. 5 Distribution of C3d and C4b differs between AD phenotypes. a In both typical and atypical AD, C1q deposition in the temporal cortex (blue border) is more diffuse than in the parietal cortex (burgundy border). b, c In typical AD, the temporal cortex shows more immunoreactivity for C3d and C4b than the parietal cortex. This distribution is inverted in atypical AD, showing more immunoreactivity for C3d and C4b in the parietal than temporal section. Bar represents 100 μm. d, e, and f Boxplots of immunoreactive area (%) for C1q, C3d, and C4b, respectively, in the temporal and parietal section of both AD phenotypes. Linear mixed model shows a different distribution for C3d and C4b over the 2 regions between phenotypes (#) (Table 5). *p < .0045 plaques, the localization of CD68 and HLA-DP/DQ/DR- we observed that also the occurrence of pTau is differ- positive microglia was less structured and positive ently distributed in typical and atypical AD. Here, we microglia appeared throughout the plaque. This data show for the first time that the distribution of both acti- supports a morphological difference between cored and vated microglia and complement factors between the coarse-grained plaques, of which the latter occurs prom- temporal and parietal lobe differentiates atypical from inently in atypical AD (Fig. 6i). typical AD cases. In addition, atypical AD cases are characterized by the presence of plaques with an abnor- Discussion mal morphology, highlighted by an alternative Atypical AD is characterized by a different distribution localization of microglia and presence of complement of NFTs when compared with typical AD. As expected, proteins. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 12 of 16 Fig. 6 Plaques in atypical AD show a different morphology compared to plaques in typical AD. a–d In typical AD cases, the morphology of amyloid-beta (a for overview, b for detail) and C4b (c for overview, d for detail) deposits come in the form of diffuse, dense, or classical cored plaques. e–h In atypical AD cases, amyloid-beta (e for overview, f for detail) and C4b (g for overview, h for detail) deposits show a distinct morphology, being coarse-grained and affecting a larger surface area (> 100 μm ) compared to classical cored plaques. Pictures are taken in the region with the highest number of amyloid-beta plaques, being temporal for typical AD and parietal for atypical AD (Fig. 2e). Bars represent 100 μm. i Boxplot ofnumberofcoarse-grained plaques quantified using C4b staining in the temporal and parietal cortex of typical and atypical AD. Linear mixed model shows a different distribution for coarse-grained plaques over the 2 regions between phenotypes (#) (Table 5). This distribution is parietally dominant in atypical AD. *p < .0045 In this study, typical and atypical AD were defined from this study do not necessarily apply for patients with according to the distribution of NFTs as well as their a logopenic or behavioral phenotype, as tau pathology clinical presentation. In our cohort, atypical AD cases seems to be differently distributed in those cases [11, were younger and more often male than typical AD 33]. Besides neuropathological studies, also clinical stud- cases. This was also seen in an earlier study by Murray ies report that an atypical clinical presentation is more and colleagues [31], indicating that an atypical distribu- common at a younger age [3]. Regarding differences in tion of pathology seems to be more common in men gender, an atypical presentation is not per se more com- and at a younger age. In line with this, another study mon in men [3]. However, AD presenting at late-onset is showed that cases with EOAD showed higher mean more common in women [34], explaining the relative levels of NFTs in the parietal lobe, when comparing high number of female subjects in typical AD cohorts. EOAD to LOAD, irrespective of clinical presentation We did not observe a different distribution of total [32]. When taking symptomology into account, this par- amyloid-beta immunoreactivity between 2 AD pheno- ietal dominant tau distribution is especially common in types. This is in line with other studies reporting that the PCA variant of atypical AD [11]. Therefore, the amyloid-beta distribution is not different between AD current atypical AD cohort defined by parietal dominant subtypes [9, 32, 35–37]. However, when quantifying for NFTs represents only a subgroup of atypical AD. Results number of dense amyloid-beta plaques, atypical AD cases Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 13 of 16 Fig. 7 Different plaque morphology in the parietal cortex of atypical AD. First row: in typical AD, classical cored plaques in the temporal cortex show an organized staining pattern with a corona showing merely amyloid-beta positivity versus a core positive for thioflavine S, amyloid-beta, and C4b. Second row: in atypical AD, fibrillar plaques in the parietal cortex show co-localization of thioflavine S, amyloid-beta, and C4b in the form of fibrils throughout the whole plaque. Third row: in typical AD, CD68-positive microglia are localized between the core and corona of classical cored plaques. Fourth row: in atypical AD, CD68-positive microglia localization is less organized. Fifth + sixth row: this different distribution within plaques between the two phenotypes also holds for HLA-DP/DQ/DR-positive microglia. Bar is applicable to all images and represents 100 μm showed a parietal dominant distribution compared to typ- In the present study, we observed a clear morpho- ical AD cases, which showed a temporal dominant distri- logical difference between classical-cored plaques in bution. This distinction in number of plaques was also typical AD and dense plaques in atypical AD cases. reported by Hoff and colleagues who compared cases with Dense plaques in atypical AD cases have a coarse- PCA to typical AD cases using stereology [38]. These re- grained structure, as observed with amyloid-beta sults indicate that although the distribution of overall immunostaining and thioflavine S staining. In addition, amyloid-beta immunoreactivity is similar, the number of these coarse-grained plaques showed a strong immuno- dense amyloid-beta plaques may differ between AD phe- reactivity for complement. Multiple studies have notypes. In addition to the number of dense amyloid-beta reported complement proteins to be associated with plaques, structural differences in amyloid-beta plaques amyloid-beta deposits [19, 39]. However, a difference in might be associated with the pathological and clinical dif- complement activation between AD subtypes has so far ferences between typical and atypical AD. not been described. The increased presence of Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 14 of 16 complement in coarse-grained plaques in atypical AD neuroinflammatory response [25]. These studies sug- cases supports a difference in amyloid structure be- gest that sex and age influence microglia activation. tween typical and atypical AD. Most likely, the amyloid However, whether differences in sex and age contrib- structure of fibrillar plaques favors a strong binding ute to differences in regional distribution, as observed and activation of complement factors, which in turn in the current study between AD subtypes, remain could act as opsonins for phagocytosis carried out by elusive. microglia [17, 40, 41]. A structural variation in amyloid-beta fibrils in combination with a difference Conclusions in binding of amyloid associated proteins may under- The results of this study are in line with the assump- lie the observed difference in the occurrence of neu- tion that the fibrillar structure and protein compos- roinflammation, pTau, and NFTs between typical and ition of different plaques may be relevant for the atypical AD. The relation between amyloid-beta, com- difference in regional vulnerability among AD pheno- plement, and microglia is underlined by a study in types. In addition, our results support a role for acti- APP transgenic mice deficient for C3 showing less vated microglia and complement factors in the cognitive problems and more amyloid-beta plaques atypical spreading of pathology in AD subtypes. In compared to APP mice not deficient for C3 [42]. The this study, we focused at 2 brain regions in a subset amyloid-beta plaques in the C3 knockout mice of atypical AD. It would be interesting to expand this showed less microglial co-localization. Together, these study on the role of neuroinflammation to various findings suggest that structural differences in amyloid brain regions in a larger atypical AD cohort. More deposits in atypical AD may directly be related to clinicopathological studies of AD variants (e.g., logo- complement and microglial activation. penic and behavioral frontal) are needed to better The distribution of NFT and pTau pathology is clearly understand the relation between amyloid-beta, pTau, associated with the presence of activated microglia in and related pathological mechanisms. Future research AD variants. Recent animal studies have shown that should focus on how variable disease mechanisms microglia are capable of both internalizing [43] and underlie the regional susceptibility in different brain excreting pTau [24, 44], suggesting that microglia con- regions leading to different clinical AD subtypes. tribute to the spreading of the pathology. Indeed, when mice are depleted for microglia, spreading of tau path- Abbreviations AD: Alzheimer’s disease; DAB: 3,3′-Diaminobenzidine tetrahydrochloride; ology is significantly reduced [23, 24]. In addition, acti- EOAD: Early-onset Alzheimer’s disease (< 65 years); FFPE: Formalin-fixed vated microglia could also contribute or induce tau paraffin-embedded; GFAP: Glial fibrillary acidic protein; hyperphosphorylation in neurons [45]. These studies in- IHC: Immunohistochemistry; IQR: Inter-quartile range; LOAD: Late-onset Alzheimer’s disease (≥ 65 years); NBB: Netherlands Brain Bank; dicate that microglial activation drives the spreading of NFT: Neurofibrillary tangle; PBS: Phosphate buffer saline; PCA: Posterior pathology and stimulates neurofibrillary degeneration. cortical atrophy; PMI: Post-mortem interval; pTau: Phosphorylated Tau; Interestingly, recent evidence from a pathological study ROI: Region of interest suggests that activation of microglia may precede tau Acknowledgements pathology in chronic traumatic encephalopathy [46], We would like to thank all brain donors and their caregivers for brain donation, which implicates that tau pathology may be a conse- the Netherlands Brain Bank and Michiel Kooreman for logistics and help in quence rather than a cause for microglial activation. selecting brain tissue samples, Martijn Heijmans and Wiesje van der Flier for advice on statistical analysis, Robert Veerhuis for help and advice on complement The aforementioned demographical differences in immunostainings, and Yolande Pijnenburg for clinical assessment. age and sex may influence the neuroinflammatory re- sponse. Former studies have shown various results on Funding the correlation between sex, age, and microglial acti- This study was funded by ZonMw grant number 70-73305-98-106. ZonMw had no role in the design of the study, collection, analysis, or interpretation vation in both humans and animal models. Schwarz of the data, or writing of the manuscript. et al. showed that during early development, male rats have more microglia within the parietal cortex Availability of data and materials compared to female rats. However, during juvenile The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. and early adulthood, this balance switches, indicating that sex hormones influence microglial colonization at Authors’ contributions different time points in rats [47]. In human studies, FB, JJMH, and BDCB designed the study. BDCB and JJMH coordinated the contradictory results are published for the effect of study and were responsible for writing the manuscript. BDCB, BL, and KNE age. In healthy controls, aging is correlated with a performed the experiments. BDCB analyzed the data. FB participated in writing the manuscript. PS made intellectual contribution and participated in more primed microglial state. Nevertheless, this was discussions. AJMR, WK, and the NBB were responsible for the autopsy shown to be different in diseased cases, in which material and neuropathological evaluation. All authors read and approved increasing age is associated with a diminished the final manuscript. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 15 of 16 Ethics approval and consent to participate NIH Public Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ Brain donors signed informed consent for autopsy and the use of tissue and 21460841. Cited 4 Oct 2017. medical records for research purposes. This study was approved by the 14. Hu X, Pickering E, Liu YC, Hall S, Fournier H, Katz E, et al. Meta-analysis for ethical committee of the VUmc. genome-wide association study identifies multiple variants at the BIN1 locus associated with late-onset Alzheimer’s disease. PLoS One. 2011;6:e16616. Competing interests Public Library of Science. Available from: http://www.ncbi.nlm.nih.gov/ The authors declare that they have no competing interests. pubmed/21390209. Cited 4 Oct 2017. 15. Sims R, van der Lee SJ, Naj AC, Bellenguez C, Badarinarayan N, Jakobsdottir J, et al. Rare coding variants in PLCG2, ABI3, and TREM2 implicate Publisher’sNote microglial-mediated innate immunity in Alzheimer’s disease. Nat Genet. Springer Nature remains neutral with regard to jurisdictional claims in 2017;49:1373–84. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ published maps and institutional affiliations. 28714976. Cited 4 Oct 2017. 16. Akiyama H, Barger S, Barnum S, Bradt B, Bauer J, Cole GM, et al. Received: 19 February 2018 Accepted: 26 April 2018 Inflammation and Alzheimer’s disease. Neurobiol Aging. 2000;21:383–421. NIH Public Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ 10858586. Cited 4 Oct 2017. References 17. Shen Y, Yang L, Li R. What does complement do in Alzheimer’s disease? Old 1. Dubois B, Feldman HH, Jacova C, Hampel H, Molinuevo JL, Blennow K, et al. molecules with new insights. Transl Neurodegener. 2013;2:21. BioMed Advancing research diagnostic criteria for Alzheimer’s disease: the IWG-2 Central. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24119446. criteria. Lancet Neurol. 2014;13:614–29. Available from: http://linkinghub. Cited 14 Apr 2017. elsevier.com/retrieve/pii/S1474442214700900. Cited 9 Aug 2017. 18. McGeer PL, Akiyama H, Itagaki S, EG MG. Activation of the classical 2. McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR, Kawas CH, complement pathway in brain tissue of Alzheimer patients. Neurosci Lett. et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations 1989;107:341–6. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ from the National Institute on Aging-Alzheimer’s Association workgroups 2559373. Cited 10 Aug 2017. on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 19. Eikelenboom P, Stam FC. Immunoglobulins and complement factors in 2011;7:263–9. NIH Public Access. Available from: senile plaques an immunoperoxidase study. Acta Neuropathol. 1982;57: http://www.ncbi.nlm.nih.gov/pubmed/21514250. Cited 9 Dec 2016. 239–42. Available from: https://link.springer.com/content/pdf/10.1007/ 3. Koedam ELGE, Lauffer V, van der Vlies AE, van der Flier WM, Scheltens P, BF00685397.pdf. Cited 11 Apr 2017. Pijnenburg YAL. Early-versus late-onset Alzheimer’s disease: more than age 20. Arends YM, Duyckaerts C, Rozemuller JM, Eikelenboom P, Hauw JJ. alone. J Alzheimers Dis. 2010;19:1401–8. Available from: http://www.ncbi. Microglia, amyloid and dementia in Alzheimer disease. A correlative study. nlm.nih.gov/pubmed/20061618. Cited 9 Dec 2016. Neurobiol Aging [Internet]. 21:39–47. Available from: http://www.ncbi.nlm. 4. Crutch SJ, Lehmann M, Schott JM, Rabinovici GD, Rossor MN, Fox NC. nih.gov/pubmed/10794847. Cited 9 Dec 2016. Posterior cortical atrophy. Lancet Neurol [Internet]. 2012;11:170–8. Available 21. Dickson DW, Farlo J, Davies P, Crystal H, Fuld P, Yen SH. Alzheimer’s disease. from: http://linkinghub.elsevier.com/retrieve/pii/S1474442211702897. Cited 9 A double-labeling immunohistochemical study of senile plaques. Am J Aug 2017. Pathol. 1988;132:86–101. Available from: http://www.ncbi.nlm.nih.gov/ 5. Gorno-Tempini ML, Hillis AE, Weintraub S, Kertesz A, Mendez M, Cappa SF, pubmed/2456021. Cited 9 Dec 2016. et al. Classification of primary progressive aphasia and its variants. 22. Li Y, Liu L, Barger SW, Griffin WST. Interleukin-1 mediates pathological Neurology. 2011;76:1006–14. American Academy of Neurology. Available effects of microglia on tau phosphorylation and on synaptophysin from: http://www.ncbi.nlm.nih.gov/pubmed/21325651. Cited 9 Aug 2017. synthesis in cortical neurons through a p38-MAPK pathway. J Neurosci. 6. Ossenkoppele R, Cohn-Sheehy BI, La Joie R, Vogel JW, Möller C, Lehmann 2003;23:1605–11. NIH Public Access. Available from: http://www.ncbi.nlm. M, et al. Atrophy patterns in early clinical stages across distinct phenotypes nih.gov/pubmed/12629164. Cited 10 Aug 2017. of Alzheimer’s disease. Hum Brain Mapp. 2015;36:4421–37. NIH Public 23. Maphis N, Xu G, Kokiko-Cochran ON, Jiang S, Cardona A, Ransohoff RM, Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26260856. et al. Reactive microglia drive tau pathology and contribute to the Cited 29 Dec 2017. spreading of pathological tau in the brain. Brain. 2015;138:1738–55. Oxford 7. Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. University Press. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ Acta Neuropathol. 1991;82:239–59. Available from: http://www.ncbi.nlm.nih. 25833819. Cited 9 Dec 2016. gov/pubmed/1759558. 24. Asai H, Ikezu S, Tsunoda S, Medalla M, Luebke J, Haydar T, et al. Depletion 8. Thal DR, Rüb U, Orantes M, Braak H. Phases of a beta-deposition in the of microglia and inhibition of exosome synthesis halt tau propagation. human brain and its relevance for the development of AD. Neurology. 2002; Nat Neurosci. 2015;18:1584–93. NIH Public Access. Available from: 58:1791–800. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ http://www.ncbi.nlm.nih.gov/pubmed/26436904. Cited 9 Dec 2016. 12084879. Cited 9 Dec 2016. 25. Hoozemans JJ, Rozemuller AJ, van Haastert ES, Eikelenboom P, van Gool 9. Tang-Wai DF, Graff-Radford NR, Boeve BF, Dickson DW, Parisi JE, Crook R, et WA. Neuroinflammation in Alzheimer’s disease wanes with age. al. Clinical, genetic, and neuropathologic characteristics of posterior cortical J Neuroinflammation. 2011;8:171. Available from: http://jneuroinflammation. atrophy. Neurology. 2004;63:1168–74. Available from: biomedcentral.com/articles/10.1186/1742-2094-8-171. Cited 9 Dec 2016. http://www.ncbi.nlm.nih.gov/pubmed/15477533. Cited 9 Dec 2016. 26. Mirra SS, Heyman A, McKeel D, Sumi SM, Crain BJ, Brownlee LM, et al. The 10. Galton CJ, Patterson K, Xuereb JH, Hodges JR. Atypical and typical Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part II. presentations of Alzheimer’s disease: a clinical, neuropsychological, Standardization of the neuropathologic assessment of Alzheimer’s disease. neuroimaging and pathological study of 13 cases. Brain. 2000;123 Pt Neurology. 1991;41:479–86. Available from: http://www.ncbi.nlm.nih.gov/ 3:484–98. Available from: http://www.ncbi.nlm.nih.gov/pubmed/10686172. pubmed/2011243. Cited 12 Dec 2016. Cited 9 Dec 2016. 27. Hoogendijk WJ, Sommer IE, Pool CW, Kamphorst W, Hofman MA, 11. Ossenkoppele R, Schonhaut DR, Schöll M, Lockhart SN, Ayakta N, Baker SL, Eikelenboom P, et al. Lack of association between depression and loss of et al. Tau PET patterns mirror clinical and neuroanatomical variability in neurons in the locus coeruleus in Alzheimer disease. Arch Gen Psychiatry. Alzheimer’s disease. Brain. 2016;139:1551–67. Available from: 1999;56:45–51. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ http://www.ncbi.nlm.nih.gov/pubmed/26962052. Cited 9 Dec 2016. 9892255. Cited 30 Aug 2017. 12. Bertram L, Lange C, Mullin K, Parkinson M, Hsiao M, Hogan MF, et al. 28. Arendt T, Morawski M, Gärtner U, Fröhlich N, Schulze F, Wohmann N, et al. Genome-wide association analysis reveals putative Alzheimer’s disease Inhomogeneous distribution of Alzheimer pathology along the isocortical susceptibility loci in addition to APOE. Am J Hum Genet. 2008;83:623–32. relief. Are cortical convolutions an Achilles heel of evolution? Brain Pathol Elsevier. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18976728. [Internet]. 2016; Available from: http://doi.wiley.com/10.1111/bpa.12442. Cited 4 Oct 2017. Cited 9 Dec 2016. 13. Naj AC, Jun G, Beecham GW, Wang L-S, Vardarajan BN, Buros J, et al. 29. Walker L, McAleese KE, Johnson M, Khundakar AA, Erskine D, Thomas AJ, et al. Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are Quantitative neuropathology: an update on automated methodologies and associated with late-onset Alzheimer’s disease. Nat Genet. 2011;43:436–41. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 16 of 16 implications for large scale cohorts. J Neural Transm. 2017;124:671–83. Available servlet/aliasResolver?alias=iospress&doi=10.3233/JAD-150704. Cited 11 Apr from: http://www.ncbi.nlm.nih.gov/pubmed/28265813. Cited 30 Aug 2017. 2017. 30. Duyckaerts C, Delatour B, Potier M-C. Classification and basic pathology of 45. Bhaskar K, Konerth M, Kokiko-Cochran ON, Cardona A, Ransohoff RM, Lamb Alzheimer disease. Acta Neuropathol. 2009;118:5–36. Available from: BT. Regulation of tau pathology by the microglial fractalkine receptor. http://www.ncbi.nlm.nih.gov/pubmed/19381658. Cited 30 Aug 2017. Neuron. 2010;68:19–31. NIH Public Access. Available from: http://www.ncbi. nlm.nih.gov/pubmed/20920788. Cited 20 Dec 2017. 31. Murray ME, Graff-Radford NR, Ross OA, Petersen RC, Duara R, Dickson DW. 46. Cherry JD, Tripodis Y, Alvarez VE, Huber B, Kiernan PT, Daneshvar DH, et al. Neuropathologically defined subtypes of Alzheimer’s disease with distinct clinical Microglial neuroinflammation contributes to tau accumulation in chronic characteristics: a retrospective study. Lancet Neurol. 2011;10:785–96. Available traumatic encephalopathy. Acta Neuropathol Commun. 2016;4:112. from: http://www.ncbi.nlm.nih.gov/pubmed/21802369. Cited 9 Dec 2016. Available from: http://actaneurocomms.biomedcentral.com/articles/10.1186/ 32. Bigio EH, Hynan LS, Sontag E, Satumtira S, White CL. Synapse loss is greater s40478-016-0382-8. Cited 30 Aug 2017. in presenile than senile onset Alzheimer disease: implications for the 47. Schwarz JM, Sholar PW, Bilbo SD. Sex differences in microglial colonization cognitive reserve hypothesis. Neuropathol Appl Neurobiol. 2002;28:218–27. of the developing rat brain. J Neurochem. 2012;120:948–63. NIH Public Blackwell Science Ltd. available from: http://doi.wiley.com/10.1046/j.1365- Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22182318. 2990.2002.00385.x. Cited 26 Oct 2017. Cited 27 Mar 2018. 33. Josephs KA, Dickson DW, Murray ME, Senjem ML, Parisi JE, Petersen RC, et al. Quantitative neurofibrillary tangle density and brain volumetric MRI analyses in Alzheimer’s disease presenting as logopenic progressive aphasia. Brain Lang. 2013;127:127–34. NIH Public Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23541297. Cited 27 Mar 2018. 34. Snyder HM, Asthana S, Bain L, Brinton R, Craft S, Dubal DB, et al. Sex biology contributions to vulnerability to Alzheimer’s disease: a think tank convened by the Women’s Alzheimer’s Research Initiative. Alzheimers Dement. 2016; 12:1186–96. Available from: http://linkinghub.elsevier.com/retrieve/pii/ S1552526016328746. Cited 14 Apr 2017. 35. Beaufils E, Ribeiro MJ, Vierron E, Vercouillie J, Dufour-Rainfray D, Cottier J-P, et al. The pattern of brain amyloid load in posterior cortical atrophy using (18)F-AV45: is amyloid the principal actor in the disease? Dement Geriatr Cogn Dis Extra. 2014;4:431–41. Karger Publishers. Available from: http://www.ncbi.nlm.nih.gov/pubmed/25538727. Cited 12 Apr 2017. 36. Wang X-D, Lu H, Shi Z, Cai L, Liu S, Liu S, et al. A pilot study on clinical and neuroimaging characteristics of Chinese posterior cortical atrophy: comparison with typical Alzheimer’s disease. PLoS One. 2015;10:e0134956. Public Library of Science, Available from: http://www.ncbi.nlm.nih.gov/ pubmed/26267071. Cited 15 Dec 2016. 37. de Souza LC, Corlier F, Habert M-O, Uspenskaya O, Maroy R, Lamari F, et al. Similar amyloid- burden in posterior cortical atrophy and Alzheimer’s disease. Brain. 2011;134:2036–43. Available from: https://www.ncbi.nlm.nih. gov/pubmed/?term=de+Souza+LC%2C+Corlier+F%2C+Habert+M-O%2C. Cited 2016 Dec 9. 38. Hof PR, Vogt BA, Bouras C, Morrison JH. Atypical form of Alzheimer’s disease with prominent posterior cortical atrophy: a review of lesion distribution and circuit disconnection in cortical visual pathways. Vis Res. 1997;37:3609– 25. Available from: http://www.sciencedirect.com.vu-nl.idm.oclc.org/science/ article/pii/S0042698996002404. Cited 12 Apr 2017. 39. Zanjani H, Finch CE, Kemper C, Atkinson J, McKeel D, Morris JC, et al. Complement activation in very early Alzheimer disease. Alzheimer Dis Assoc Disord [Internet]. 19:55–66. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/15942322. Cited 14 Apr 2017. 40. Wyss-Coray T, Yan F, Lin AH-T, Lambris JD, Alexander JJ, Quigg RJ, et al. Prominent neurodegeneration and increased plaque formation in complement-inhibited Alzheimer’s mice. Proc Natl Acad Sci U S A. 2002;99: 10837–42. Available from: http://www.pnas.org/cgi/doi/10.1073/pnas. 162350199. Cited 14 Apr 2017. 41. Maier M, Peng Y, Jiang L, Seabrook TJ, Carroll MC, Lemere CA. Complement C3 deficiency leads to accelerated amyloid beta plaque deposition and neurodegeneration and modulation of the microglia/macrophage phenotype in amyloid precursor protein transgenic mice. J Neurosci. 2008; 28:6333–41. NIH Public Access. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/18562603. Cited 14 Apr 2017. 42. Shi Q, Chowdhury S, Ma R, Le KX, Hong S, Caldarone BJ, et al. Complement C3 deficiency protects against neurodegeneration in aged plaque-rich APP/ PS1 mice. Sci Transl Med. 2017;9:eaaf6295. Available from: http://www.ncbi. nlm.nih.gov/pubmed/28566429. Cited 30 Aug 2017. 43. Luo W, Liu W, Hu X, Hanna M, Caravaca A, Paul SM. Microglial internalization and degradation of pathological tau is enhanced by an anti- tau monoclonal antibody. Sci Rep. 2015;5:11161. Available from: http://www.nature.com/articles/srep11161. Cited 30 Aug 2017. 44. Bolós M, Llorens-Martín M, Jurado-Arjona J, Hernández F, Rábano A, Avila J. Direct evidence of internalization of tau by microglia in vitro and in vivo. J Alzheimers Dis. 2015;50:77–87. Available from: http://www.medra.org/ http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Neuroinflammation Springer Journals

Neuroinflammation is increased in the parietal cortex of atypical Alzheimer’s disease

Free
16 pages

Loading next page...
 
/lp/springer_journal/neuroinflammation-is-increased-in-the-parietal-cortex-of-atypical-GXMg0XvF0c
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s).
Subject
Biomedicine; Neurosciences; Neurology; Neurobiology; Immunology
eISSN
1742-2094
D.O.I.
10.1186/s12974-018-1180-y
Publisher site
See Article on Publisher Site

Abstract

Background: While most patients with Alzheimer’s disease (AD) present with memory complaints, 30% of patients with early disease onset present with non-amnestic symptoms. This atypical presentation is thought to be caused by a different spreading of neurofibrillary tangles (NFT) than originally proposed by Braak and Braak. Recent studies suggest a prominent role for neuroinflammation in the spreading of tau pathology. Methods: We aimed to explore whether an atypical spreading of pathology in AD is associated with an atypical distribution of neuroinflammation. Typical and atypical AD cases were selected based on both NFT distribution and amnestic or non-amnestic clinical presentation. Immunohistochemistry was performed on the temporal pole and superior parietal lobe of 10 typical and 9 atypical AD cases. The presence of amyloid-beta (N-terminal; IC16), pTau (AT8), reactive astrocytes (GFAP), microglia (Iba1, CD68, and HLA-DP/DQ/DR), and complement factors (C1q, C3d, C4b, and C5b-9) was quantified by image analysis. Differences in lobar distribution patterns of immunoreactivity were statistically assessed using a linear mixed model. Results: We found a temporal dominant distribution for amyloid-beta, GFAP, and Iba1 in both typical and atypical AD. Distribution of pTau, CD68, HLA-DP/DQ/DR, C3d, and C4b differed between AD variants. Typical AD cases showed a temporal dominant distribution of these markers, whereas atypical AD cases showed a parietal dominant distribution. Interestingly, when quantifying for the number of amyloid-beta plaques instead of stained surface area, atypical AD cases differed in distribution pattern from typical AD cases. Remarkably, plaque morphology and localization of neuroinflammation within the plaques was different between the two phenotypes. Conclusions: Our data show a different localization of neuroinflammatory markers and amyloid-beta plaques between AD phenotypes. In addition, these markers reflect the atypical distribution of tau pathology in atypical AD, suggesting that neuroinflammation might be a crucial link between amyloid-beta deposits, tau pathology, and clinical symptoms. Keywords: Atypical Alzheimer’s disease, Microglia, Complement, Human brain tissue, Neuroinflammation, Amyloid-beta plaque * Correspondence: b.boon@vumc.nl Department of Neurology, Alzheimer Center, Amsterdam Neuroscience, VU University Medical Center, Amsterdam, The Netherlands Department of Pathology, Amsterdam Neuroscience, VU University Medical Center, Amsterdam, The Netherlands © 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. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 2 of 16 Background Evidence for the correlation of pTau, neuroinflamma- Patients with Alzheimer’s disease (AD) typically tion, and microglia in AD subtypes is lacking. In this present with episodic memory impairment followed study, we aimed to explore whether an atypical spread- by deterioration of executive functioning, praxis, and ing of NFT pathology in non-amnestic AD is associated visuospatial skills. However, AD patients may also with an atypical distribution of neuroinflammation. In a present with an atypical phenotype [1, 2]. An atypical well-defined cohort of typical and atypical AD, we presentation is seen in 10% of the late-onset AD assessed and compared the presence of pTau, amyloid- (LOAD) patients (≥ 65 years of age) and up to 30% of beta, (activated) glial cells, and complement proteins in the early onset (< 65 years) AD (EOAD) patients [3]. temporal and parietal cortical areas. So far, three variants of atypical AD have been de- scribed: the posterior cortical atrophy (PCA) variant Methods characterized by visuoperceptual problems [4], the Post-mortem brain tissue logopenic variant characterized by aphasia [5], and Post-mortem brain tissue was obtained from the the frontal variant associated with behavioral changes Netherlands Brain Bank (NBB; Amsterdam, The [1, 2]. In addition to clinical differences, these differ- Netherlands). Donors signed informed consent for brain ent AD variants show syndrome-specific atrophy pat- autopsy, and the use of brain tissue and medical records terns on MRI [6]. for research purposes. Neuropathological diagnosis was AD is characterized by the deposition of amyloid-beta based on histochemical stainings including hematoxylin plaques and the formation of neurofibrillary tangles and eosin, congo red staining, Bodian or Gallyas and (NFT) in the brain. During disease progression, both methenamine silver stainings, and immunohistochemical plaques and NFTs are assumed to spread through the stainings for amyloid-beta, pTau, alpha-synuclein, and brain in a fixed order [7, 8]. However, the typical NFT p62. These stainings were performed on formalin-fixed distribution as originally described by Braak and Braak paraffin-embedded (FFPE) brain tissue of multiple brain [9] does not seem to hold for all AD cases. Clinicopatho- regions including the frontal cortex, temporal pole, logical studies indicated that AD patients with an atyp- superior parietal lobe, occipital pole, amygdala, and the ical phenotype have an atypical NFT distribution [9, 10]. hippocampus. Neuropathological diagnosis of AD was Furthermore, this atypical NFT distribution was demon- based on Braak stages for NFT and amyloid [7], Thal strated in living AD patients using the tau tracer phases for amyloid-beta [8], and CERAD criteria for [ F]AV1451 [11]. While the atypical distribution of NFTs neuritic plaques [26]. corresponds with the observed clinical phenotype, the cause of this difference in NFT spreading between AD Selection of typical and atypical AD cases variants remains elusive. Between 1996 and 2014, 352 AD cases came to autopsy There is accumulating evidence that inflammation and were semi-quantitatively scored by two neuropathol- plays a prominent role in the pathogenesis of AD. Re- ogists (WK, AR) for the NFT load using Bodian or cently, genome-wide association studies have identified Gallyas staining in the temporal pole, the frontal, super- several genes involved in inflammation, especially those ior parietal, and occipital cortex as previously described engaged in microglia function, as risk factors for devel- by Hoogendijk et al. [27]. The NFT load was scored in a oping AD [12–15]. The AD brain shows an increased 0.4-mm area as being absent (0), sparse (1), mild (2; 2 presence of activated microglia, reactive astrocytes, pro- to 3 NFTs), or severe (3; > 3 NFTs) for each brain region inflammatory cytokines, acute phase proteins, and acti- separately. From this cohort, we selected cases with an vated complement proteins compared to controls [16]. NFT score of 3 in either the temporal or parietal section, Complement proteins co-localize with NFTs [17, 18], as or in both sections, resulting in 296 cases (for flowchart, well as with amyloid-beta deposits [19], and are actively see Fig. 1). For 142 cases, the NFT score was higher in involved in the formation of these pathological struc- the temporal section than the parietal section. These tures. Clusters of activated microglia are found in amyl- cases were referred to as having a typical NFT oid plaques, and the presence of activated microglia distribution [7]. In 126 of 296 cases, an NFT score of 3 increases with disease severity [20, 21]. Recent disease wasfound in the temporal aswell as the parietal section. A models suggest that microglia are actively involved in higher NFT score in the parietal compared to the temporal the spreading of phosphorylated Tau (pTau) [22–24]. section was observed in 28 cases and was defined as a Tau pathology is heavily reduced in disease-modeled parietal dominant and thus atypical NFT distribution. mice that are depleted for microglia compared to To study the distribution of neuroinflammation in typ- their microglia-positive peers [23, 24]. In the human ical and atypical AD, we further refined our cohort to brain, the presence of activated microglia correlates include only cases with a concordance between clinical with Braak staging for NFTs [25]. presentation and NFT distribution. The clinical Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 3 of 16 Fig. 1 Flowchart of neuropathologically assessed and semi-quantitatively scored AD cohort. Between 1996 and 2014, 352 AD cases came to autopsy and were semi-quantitatively scored for NFTs as described by Hoogendijk et al. [27]. In 296 cases, an NFT score of ≥ 3 in the temporal and/or parietal cortex was observed. Typical NFT distribution was defined as a higher NFT score in the temporal compared to the parietal cortex. Atypical NFT distribution was defined as a higher NFT score in the parietal compared to the temporal section. From the cases with typical and atypical NFT distribution, 18 cases per group were selected for which the clinical phenotype was stratified as either amnestic or non-amnestic (results shown in Table 2). From these clinical phenotyped cases, 9 cases with an atypical NFT as well as non-amnestic clinical presentation were compared with 10 cases with an amnestic presentation and typical NFT distri- bution using immunohistochemistry. Mean ± SD is shown for age at death and disease duration in years. AD Alzheimer’s disease, NFT neurofibrillary tangle phenotype of 36 cases with atypical and typical NFT randomly chose 18 cases with both typical NFT distribu- pathology was retrospectively assessed. In 18 out of 28 tion and sufficient clinical information for clinical phe- cases with an atypical NFT distribution, the available notyping. Clinical assessment was performed clinical information was sufficient to come to a retro- retrospectively and independently by 2 cognitive neurol- spective clinical diagnosis (see Table 1 for demograph- ogists (YP and FB) using the NIA-AA criteria [2]. Both ics). To have an equal group for comparison, we clinicians were blinded to the pathological stratification Table 1 Demographics of 36 cases with typical and atypical NFT distribution for which extensive retrospective clinical assessment was performed Typical NFT distribution Atypical NFT distribution n =18 n =18 Phenotype Amnestic (n = 16) Non-amnestic (n = 2) Amnestic (n = 5) Non-amnestic (n = 13) Male, n (%) 6 (37) 0 3 (60) 8 (62) Age at death 82 (± 7) 88 (± 5) 71 (± 11) 67(± 7) Disease duration 8 (± 5) 7 (± 4) 11 (± 5) 8 (± 4) NFT stage [7] n per stage 4/5/6 2/10/4 1/1/0 0/2/3 0/7/6 Amyloid stage [7] n per stage O/A/B/C 0/0/16 0/0/2 0/0/5 0/1/12 Data are mean ± SD. Age at death and disease duration shown in years NFT neurofibrillary tangle Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 4 of 16 when assessing the clinical phenotype. The clinical Immunohistochemistry (IHC) stratification was based on (collateral) history and cogni- IHC was performed to detect pTau (AT8); amyloid-beta tive examination documented by the clinical neurologist. (N-terminal; IC16); reactive astrocytes (GFAP); microglia Cases with first-degree relatives affected by EOAD were (Iba1); activated microglia (CD68 and HLA-DP/DQ/DR); excluded to minimize the risk of genetic AD. Other ex- and complement proteins C1q, C3d, C4b, and C5b-9 clusion criteria were sepsis, other neurodegenerative or (Table 3). FFPE sections (5-μmthick) from the temporal psychiatric diseases, significant cerebrovascular disease, pole and superior parietal lobe of the right hemisphere post-mortem interval > 12 h, and prior known genetic were used. mutations. Due to our exclusion criteria and availability IHC for pTau, amyloid-beta, GFAP, and Iba1 was per- of archived brain tissue samples, our inclusion was lim- formed using the Ventana BenchMark ULTRA staining ited to 9 atypical AD cases and 10 typical AD cases of system (Roche, Basel, Switzerland). Tissue sections were which the temporal pole and superior parietal lobe were mounted on TOMO adhesive glass slides (Matsunami, assessed by immunohistochemistry (see Fig. 1 for inclu- Osaka, Japan) and deparaffinized. After blocking for sion flowchart, Table 2 for patient details, and Table 4 endogenous peroxidase, antigen retrieval was performed for demographics in the “Results” section). Cases were by heating sections at 100 °C in Cell Conditioning 1 not intentionally matched for disease duration, brain solution (pH 8.5) (Roche) for different durations per weight, ApoE status, or post-mortem interval. antibody (see Table 3). For detection of primary Table 2 Clinical and neuropathological characteristics of typical and atypical AD cases Case Phenotype Symptoms at clinical presentation Sex Age at Disease NFT Amyloid Brain Cause of death PMI ApoE death duration stage [7] stage [7] weight (grams) 1 Typical AD Memory F 92 8 5 C 933 Heart failure 7:00 43 2 Typical AD Memory, disorientation F 84 4 4 C 908 Cardiogenic 4:15 32 shock 3 Typical AD Memory, disorientation F 84 9 5 C 827 Cachexia 6:40 43 4 Typical AD Memory F 89 5 5 C 962 Pneumonia 6:28 43 5 Typical AD Memory F 83 6 6 C 1100 Dehydration 6:17 43 6 Typical AD Memory, behavior F 91 3 4 C 1026 Cachexia 6:25 33 7 Typical AD Memory, F 77 2 5 C 999 Pneumonia 6:05 33 8 Typical AD Memory, behavior M 70 2 6 C 1261 Metastasized 6:20 43 colon carcinoma 9 Typical AD Memory F 76 12 5 C 1223 Unknown 10:45 44 10 Typical AD Memory M 60 2 6 C 1191 Cachexia 6:15 43 11 Atypical AD Aphasia, dyscalculia, agraphia, F 65 6 6 C 975 Pneumonia 5:40 33 left-right agnosia, visuoconstruction problems 12 Atypical AD Aphasia, dyslexia, apraxia, M 65 2 6 C 1057 Cardiac 8:50 44 visuoconstruction problems insufficiency 13 Atypical AD Aphasia, acalculia, M 64 7 5 C 1135 Pneumonia 4:45 42 fingeragnosia, apraxia 14 Atypical AD Parkinsonism, falling, alien F 67 3 5 C 817 Epileptic insult 7:35 33 hand syndrome 15 Atypical AD Aphasia, apathy, agitation M 59 6 6 C 1300 Cachexia 5:05 44 16 Atypical AD Aphasia, dyscalculia, dyslexia, M 62 3 6 C 1116 Malign 4:15 43 disorientation neuroleptic syndrome 17 Atypical AD Aphasia, dyslexia, apathy, apraxia M 65 1 5 C 1150 Euthanasia 6:50 43 18 Atypical AD Aphasia, dyslexia, apraxia, M 62 6 5 B 1153 Cachexia 4:40 33 visuospatial problems, behavior 19 Atypical AD Visual hallucinations, psychosis M 61 6 6 C 1355 Pneumonia 5:00 43 Age at death and disease duration in years; post-mortem interval in hours to minutes. Typical AD defined as more neurofibrillary tangles in the temporal compared to the parietal cortex by semi-quantitative scoring as described by Hoogendijk et al. [27] and an amnestic presentation during life. Atypical AD defined as more neurofibrillary tangles assessed by semi-quantitative scoring in the parietal compared to the temporal cortex and a non-amnestic presentation AD Alzheimer’s disease, F female, M male, NFT neurofibrillary tangle, PMI post-mortem interval Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 5 of 16 Table 3 Characteristics of primary antibodies and staining details Antibody Antigen Species Origin details Dilution Incubation Antigen retrieval Detection time method pTau, clone AT8 Tau phosphorylated at Ser202 Mouse ThermoFisher, Pittsburgh, USA 1:10000 32 min Heat-induced (pH 8.5) Optiview and Thr205 IgG1 at 36 °C for 24 min Amyloid-beta, N-terminus of amyloid-beta Mouse Dr. Carsten Korth, University 1:25 32 min Heat-induced (pH 8.5) Optiview clone IC-16 (aa 1-16) IgG2a of Dusseldorf, Germany at 36 °C for 24 min GFAP, clone Glial fibrillary acidic protein Mouse Roche, Basel, Switzerland 1: 2 8 min Heat-induced (pH 8.5) Optiview EP672Y at 37 °C for 32 min Iba1 C-terminus of Iba1 Rabbit Wako Pure Chemical 1:4000 32 min Heat-induced (pH 8.5) Optiview Industries, Osaka, Japan at 36 °C for 16 min CD68, clone CD68 Mouse Dako, Glostrup, Denmark 1:1200 Overnight Heat-induced (pH 6.0) EnVision KP1 IgG1 at 4°C by autoclave HLA-DP/DQ/DR, Alpha and beta-chains of all Mouse Dako 1:800 Overnight Heat-induced (pH 6.0) EnVision clone CR3/43 products of the DP, DQ, and IgG1 at 4°C by autoclave DR subregions C1q C1q Rabbit Dako 1:25600 Overnight Heat-induced (pH 6.0) EnVision at 4°C by autoclave C3d C3d Rabbit Dako 1:3200 Overnight Heat-induced (pH 6.0) EnVision at 4°C by autoclave C4b C4b Rabbit Abcam, Cambridge, United 1:1600 Overnight Heat-induced (pH 6.0) EnVision Kingdom at 4°C by autoclave C5b-9, clone Neoepitope on C9 in the Mouse Hycult Biotech, Plymouth 1:400 Overnight Heat-induced (pH 6.0) EnVision WU13-15 membrane attack complex meeting, USA at 4°C by autoclave antibodies with 3,3′-diaminobenzidine tetrahydrochlo- perpendicular to the cortical surface of the cortex were ride (DAB), Optiview DAB IHC detection kit (Roche) photographed. Total surface, depending on the width of was used. Finally, the sections were mounted with Cov- the cortex, could vary for each ROI and contained at erslipping film (Sakura Tissue-Tek, Leiden, The least 2 columns. Images were taken using a × 10 object- Netherlands). ive on an Olympus BX 41 photomicroscope with a Leica IHC for CD68, HLA-DP/DQ/DR, C1q, C3d, C4b, and MC 170 HD digital camera. The presence of DAB stain- C5b-9 was performed manually. The sections were ing was quantified with ImageJ (NIH) using the color mounted on SuperFrost Plus glass slides (Menzel-Gläser, threshold plugin. Our outcome measurement was the Braunschweig, Germany) and deparaffinized. Subse- percentage of DAB-stained area per marker, also referred quently, the sections were blocked for endogenous per- to as immunoreactivity. In addition to immunoreactivity, oxidase using 0.3% hydrogen peroxide in phosphate we quantified the number of amyloid-beta and C4b pla- buffer saline (PBS; pH 7.4). The sections were immersed ques. For the amyloid-beta plaques, diffuse deposits were in sodium citrate buffer (10 mM sodium citrate, 5 M not taken into account and only dense plaques were NaOH, dH O, pH 6.0) and heated to 120 °C in an auto- quantified, defined as particles with an immunoreactive clave for antigen retrieval. Primary antibodies were surface area of 100 μm or more [29, 30]. C4b-positive diluted in normal antibody diluent (ImmunoLogic, Dui- deposits of the same surface area were quantified as a ven, The Netherlands) and incubated overnight at 4 °C. measurement of the atypical appearing plaques as de- Primary antibodies were detected using EnVision (Dako, scribed in the “Results” section. Glostrup, Denmark). Between steps, the sections were washed in PBS. Subsequently, antibodies were visualized Fluorescent triple stainings with DAB (Dako). After counterstaining with Co-localization of C4b and CD68 or HLA-DP/DQ/DR hematoxylin, the sections were mounted with Entellan with amyloid-beta and thioflavine S was visualized in the (Merck, Darmstadt, Germany). parietal section of 4 atypical AD cases and the temporal section of 2 typical AD cases. The typical AD cases Image analyses and quantitative assessment of served as positive controls and reference since immunostainings localization of complement and microglia in classical- For quantitative assessment, 2 regions of interest (ROI) cored plaques is widely described in literature (for CD68 were randomly selected within non-curved areas of each [20]/for complement [16, 19]). section containing all 6 cortical layers [28]. Within each After deparaffinization, the sections were submerged ROI, contiguous microscopic fields arranged in columns in sodium citrate buffer and heated to 120 °C in an Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 6 of 16 autoclave. Subsequently, the sections were incubated model. Statistical significance was set at p <.05 for compari- using different combinations of primary antibodies: son of baseline characteristics. mouse IgG2a-anti-amyloid-beta (1:200), rabbit-anti-C4b (1:200), and mouse IgG1-anti-HLA-DP/DQ/DR (1:25) or Results mouse IgG1-anti-CD68 (1:300). Antibodies were diluted Atypical AD cases are younger than typical AD cases in normal antibody diluent (ImmunoLogic) and incu- The post-mortem cohort used for IHC consisted of 9 bated overnight at 4 °C. Subsequently, the sections were atypical AD cases and 10 typical AD cases. For a sum- incubated with the following secondary antibodies: goat- mary of initial clinical symptoms at presentation for each anti-mouse IgG2a Alexa Fluor dye 594, goat-anti-mouse case, see Table 2. Most atypical AD cases presented with IgG1 Alexa Fluor dye 647, and donkey-anti-rabbit Alexa symptoms of aphasia, consisting of word-finding difficul- Fluor dye 647 (1:250 dilution, ThermoFisher, Waltham, ties and spelling mistakes, combined with apraxia. None USA). For visualization of amyloid structures, the sec- of our atypical cases retrospectively met the criteria for tions were counterstained with thioflavine S (1% in an isolated primary progressive aphasia [5]. One case dH O) and subsequently rinsed in 70% ethanol. Autoflu- presented with Parkinsonism and an alien hand syn- orescence was blocked with 0.1% Sudan black in 70% drome, fitting a corticobasal syndrome during life. All 10 ethanol for 5 min. Between steps, the sections were typical AD cases presented with memory complaints as rinsed with PBS. Finally, the sections were enclosed with most prominent initial symptom. Demographic character- 80% glycerol/20% tris buffered saline. Representative pic- istics of both AD phenotypes used for immunohistochem- tures were taken with a Leica DMi8 inverted fluorescent ical analysis are shown in Table 4. Similar to the large microscope equipped with a Leica DFC300 G camera. cohort of 296 AD subjects (see Fig. 1), atypical AD pa- tients selected for IHC analysis were younger at age of Statistical analysis death and more often male. The disease duration, brain Demographics of the typical and atypical AD groups were weight, post-mortem interval, disease severity, and ApoE compared using Fisher’s exact test for categorical and genotype did not differ between groups. Mann-Whitney U test for numerical and not normally dis- tributed data. Outcome measures were compared between Distribution of pTau and amyloid-beta in typical and the 2 AD groups by using linear mixed model analysis. Lin- atypical AD ear mixed model analysis was used to adjust for the nested Immunohistochemistry for pTau showed neuronal inclu- observations within cases. In the linear mixed model ana- sions as well as neuritic threads (Fig. 2a). Typical AD lyses, the group variable (typical versus atypical AD), the re- cases showed more pTau immunoreactivity in the tem- gion (temporal versus parietal), and the interaction between poral compared to the parietal section (Fig. 2c). This group and region were added. Correcting for age and sex was contrary to the pTau distribution in atypical AD made the model less stable and was therefore not per- cases, in which the parietal section showed more formed. An assumption to apply a linear mixed model is that residuals of outcome measurements are normally dis- tributed. To meet this assumption, all outcome variables Table 4 Demographic characteristics of the AD cases used for (pTau, amyloid-beta, GFAP, Iba1, HLA-DP/DQ/DR, CD68, immunohistochemical analysis C1q, C3d, C4b, number of amyloid-beta plaques, and num- Typical AD Atypical AD p-value ber of C4b plaques) were transformed by taking the natural (n = 10) (n =9) log of the (variable + 1). The covariance structure was set Male, n 27<.05 to unstructured. Using the linear mixed model, we an- Age at death 81 (± 10) 63 (± 3) < .01 swered if the difference in outcome measurement over the Disease duration 5 (± 3) 4 (± 2) .78 2 regions was different between the 2 AD phenotypes (re- Brain weight (grams) 1043 (± 146) 1117 (± 161) .32 gion × phenotype), also referred to as interaction effect. Both phenotypes showed a similar distribution over the 2 PMI (h:min) 6:19 (± 1:48) 5:51 (± 1:33) .66 regions if no interaction effect was found. Statistical analysis NFT stage [7] 2/5/3 0/4/5 .46 n per stage 4/5/6 was performed in IBM SPSS statistics version 22.0 (IBM SPSS Statistics, Armonk, NY, USA). Bonferroni correction Amyloid stage [7] 0/0/10 0/1/8 .47 n per stage O/A/B/C was used to correct for multiple testing. Since we tested 11 outcome measurements (amyloid-beta, pTau, GFAP, Iba1, ApoE genotype 1/2/0/6/1 0/3/1/3/2 .48 n per category CD68, HLA-DP/DQ/DR, C1q, C3d, C4b, # amyloid-beta 32/33/42/43/44 plaques, and # C4b plaques), statistical significance was set Data in mean (± SD). Age at death and disease duration in years. Mann-Whitney at p < .0045 (p < .05/11 outcome measurements) for each U test for continuous data. Fisher’s exact test for categorical data effect (region × phenotype and region) of the linear mixed AD Alzheimer’s disease, NFT neurofibrillary tangle, PMI post-mortem interval Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 7 of 16 Fig. 2 pTau and amyloid-beta distribution in typical AD and atypical AD. a In typical AD, the temporal cortex (blue border) shows more immunoreactivity for pTau than the parietal cortex (burgundy border). This distribution is inversed in atypical AD (boxplot in c). b Although both typical and atypical AD show more overall amyloid-beta immunoreactivity in the temporal cortex compared to the parietal region, the atypical AD group shows increased number of amyloid-beta plaques in the parietal compared to temporal section (boxplot in e). Bar represents 100 μm. c, d,and e Boxplots showing pTau immunore- active area (%), amyloid-beta immunoreactive area (%), and the number of amyloid-beta plaques, respectively, in the temporal and parietal section ofboth AD phenotypes. Data shown as median (bar), 1st and 3rd quartile (box boundaries), and min to max (error bars). A difference in distribution over the two regions between the two AD phenotypes is indicated by # (Table 5), * p < .0045 immunoreactivity compared to the temporal section. In the atypical AD group. Besides a difference in plaque addition, the distribution of pTau over the 2 regions dif- number, we observed a contrast in plaque morphology fered significantly between the 2 phenotypes (Table 5). between the two groups, which will be addressed below. Amyloid-beta immunoreactivity was present in the form of diffuse deposits, dense plaques, and classical Glial activation is increased in atypical AD cored plaques (Fig. 2b). Both the typical and atypical AD Staining for GFAP-positive astrocytes showed variably group showed more immunoreactivity for amyloid-beta sized star-like GFAP-positive structures in all AD cases in the temporal than parietal section (Fig. 2d), and no (Fig. 3a). Both phenotypes showed higher levels of difference in distribution was observed (Table 5). While immunoreactivity for GFAP in the temporal section total immunoreactivity levels did not show differences compared to the parietal section. Atypical AD cases between the 2 phenotypes, the number of dense plaques showed relatively more GFAP immunoreactivity in the with an immunoreactive surface area of 100 μm or parietal cortex compared to typical AD cases (Fig. 3c; more did show a significant difference in distribution Table 5). over the 2 regions between the 2 phenotypes (Fig. 2e) Iba1 immunostaining showed positivity in both the (Table 5). In contrast to the typical AD group, more cell soma as well as the processes of the microglia, dense plaques were observed in the parietal section of mostly in the form of ramified microglia (Fig. 3b). Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 8 of 16 Table 5 Results for the linear mixed model of transformed immunohistochemistry variables Transformed variable Fixed Effect Phenotype Results for linear mixed model Beta-coefficient 95% CI pTau Region x phenotype 1.53* [1.33 – 1.72] Region Typical AD - 0.82* [- 0.95 – - 0.68] Atypical AD 0.71* [0.57 – 0.85] Amyloid-beta Region x phenotype 0.04 [- 0.06 – 0.15] Region Both - 0.46* [- 0.54 – - 0.39] GFAP Region x phenotype 1.25* [1.02 – 1.48] Region Typical AD - 1.71* [- 1.87 – - 1.56] Atypical AD - 0.46* [- 0.63 – - 0.29] Iba1 Region x phenotype 0.10 [0.002 – 0.19] Region Both - 0.47* [- 0.53 – - 0.41] CD68 Region x phenotype - 0.43* [0.38 – 0.48] Region Typical AD - 0.13* [- 0.17 – - 0.10] Atypical AD 0.29* [0.26 – 0.33] HLA-DP/DQ/DR Region x phenotype 0.97* [0.88 – 1.06] Region Typical AD - 0.08* [- 0.14 – - 0.03] Atypical AD 0.89* [0.82 – 0.96] C1q Region x phenotype - 0.05 [- 0.15 – 0.06] Region Both 0.03 [- 0.05 – 0.10] C3d Region x phenotype 0.74* [0.62 – 0.86] Region Typical AD - 0.39* [ - 0.47 – - 0.31] Atypical AD 0.36* [0.26 – 0.45] C4b Region x phenotype 0.95* [0.84 – 1.07] Region Typical AD - 0.24* [- 0.32 – - 0.17] Atypical AD 0.71* [0.62 – 0.80] # of amyloid-beta plaques Region x phenotype 0.58* [0.41 – 0.75] Region Typical AD - 0.33* [- 0.44 – - 0.21] Atypical AD 0.25* [0.13 – 0.37] # of C4b plaques Region x phenotype 1.84* [1.58 – 2.11] Region Typical AD - 0.39* [- 0.56 – - 0.22] Atypical AD 1.46* [1.25 – 1.66] Results of the linear mixed model for analyzed immunohistochemistry variables are shown. All variables were transformed: ln(variable + 1). We tested if the distribution over the 2 regions was different between the 2 AD phenotypes, defined as the interaction effect: region × phenotype. When an interaction effect was found, the beta-coefficient is shown per phenotype. To correct for multiple testing, a p value < .0045 was considered significant (p < .05/11 outcome measurements) and indicated with * AD Alzheimer’s disease, CI confidence interval # Number Both AD groups showed a similar temporal domi- CD68 and HLA-DP/DQ/DR in the parietal compared to nancy for Iba1. the temporal section, which was in contrast to the typ- Activated microglia were stained using CD68 and ical AD cases. In addition, the levels of CD68 and HLA- HLA-DP/DQ/DR. Whereas CD68 positivity was mostly DP/DQ/DR immunoreactivity were relatively high in the found in the soma of microglia, HLA-DP/DQ/DR parietal section of atypical AD compared to the temporal showed a prominent staining in the processes of micro- section of typical AD. glia (Fig. 4a, b). Both markers showed a different distribution over the Increased presence of complement proteins in atypical AD 2 regions between the 2 AD phenotypes (Fig. 4c, d;Table 5). To visualize different parts of the complement cascade, we Atypical AD cases showed more immunoreactivity for stained for C1q, C3d, and C4b, representing the start of the Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 9 of 16 Fig. 3 GFAP and Iba1 immunoreactivity is temporal dominant in both AD phenotypes. a, b The temporal cortex (blue border) shows more GFAP and Iba1 immunoreactivity respectively than the parietal cortex (burgundy border) in typical and atypical AD. Bar represents 100 μm. c, d Boxplot showing GFAP and Iba1 immunoreactive area (%) in the temporal and parietal section, respectively, of both AD phenotypes. Atypical AD shows relatively more GFAP immunoreactivity in the parietal cortex than typical AD (Table 5). Iba1 distribution is not different between the two phenotypes. Data shown as median (bar), 1st and 3rd quartile (box boundaries), and min to max (error bars). *p <.0045 cascade, as well as for C5b-9, defining the end-stage of the the parietal cortex (Fig. 5e, f; Table 5). In contrast, atyp- cascade and forming the membrane attack complex. ical AD showed higher immunoreactivity levels for both C1q immunoreactivity was observed as a weak diffuse markers in the parietal section compared to the tem- staining in the form of plaque-like structures, as punctu- poral section. ate staining of the neuropil, and sometimes in neurons Analysis of C5b-9 exposed very low to no immunore- (Fig. 5a). While the plaque-like structures were more activity in both regions of both phenotypes (data not often observed in the parietal section, the punctate stain- shown). For this reason, no quantification or statistical ing of the neuropil was more prominent in the temporal analysis was performed for C5b-9. The few structures section of both AD groups. No difference in C1q immu- that were C5b-9 positive included parenchymal and noreactivity was observed between regions or AD phe- meningeal vessels. The serum within these vessels also notypes (Fig. 5d; Table 5). stained positive for C5b-9. Compared to C1q, staining for C3d and C4b showed Our data show a different distribution in temporal an intense plaque-like staining, which had a morphology and parietal regions between both AD phenotypes for resembling that of compact and classical cored plaques complement factors C3d and C4b. Both complement (Fig. 5b, c). Levels of immunoreactivity for C3d and C4b factors were mostly abundant in the parietal section in typical AD were higher in the temporal compared to of atypical AD. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 10 of 16 Fig. 4 CD68 and HLA-DP/DQ/DR immunoreactivity show a parietal dominant distribution in atypical AD. a, b In typical AD, the temporal section (blue border) shows more CD68 and HLA-DP/DQ/DR immunoreactivity than the parietal section (burgundy border). In contrast, atypical AD shows more CD68 and HLA-DP/DQ/DR immunoreactivity in the parietal than the temporal section. Note the relative difference in immunoreactivity in the parietal section of atypical AD compared to the temporal section in typical AD. Bar represents 100 μm. c, d Boxplot showing CD68 and HLA- DP/DQ/DR immunoreactive area (%) in the temporal and parietal section, respectively, of both AD phenotypes. Linear mixed model shows a different distribution over the 2 regions between phenotypes (#) (Table 5). *p < .0045 Different plaque appearance in typical and atypical AD This staining pattern was compared with that of Looking at the morphology of plaques stained by classical-cored plaques observed in typical AD. In amyloid-beta and C4b, we observed a difference in classical-cored plaques, C4b, amyloid-beta, and thiofla- appearance between the 2 AD phenotypes. Amyloid-beta vine S co-localized in the core of the plaque, while the and C4b immunoreactive plaques in the parietal section corona only stained for amyloid-beta. Compared to of atypical AD cases showed a more granular compos- cored plaques, coarse-grained plaques showed a fibrillar, ition compared to deposits in the temporal section of less organized morphology with co-localization of C4b, this phenotype or compared to deposits in both regions amyloid-beta, and thioflavine S all over the plaque sur- of the typical AD cases (Fig. 6). The surface area positive face. Since an increased presence of activated microglia for C4b of these coarse-grained plaques was larger (> in atypical AD was observed, the localization of CD68 100 μm ) than that of typical plaques. Atypical AD cases and HLA-DP/DQ/DR with cored and coarse-grained had more of these C4b plaques in the parietal compared plaques was compared (Fig. 7). Like cored plaques, to the temporal cortex, which was contrary to typical coarse-grained plaques were associated with clusters of AD cases (Fig. 6i; Table 5). The coarse-grained plaques CD68 and HLA-DP/DQ/DR-positive microglia. In cored observed in the parietal cortex of atypical AD triple- plaques, activated microglia were located between the stained for C4b, amyloid-beta, and thioflavine S (Fig. 7). core and corona of the plaque. In coarse-grained Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 11 of 16 Fig. 5 Distribution of C3d and C4b differs between AD phenotypes. a In both typical and atypical AD, C1q deposition in the temporal cortex (blue border) is more diffuse than in the parietal cortex (burgundy border). b, c In typical AD, the temporal cortex shows more immunoreactivity for C3d and C4b than the parietal cortex. This distribution is inverted in atypical AD, showing more immunoreactivity for C3d and C4b in the parietal than temporal section. Bar represents 100 μm. d, e, and f Boxplots of immunoreactive area (%) for C1q, C3d, and C4b, respectively, in the temporal and parietal section of both AD phenotypes. Linear mixed model shows a different distribution for C3d and C4b over the 2 regions between phenotypes (#) (Table 5). *p < .0045 plaques, the localization of CD68 and HLA-DP/DQ/DR- we observed that also the occurrence of pTau is differ- positive microglia was less structured and positive ently distributed in typical and atypical AD. Here, we microglia appeared throughout the plaque. This data show for the first time that the distribution of both acti- supports a morphological difference between cored and vated microglia and complement factors between the coarse-grained plaques, of which the latter occurs prom- temporal and parietal lobe differentiates atypical from inently in atypical AD (Fig. 6i). typical AD cases. In addition, atypical AD cases are characterized by the presence of plaques with an abnor- Discussion mal morphology, highlighted by an alternative Atypical AD is characterized by a different distribution localization of microglia and presence of complement of NFTs when compared with typical AD. As expected, proteins. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 12 of 16 Fig. 6 Plaques in atypical AD show a different morphology compared to plaques in typical AD. a–d In typical AD cases, the morphology of amyloid-beta (a for overview, b for detail) and C4b (c for overview, d for detail) deposits come in the form of diffuse, dense, or classical cored plaques. e–h In atypical AD cases, amyloid-beta (e for overview, f for detail) and C4b (g for overview, h for detail) deposits show a distinct morphology, being coarse-grained and affecting a larger surface area (> 100 μm ) compared to classical cored plaques. Pictures are taken in the region with the highest number of amyloid-beta plaques, being temporal for typical AD and parietal for atypical AD (Fig. 2e). Bars represent 100 μm. i Boxplot ofnumberofcoarse-grained plaques quantified using C4b staining in the temporal and parietal cortex of typical and atypical AD. Linear mixed model shows a different distribution for coarse-grained plaques over the 2 regions between phenotypes (#) (Table 5). This distribution is parietally dominant in atypical AD. *p < .0045 In this study, typical and atypical AD were defined from this study do not necessarily apply for patients with according to the distribution of NFTs as well as their a logopenic or behavioral phenotype, as tau pathology clinical presentation. In our cohort, atypical AD cases seems to be differently distributed in those cases [11, were younger and more often male than typical AD 33]. Besides neuropathological studies, also clinical stud- cases. This was also seen in an earlier study by Murray ies report that an atypical clinical presentation is more and colleagues [31], indicating that an atypical distribu- common at a younger age [3]. Regarding differences in tion of pathology seems to be more common in men gender, an atypical presentation is not per se more com- and at a younger age. In line with this, another study mon in men [3]. However, AD presenting at late-onset is showed that cases with EOAD showed higher mean more common in women [34], explaining the relative levels of NFTs in the parietal lobe, when comparing high number of female subjects in typical AD cohorts. EOAD to LOAD, irrespective of clinical presentation We did not observe a different distribution of total [32]. When taking symptomology into account, this par- amyloid-beta immunoreactivity between 2 AD pheno- ietal dominant tau distribution is especially common in types. This is in line with other studies reporting that the PCA variant of atypical AD [11]. Therefore, the amyloid-beta distribution is not different between AD current atypical AD cohort defined by parietal dominant subtypes [9, 32, 35–37]. However, when quantifying for NFTs represents only a subgroup of atypical AD. Results number of dense amyloid-beta plaques, atypical AD cases Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 13 of 16 Fig. 7 Different plaque morphology in the parietal cortex of atypical AD. First row: in typical AD, classical cored plaques in the temporal cortex show an organized staining pattern with a corona showing merely amyloid-beta positivity versus a core positive for thioflavine S, amyloid-beta, and C4b. Second row: in atypical AD, fibrillar plaques in the parietal cortex show co-localization of thioflavine S, amyloid-beta, and C4b in the form of fibrils throughout the whole plaque. Third row: in typical AD, CD68-positive microglia are localized between the core and corona of classical cored plaques. Fourth row: in atypical AD, CD68-positive microglia localization is less organized. Fifth + sixth row: this different distribution within plaques between the two phenotypes also holds for HLA-DP/DQ/DR-positive microglia. Bar is applicable to all images and represents 100 μm showed a parietal dominant distribution compared to typ- In the present study, we observed a clear morpho- ical AD cases, which showed a temporal dominant distri- logical difference between classical-cored plaques in bution. This distinction in number of plaques was also typical AD and dense plaques in atypical AD cases. reported by Hoff and colleagues who compared cases with Dense plaques in atypical AD cases have a coarse- PCA to typical AD cases using stereology [38]. These re- grained structure, as observed with amyloid-beta sults indicate that although the distribution of overall immunostaining and thioflavine S staining. In addition, amyloid-beta immunoreactivity is similar, the number of these coarse-grained plaques showed a strong immuno- dense amyloid-beta plaques may differ between AD phe- reactivity for complement. Multiple studies have notypes. In addition to the number of dense amyloid-beta reported complement proteins to be associated with plaques, structural differences in amyloid-beta plaques amyloid-beta deposits [19, 39]. However, a difference in might be associated with the pathological and clinical dif- complement activation between AD subtypes has so far ferences between typical and atypical AD. not been described. The increased presence of Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 14 of 16 complement in coarse-grained plaques in atypical AD neuroinflammatory response [25]. These studies sug- cases supports a difference in amyloid structure be- gest that sex and age influence microglia activation. tween typical and atypical AD. Most likely, the amyloid However, whether differences in sex and age contrib- structure of fibrillar plaques favors a strong binding ute to differences in regional distribution, as observed and activation of complement factors, which in turn in the current study between AD subtypes, remain could act as opsonins for phagocytosis carried out by elusive. microglia [17, 40, 41]. A structural variation in amyloid-beta fibrils in combination with a difference Conclusions in binding of amyloid associated proteins may under- The results of this study are in line with the assump- lie the observed difference in the occurrence of neu- tion that the fibrillar structure and protein compos- roinflammation, pTau, and NFTs between typical and ition of different plaques may be relevant for the atypical AD. The relation between amyloid-beta, com- difference in regional vulnerability among AD pheno- plement, and microglia is underlined by a study in types. In addition, our results support a role for acti- APP transgenic mice deficient for C3 showing less vated microglia and complement factors in the cognitive problems and more amyloid-beta plaques atypical spreading of pathology in AD subtypes. In compared to APP mice not deficient for C3 [42]. The this study, we focused at 2 brain regions in a subset amyloid-beta plaques in the C3 knockout mice of atypical AD. It would be interesting to expand this showed less microglial co-localization. Together, these study on the role of neuroinflammation to various findings suggest that structural differences in amyloid brain regions in a larger atypical AD cohort. More deposits in atypical AD may directly be related to clinicopathological studies of AD variants (e.g., logo- complement and microglial activation. penic and behavioral frontal) are needed to better The distribution of NFT and pTau pathology is clearly understand the relation between amyloid-beta, pTau, associated with the presence of activated microglia in and related pathological mechanisms. Future research AD variants. Recent animal studies have shown that should focus on how variable disease mechanisms microglia are capable of both internalizing [43] and underlie the regional susceptibility in different brain excreting pTau [24, 44], suggesting that microglia con- regions leading to different clinical AD subtypes. tribute to the spreading of the pathology. Indeed, when mice are depleted for microglia, spreading of tau path- Abbreviations AD: Alzheimer’s disease; DAB: 3,3′-Diaminobenzidine tetrahydrochloride; ology is significantly reduced [23, 24]. In addition, acti- EOAD: Early-onset Alzheimer’s disease (< 65 years); FFPE: Formalin-fixed vated microglia could also contribute or induce tau paraffin-embedded; GFAP: Glial fibrillary acidic protein; hyperphosphorylation in neurons [45]. These studies in- IHC: Immunohistochemistry; IQR: Inter-quartile range; LOAD: Late-onset Alzheimer’s disease (≥ 65 years); NBB: Netherlands Brain Bank; dicate that microglial activation drives the spreading of NFT: Neurofibrillary tangle; PBS: Phosphate buffer saline; PCA: Posterior pathology and stimulates neurofibrillary degeneration. cortical atrophy; PMI: Post-mortem interval; pTau: Phosphorylated Tau; Interestingly, recent evidence from a pathological study ROI: Region of interest suggests that activation of microglia may precede tau Acknowledgements pathology in chronic traumatic encephalopathy [46], We would like to thank all brain donors and their caregivers for brain donation, which implicates that tau pathology may be a conse- the Netherlands Brain Bank and Michiel Kooreman for logistics and help in quence rather than a cause for microglial activation. selecting brain tissue samples, Martijn Heijmans and Wiesje van der Flier for advice on statistical analysis, Robert Veerhuis for help and advice on complement The aforementioned demographical differences in immunostainings, and Yolande Pijnenburg for clinical assessment. age and sex may influence the neuroinflammatory re- sponse. Former studies have shown various results on Funding the correlation between sex, age, and microglial acti- This study was funded by ZonMw grant number 70-73305-98-106. ZonMw had no role in the design of the study, collection, analysis, or interpretation vation in both humans and animal models. Schwarz of the data, or writing of the manuscript. et al. showed that during early development, male rats have more microglia within the parietal cortex Availability of data and materials compared to female rats. However, during juvenile The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. and early adulthood, this balance switches, indicating that sex hormones influence microglial colonization at Authors’ contributions different time points in rats [47]. In human studies, FB, JJMH, and BDCB designed the study. BDCB and JJMH coordinated the contradictory results are published for the effect of study and were responsible for writing the manuscript. BDCB, BL, and KNE age. In healthy controls, aging is correlated with a performed the experiments. BDCB analyzed the data. FB participated in writing the manuscript. PS made intellectual contribution and participated in more primed microglial state. Nevertheless, this was discussions. AJMR, WK, and the NBB were responsible for the autopsy shown to be different in diseased cases, in which material and neuropathological evaluation. All authors read and approved increasing age is associated with a diminished the final manuscript. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 15 of 16 Ethics approval and consent to participate NIH Public Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ Brain donors signed informed consent for autopsy and the use of tissue and 21460841. Cited 4 Oct 2017. medical records for research purposes. This study was approved by the 14. Hu X, Pickering E, Liu YC, Hall S, Fournier H, Katz E, et al. Meta-analysis for ethical committee of the VUmc. genome-wide association study identifies multiple variants at the BIN1 locus associated with late-onset Alzheimer’s disease. PLoS One. 2011;6:e16616. Competing interests Public Library of Science. Available from: http://www.ncbi.nlm.nih.gov/ The authors declare that they have no competing interests. pubmed/21390209. Cited 4 Oct 2017. 15. Sims R, van der Lee SJ, Naj AC, Bellenguez C, Badarinarayan N, Jakobsdottir J, et al. Rare coding variants in PLCG2, ABI3, and TREM2 implicate Publisher’sNote microglial-mediated innate immunity in Alzheimer’s disease. Nat Genet. Springer Nature remains neutral with regard to jurisdictional claims in 2017;49:1373–84. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ published maps and institutional affiliations. 28714976. Cited 4 Oct 2017. 16. Akiyama H, Barger S, Barnum S, Bradt B, Bauer J, Cole GM, et al. Received: 19 February 2018 Accepted: 26 April 2018 Inflammation and Alzheimer’s disease. Neurobiol Aging. 2000;21:383–421. NIH Public Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ 10858586. Cited 4 Oct 2017. References 17. Shen Y, Yang L, Li R. What does complement do in Alzheimer’s disease? Old 1. Dubois B, Feldman HH, Jacova C, Hampel H, Molinuevo JL, Blennow K, et al. molecules with new insights. Transl Neurodegener. 2013;2:21. BioMed Advancing research diagnostic criteria for Alzheimer’s disease: the IWG-2 Central. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24119446. criteria. Lancet Neurol. 2014;13:614–29. Available from: http://linkinghub. Cited 14 Apr 2017. elsevier.com/retrieve/pii/S1474442214700900. Cited 9 Aug 2017. 18. McGeer PL, Akiyama H, Itagaki S, EG MG. Activation of the classical 2. McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR, Kawas CH, complement pathway in brain tissue of Alzheimer patients. Neurosci Lett. et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations 1989;107:341–6. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ from the National Institute on Aging-Alzheimer’s Association workgroups 2559373. Cited 10 Aug 2017. on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 19. Eikelenboom P, Stam FC. Immunoglobulins and complement factors in 2011;7:263–9. NIH Public Access. Available from: senile plaques an immunoperoxidase study. Acta Neuropathol. 1982;57: http://www.ncbi.nlm.nih.gov/pubmed/21514250. Cited 9 Dec 2016. 239–42. Available from: https://link.springer.com/content/pdf/10.1007/ 3. Koedam ELGE, Lauffer V, van der Vlies AE, van der Flier WM, Scheltens P, BF00685397.pdf. Cited 11 Apr 2017. Pijnenburg YAL. Early-versus late-onset Alzheimer’s disease: more than age 20. Arends YM, Duyckaerts C, Rozemuller JM, Eikelenboom P, Hauw JJ. alone. J Alzheimers Dis. 2010;19:1401–8. Available from: http://www.ncbi. Microglia, amyloid and dementia in Alzheimer disease. A correlative study. nlm.nih.gov/pubmed/20061618. Cited 9 Dec 2016. Neurobiol Aging [Internet]. 21:39–47. Available from: http://www.ncbi.nlm. 4. Crutch SJ, Lehmann M, Schott JM, Rabinovici GD, Rossor MN, Fox NC. nih.gov/pubmed/10794847. Cited 9 Dec 2016. Posterior cortical atrophy. Lancet Neurol [Internet]. 2012;11:170–8. Available 21. Dickson DW, Farlo J, Davies P, Crystal H, Fuld P, Yen SH. Alzheimer’s disease. from: http://linkinghub.elsevier.com/retrieve/pii/S1474442211702897. Cited 9 A double-labeling immunohistochemical study of senile plaques. Am J Aug 2017. Pathol. 1988;132:86–101. Available from: http://www.ncbi.nlm.nih.gov/ 5. Gorno-Tempini ML, Hillis AE, Weintraub S, Kertesz A, Mendez M, Cappa SF, pubmed/2456021. Cited 9 Dec 2016. et al. Classification of primary progressive aphasia and its variants. 22. Li Y, Liu L, Barger SW, Griffin WST. Interleukin-1 mediates pathological Neurology. 2011;76:1006–14. American Academy of Neurology. Available effects of microglia on tau phosphorylation and on synaptophysin from: http://www.ncbi.nlm.nih.gov/pubmed/21325651. Cited 9 Aug 2017. synthesis in cortical neurons through a p38-MAPK pathway. J Neurosci. 6. Ossenkoppele R, Cohn-Sheehy BI, La Joie R, Vogel JW, Möller C, Lehmann 2003;23:1605–11. NIH Public Access. Available from: http://www.ncbi.nlm. M, et al. Atrophy patterns in early clinical stages across distinct phenotypes nih.gov/pubmed/12629164. Cited 10 Aug 2017. of Alzheimer’s disease. Hum Brain Mapp. 2015;36:4421–37. NIH Public 23. Maphis N, Xu G, Kokiko-Cochran ON, Jiang S, Cardona A, Ransohoff RM, Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26260856. et al. Reactive microglia drive tau pathology and contribute to the Cited 29 Dec 2017. spreading of pathological tau in the brain. Brain. 2015;138:1738–55. Oxford 7. Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. University Press. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ Acta Neuropathol. 1991;82:239–59. Available from: http://www.ncbi.nlm.nih. 25833819. Cited 9 Dec 2016. gov/pubmed/1759558. 24. Asai H, Ikezu S, Tsunoda S, Medalla M, Luebke J, Haydar T, et al. Depletion 8. Thal DR, Rüb U, Orantes M, Braak H. Phases of a beta-deposition in the of microglia and inhibition of exosome synthesis halt tau propagation. human brain and its relevance for the development of AD. Neurology. 2002; Nat Neurosci. 2015;18:1584–93. NIH Public Access. Available from: 58:1791–800. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ http://www.ncbi.nlm.nih.gov/pubmed/26436904. Cited 9 Dec 2016. 12084879. Cited 9 Dec 2016. 25. Hoozemans JJ, Rozemuller AJ, van Haastert ES, Eikelenboom P, van Gool 9. Tang-Wai DF, Graff-Radford NR, Boeve BF, Dickson DW, Parisi JE, Crook R, et WA. Neuroinflammation in Alzheimer’s disease wanes with age. al. Clinical, genetic, and neuropathologic characteristics of posterior cortical J Neuroinflammation. 2011;8:171. Available from: http://jneuroinflammation. atrophy. Neurology. 2004;63:1168–74. Available from: biomedcentral.com/articles/10.1186/1742-2094-8-171. Cited 9 Dec 2016. http://www.ncbi.nlm.nih.gov/pubmed/15477533. Cited 9 Dec 2016. 26. Mirra SS, Heyman A, McKeel D, Sumi SM, Crain BJ, Brownlee LM, et al. The 10. Galton CJ, Patterson K, Xuereb JH, Hodges JR. Atypical and typical Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part II. presentations of Alzheimer’s disease: a clinical, neuropsychological, Standardization of the neuropathologic assessment of Alzheimer’s disease. neuroimaging and pathological study of 13 cases. Brain. 2000;123 Pt Neurology. 1991;41:479–86. Available from: http://www.ncbi.nlm.nih.gov/ 3:484–98. Available from: http://www.ncbi.nlm.nih.gov/pubmed/10686172. pubmed/2011243. Cited 12 Dec 2016. Cited 9 Dec 2016. 27. Hoogendijk WJ, Sommer IE, Pool CW, Kamphorst W, Hofman MA, 11. Ossenkoppele R, Schonhaut DR, Schöll M, Lockhart SN, Ayakta N, Baker SL, Eikelenboom P, et al. Lack of association between depression and loss of et al. Tau PET patterns mirror clinical and neuroanatomical variability in neurons in the locus coeruleus in Alzheimer disease. Arch Gen Psychiatry. Alzheimer’s disease. Brain. 2016;139:1551–67. Available from: 1999;56:45–51. Available from: http://www.ncbi.nlm.nih.gov/pubmed/ http://www.ncbi.nlm.nih.gov/pubmed/26962052. Cited 9 Dec 2016. 9892255. Cited 30 Aug 2017. 12. Bertram L, Lange C, Mullin K, Parkinson M, Hsiao M, Hogan MF, et al. 28. Arendt T, Morawski M, Gärtner U, Fröhlich N, Schulze F, Wohmann N, et al. Genome-wide association analysis reveals putative Alzheimer’s disease Inhomogeneous distribution of Alzheimer pathology along the isocortical susceptibility loci in addition to APOE. Am J Hum Genet. 2008;83:623–32. relief. Are cortical convolutions an Achilles heel of evolution? Brain Pathol Elsevier. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18976728. [Internet]. 2016; Available from: http://doi.wiley.com/10.1111/bpa.12442. Cited 4 Oct 2017. Cited 9 Dec 2016. 13. Naj AC, Jun G, Beecham GW, Wang L-S, Vardarajan BN, Buros J, et al. 29. Walker L, McAleese KE, Johnson M, Khundakar AA, Erskine D, Thomas AJ, et al. Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are Quantitative neuropathology: an update on automated methodologies and associated with late-onset Alzheimer’s disease. Nat Genet. 2011;43:436–41. Boon et al. Journal of Neuroinflammation (2018) 15:170 Page 16 of 16 implications for large scale cohorts. J Neural Transm. 2017;124:671–83. Available servlet/aliasResolver?alias=iospress&doi=10.3233/JAD-150704. Cited 11 Apr from: http://www.ncbi.nlm.nih.gov/pubmed/28265813. Cited 30 Aug 2017. 2017. 30. Duyckaerts C, Delatour B, Potier M-C. Classification and basic pathology of 45. Bhaskar K, Konerth M, Kokiko-Cochran ON, Cardona A, Ransohoff RM, Lamb Alzheimer disease. Acta Neuropathol. 2009;118:5–36. Available from: BT. Regulation of tau pathology by the microglial fractalkine receptor. http://www.ncbi.nlm.nih.gov/pubmed/19381658. Cited 30 Aug 2017. Neuron. 2010;68:19–31. NIH Public Access. Available from: http://www.ncbi. nlm.nih.gov/pubmed/20920788. Cited 20 Dec 2017. 31. Murray ME, Graff-Radford NR, Ross OA, Petersen RC, Duara R, Dickson DW. 46. Cherry JD, Tripodis Y, Alvarez VE, Huber B, Kiernan PT, Daneshvar DH, et al. Neuropathologically defined subtypes of Alzheimer’s disease with distinct clinical Microglial neuroinflammation contributes to tau accumulation in chronic characteristics: a retrospective study. Lancet Neurol. 2011;10:785–96. Available traumatic encephalopathy. Acta Neuropathol Commun. 2016;4:112. from: http://www.ncbi.nlm.nih.gov/pubmed/21802369. Cited 9 Dec 2016. Available from: http://actaneurocomms.biomedcentral.com/articles/10.1186/ 32. Bigio EH, Hynan LS, Sontag E, Satumtira S, White CL. Synapse loss is greater s40478-016-0382-8. Cited 30 Aug 2017. in presenile than senile onset Alzheimer disease: implications for the 47. Schwarz JM, Sholar PW, Bilbo SD. Sex differences in microglial colonization cognitive reserve hypothesis. Neuropathol Appl Neurobiol. 2002;28:218–27. of the developing rat brain. J Neurochem. 2012;120:948–63. NIH Public Blackwell Science Ltd. available from: http://doi.wiley.com/10.1046/j.1365- Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22182318. 2990.2002.00385.x. Cited 26 Oct 2017. Cited 27 Mar 2018. 33. Josephs KA, Dickson DW, Murray ME, Senjem ML, Parisi JE, Petersen RC, et al. Quantitative neurofibrillary tangle density and brain volumetric MRI analyses in Alzheimer’s disease presenting as logopenic progressive aphasia. Brain Lang. 2013;127:127–34. NIH Public Access. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23541297. Cited 27 Mar 2018. 34. Snyder HM, Asthana S, Bain L, Brinton R, Craft S, Dubal DB, et al. Sex biology contributions to vulnerability to Alzheimer’s disease: a think tank convened by the Women’s Alzheimer’s Research Initiative. Alzheimers Dement. 2016; 12:1186–96. Available from: http://linkinghub.elsevier.com/retrieve/pii/ S1552526016328746. Cited 14 Apr 2017. 35. Beaufils E, Ribeiro MJ, Vierron E, Vercouillie J, Dufour-Rainfray D, Cottier J-P, et al. The pattern of brain amyloid load in posterior cortical atrophy using (18)F-AV45: is amyloid the principal actor in the disease? Dement Geriatr Cogn Dis Extra. 2014;4:431–41. Karger Publishers. Available from: http://www.ncbi.nlm.nih.gov/pubmed/25538727. Cited 12 Apr 2017. 36. Wang X-D, Lu H, Shi Z, Cai L, Liu S, Liu S, et al. A pilot study on clinical and neuroimaging characteristics of Chinese posterior cortical atrophy: comparison with typical Alzheimer’s disease. PLoS One. 2015;10:e0134956. Public Library of Science, Available from: http://www.ncbi.nlm.nih.gov/ pubmed/26267071. Cited 15 Dec 2016. 37. de Souza LC, Corlier F, Habert M-O, Uspenskaya O, Maroy R, Lamari F, et al. Similar amyloid- burden in posterior cortical atrophy and Alzheimer’s disease. Brain. 2011;134:2036–43. Available from: https://www.ncbi.nlm.nih. gov/pubmed/?term=de+Souza+LC%2C+Corlier+F%2C+Habert+M-O%2C. Cited 2016 Dec 9. 38. Hof PR, Vogt BA, Bouras C, Morrison JH. Atypical form of Alzheimer’s disease with prominent posterior cortical atrophy: a review of lesion distribution and circuit disconnection in cortical visual pathways. Vis Res. 1997;37:3609– 25. Available from: http://www.sciencedirect.com.vu-nl.idm.oclc.org/science/ article/pii/S0042698996002404. Cited 12 Apr 2017. 39. Zanjani H, Finch CE, Kemper C, Atkinson J, McKeel D, Morris JC, et al. Complement activation in very early Alzheimer disease. Alzheimer Dis Assoc Disord [Internet]. 19:55–66. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/15942322. Cited 14 Apr 2017. 40. Wyss-Coray T, Yan F, Lin AH-T, Lambris JD, Alexander JJ, Quigg RJ, et al. Prominent neurodegeneration and increased plaque formation in complement-inhibited Alzheimer’s mice. Proc Natl Acad Sci U S A. 2002;99: 10837–42. Available from: http://www.pnas.org/cgi/doi/10.1073/pnas. 162350199. Cited 14 Apr 2017. 41. Maier M, Peng Y, Jiang L, Seabrook TJ, Carroll MC, Lemere CA. Complement C3 deficiency leads to accelerated amyloid beta plaque deposition and neurodegeneration and modulation of the microglia/macrophage phenotype in amyloid precursor protein transgenic mice. J Neurosci. 2008; 28:6333–41. NIH Public Access. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/18562603. Cited 14 Apr 2017. 42. Shi Q, Chowdhury S, Ma R, Le KX, Hong S, Caldarone BJ, et al. Complement C3 deficiency protects against neurodegeneration in aged plaque-rich APP/ PS1 mice. Sci Transl Med. 2017;9:eaaf6295. Available from: http://www.ncbi. nlm.nih.gov/pubmed/28566429. Cited 30 Aug 2017. 43. Luo W, Liu W, Hu X, Hanna M, Caravaca A, Paul SM. Microglial internalization and degradation of pathological tau is enhanced by an anti- tau monoclonal antibody. Sci Rep. 2015;5:11161. Available from: http://www.nature.com/articles/srep11161. Cited 30 Aug 2017. 44. Bolós M, Llorens-Martín M, Jurado-Arjona J, Hernández F, Rábano A, Avila J. Direct evidence of internalization of tau by microglia in vitro and in vivo. J Alzheimers Dis. 2015;50:77–87. Available from: http://www.medra.org/

Journal

Journal of NeuroinflammationSpringer Journals

Published: May 29, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

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

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

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.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off