Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 https://doi.org/10.1186/s12974-021-02169-0 RESEARCH Open Access Cognitive and brain cytokine profile of non-demented individuals with cerebral amyloid-beta deposition Lisi Flores-Aguilar1, M. Florencia Iulita2, Chiara Orciani3, Neil Tanna4, Jingyun Yang5,6, David A. Bennett5,6 and A. Claudio Cuello1,3,4,7* Abstract Background: Brain inflammation has been increasingly associated with early amyloid accumulation in Alzheimer’s disease models; however, evidence of its occurrence in humans remains scarce. To elucidate whether amyloid deposition is associated with neuroinflammation and cognitive deficits, we studied brain inflammatory cytokine expression and cognitive decline in non-demented elderly individuals with and without cerebral amyloid-beta deposition. Methods: Global cognition, episodic, working, and semantic memory, perceptual speed, visuospatial ability, and longitudinal decline (5.7 ± 3.6 years) in each cognitive domain were compared between elderly individuals (66–79 years) with and without cerebral amyloid-beta deposition. The expression of 20 inflammatory cytokines was analyzed in frozen temporal, parietal, and frontal cortices and compared between older individuals with and without amyloid-beta deposition in each brain region. Correlation analyses were performed to analyze associations between amyloid-beta load, cytokine expression, and cognitive decline. Results: Individuals with cortical amyloid-beta deposition displayed deficits and a faster rate of cognitive decline in perceptual speed as compared with those individuals without amyloid-beta. This decline was positively associated with cortical amyloid-beta levels. Elderly individuals with amyloid-beta deposition had higher levels of IL-1β, IL-6, and eotaxin-3 in the temporal cortex accompanied by an increase in MCP-1 and IL-1β in the parietal cortex and a trend towards higher levels of IL-1β and MCP-1 in the frontal cortex as compared with age-matched amyloid-free individuals. Brain IL-1β levels displayed a positive association with cortical amyloid burden in each brain region. Finally, differential cytokine expression in each cortical region was associated with cognitive decline. Conclusions: Elderly individuals with amyloid-beta neuropathology but no symptomatic manifestation of dementia, exhibit cognitive decline and increased brain cytokine expression. Such observations suggest that increased cytokine expression might be an early event in the Alzheimer’s continuum. Keywords: Amyloid-beta, Cognition, Cytokines, Alzheimer’s disease, Neuroinflammation, Perceptual speed, No cognitive impairment * Correspondence: claudio.cuello@mcgill.ca 1 Department of Anatomy and Cell Biology, McGill University, Montreal, Quebec, Canada 3 Department of Neurology and Neurosurgery, McGill University, Montreal, Canada Full list of author information is available at the end of the article © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 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 in a credit line to the data. Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 2 of 13 Background temporal (n = 28), parietal (n = 26), and frontal (n = 27) Mounting evidence from human and animal studies cortices from individuals under age 80 who had been indicates that neuroinflammation plays an important clinically diagnosed as NCI at the last neuropsycho- role in the development and progression of Alzheimer’s logical evaluation before death [14]. We excluded cases disease (AD). Specifically, epidemiological studies revealed over age 80 to avoid for potentially confounding effects that cognitively healthy individuals under long-term ad- of ‘SuperAgers’, a group that displays less prominent ministration of anti-inflammatories have a reduced risk of Alzheimer’s-related changes [15]. developing symptomatic AD [1, 2]. The benefit of early anti-inflammatory treatment suggests that a detrimental Neuropathological assessment neuroinflammatory process is active at preclinical AD Neuropathological examination of each case was stages [1, 3, 4]. Indeed, studies in transgenic models of performed by an experienced pathologist blinded to the amyloid pathology have demonstrated the presence of a clinical diagnosis at the Rush Alzheimer’s Disease disease-aggravating neuroinflammatory process concomi- Center. A total of 8 brain regions were analyzed per case tant with the intraneuronal accumulation of amyloid-beta (hippocampus, entorhinal cortex, midfrontal cortex, in- (Aβ) oligomers [5–9]. While an incipient inflammatory ferior temporal, angular gyrus, calcarine cortex, anterior process has been well characterized in AD animal models cingulate cortex, and superior frontal cortex). Aβ values and has been associated with the abnormal deposition of for each brain region represent the percentage covered Aβ, evidence from human studies remains scarce. by 4G8 immunoreactivity in 1–2 sections per brain re- In the general population, it has been estimated that gion [10]. The global Aβ value represents the average of about one-third of older non-cognitively impaired indi- 4G8 immunoreactivity across all cortical areas analyzed. viduals (NCI) display brain AD neuropathology [10–13]. Neurofibrillary tangle pathology was determined by the Given that aging is the major risk factor for AD and that stereological assessment of AT8 immunoreactivity across AD has a long, pre-symptomatic phase where neuro- the entorhinal cortex, CA1, superior frontal cortex, mid pathology starts to ‘silently’ build-up; it is then expected frontal cortex, inferior temporal cortex, angular gyrus, that a proportion of NCI individuals with AD neuro- cingulate gyrus, and calcarine cortex. Final values repre- pathology should be at the earliest, asymptomatic stages sent cortical AT8 density per mm2 [10]. Additional of AD. neuropathological and genetic information was available, Towards elucidating whether incipient AD pathology including the modified Consortium to Establish a Registry is associated with inflammatory marker expression and for Alzheimer’s Disease (CERAD), modified National In- cognitive decline, we analyzed cognitive changes and stitute on Aging (NIA)-Reagan, Braak scores, and APOE pro-inflammatory cytokine expression in the temporal, genotype [10]. Modified CERAD, NIA-Reagan scores, and parietal, and frontal cortices of a clinically and neuro- Braak scores were based on counts of neuritic plaques pathologically characterized cohort of non-demented and neurofibrillary tangles according to Bielschowsky elderly individuals stratified according to the presence or silver staining. Demographic details are illustrated in absence of cerebral Aβ. Table 1 and in Additional files 1, 2 and 3, Supplemental Our results indicate that non-demented individuals Tables 1-3. harboring amyloid pathology display impairments in perceptual speed accompanied by an increase in key Immunohistochemical detection and image acquisition of cerebral pro-inflammatory markers across different brain amyloid and tau pathologies in the temporal, parietal, regions, strengthening the evidence for an early dysregu- and frontal cortices lation of inflammatory markers at presumable preclinical Aβ (4G8 immunoreactivity) and tau neuropathology stages of AD. (AT8 immunoreactivity) was assessed as described in [10] at the Rush Alzheimer’s Disease Center. Z-stack Methods Brightfield images of cortical laminae III and V from the Brain tissue temporal, parietal, and frontal cortices were taken with a A total of 81 brain tissue samples were obtained from Zeiss AxioImager M2 Imaging microscope equipped American old clergy members and community-dwelling with an AxioCam 506 color digital camera (Zeiss, persons involved in the Religious Orders Study (ROS) Canada), and Zeiss ZenPro software v.2.3 (Zeiss, and Rush Memory and Aging Project (MAP) [10]. Canada) was used. One representative brain section per Briefly, ROSMAP comprises community cohorts of region per case was imaged. aging and AD that enroll cognitively healthy individuals who agree to an annual detailed cognitive evaluation and Cognitive assessment brain donation at the time of death. For this study, we Global cognitive function was calculated using a battery randomly selected post-mortem brain tissue from the of 19 cognitive tests to assess five domains of cognitive Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 3 of 13 Table 1 Demographics of all the study cases with or without global Aβ deposition Group gAβ - gAβ + p-value Sample size n = 14 n = 16 Age range (years) 66.21 – 79.53 67.37 – 79.63 mean ± SD 75.83 ± 3.605 75.32 ± 3.708 0.821 Sex (F/M) 6/8 9/7 0.715 PMI range 3 - 29 2.5 – 29.58 mean ± SD 11.56 ± 7.238 8.748 ± 7.803 0.07 Cogn global mean ± SD 0.4162 ± 0.3604 0.3103 ± 0.3535 0.423 Years of education mean ± SD 18.29 ± 3.832 17.56 ± 3.245 0.58 Apoe4 0 5 (31.25 %) 0.044 ApoE distribution ε 2/2 0 0 ε 2/3 2 (14.28 %) 2 (12.5 %) ε 2/4 0 0 ε 3/3 12 (85.71%) 9 (56.25 %) ε 3/4 0 5 (31.25 %) ε 4/4 0 0 Braak score 0 – II 9 (64.29 %) 7 (43.75 %) 0.298 III - IV 5 (35.71 %) 9 (56.25 %) Braak score distribution 0 1 (7.14 %) 0 I 4 (28.57 %) 4 (25%) II 4 (28.57 %) 3 (18.75 %) III 4 (28.57 %) 5 (31.25 %) IV 1 (7.14 %) 4 (25 %) V 0 0 VI 0 0 CERAD possible or no AD 14 (100 %) 6 (37.5 %) 0.0003 probable or definite AD 0 10 (62.5 %) CERAD distribution no AD 13 (92.86 %) 5 (31.25 %) possible AD 1 (7.14 %) 1 (6.25 %) probable AD 0 9 (56.25 %) definite AD 0 1 (6.25 %) NIA-Reagan low or no likelihood 14 (100%) 9 (56.25%) intermediate/high likelihood 0 7 (43.75%) NIA-Reagan distribution no likelihood 1 (7.14%) 0 0.0073 low likelihood 13 (92.85%) 9 (56.25%) intermediate likelihood 0 7 (43.75%) high likelihood 0 0 p-tau mean ± SD 1.946 ± 2.31 1.274 ± 1.270 0.918 Aβ-IR mean ± SD 0 1.962 ± 1.946 Abbreviations: PMI post-mortem interval, Cogn cognition, gAβ global amyloid beta, F female, M male, IR immunoreactivity, SD standard deviation. Data are presented as mean ± SD function: (i) episodic memory (word list memory, recall span forward and backward, and digit ordering tests), and recognition, east Boston story immediate and de- (iv) perceptual speed (symbol digits modality, number layed, and logical memory Ia immediate and IIa delayed comparison, and Stroop color naming and reading), and tests), (ii) semantic memory (Boston naming, verbal flu- (v) visuospatial ability (judgment of line orientation, and ency, and reading tests), (iii) working memory (digit standard progressive matrices tests). A composite score Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 4 of 13 of each cognitive domain was obtained by converting A, (plate 2) and Eotaxin-3, Interferon-inducible protein raw scores to z scores using the baseline mean and (IP)-10, Macrophage inflammatory protein (MIP)-1β, standard deviation of all persons in the ROSMAP stud- Monocyte chemoattractant protein (MCP)-1, ies. The z scores were averaged to yield a final composite Macrophage-derived chemokine (MDC), Thymus and score representing global cognitive function [10]. activation-regulated chemokine (TARC) in plate 3. Sam- ples were normalized to 3 μg/μl in lysis buffer with 1% Estimated slopes from the random-effects model for each blocker A (Meso-Scale Discovery) and 1× of protease cognitive domain and phosphatase inhibitors. Samples were loaded in du- A linear mixed-effects model with each cognitive do- plicates according to the manufacturer’s recommenda- main (episodic memory, semantic memory, working tions and averaged for analysis. The expression levels of memory, perceptual speed, and visuospatial ability) as each cytokine were fitted to its corresponding standard the longitudinal outcome was utilized to estimate the curve. Signal was measured with a SECTOR Imager person-specific rate of change in each cognitive domain 6000 reader (Meso-Scale Discovery), and data were over time (random slope) [16]. The mean time of follow- processed with the Discovery Workbench 4.0 software. up for the cognitive evaluation was 5.7 ± 3.6 years (inter- VEGF-A signal was below the lower limit of detection, quartile range: 3–9 years). The statistical model controls and consequently, there is no data presented in this for age at baseline, sex, and years of formal education. study. Cognitive differences in domain-specific cognition, and the rate of cognitive decline were analyzed within the Statistical analysis NCI population. As various brain regions are involved in Data are presented in boxplots where the median is rep- the different cognitive domains that were assessed, we resented by the horizontal line and the whiskers repre- stratified the NCI population according to their global sent the 25th and 75th percentiles. Two-tailed Student’s Aβ deposition score. Thus, those with positive values of or Mann–Whitney tests were used for two-group com- global Aβ deposition were classified as gAβ+ and those parisons, according to data normality analysis. Data without Aβ deposition were classified as gAβ− (Table 1). points above or under 1.5 times the interquartile range were considered as outliers and excluded from analyses. Segregation of NCI individuals according to Aβ pathology For three-group comparisons, one-way ANOVA or NCI individuals were segregated into two groups Kruskal–Wallis tests were used followed by Tukey or depending on the presence or absence of Aβ in each Dunn’s post hoc tests, respectively. Fisher’s exact test specific cortical area that was analyzed (temporal, was used to assess differences in sex, APOE genotype, parietal, and frontal cortices). Therefore, those who had Braak score, CERAD, and NIA-Reagan criteria. Pearson’s zero amyloid immunoreactivity in the examined cortical or Spearman’s correlation coefficients were assessed in area (Additional files 1, 2 and 3, Supplemental Tables correlation analyses. For correlation and partial correl- 1-3) were classified as Aβ−, whereas those harboring ation analyses, data were log-transformed. Partial correl- Aβ deposition were classified as Aβ+. ation analyses of cytokine expression and Aβ or p-tau were controlled for post-mortem interval. Linear regres- Brain cytokine analysis sion analyses of cognitive decline and cytokine expression Fifty milligrams of frozen brain tissue per case was were adjusted for age at death and years of education. No homogenized in ice-cooled lysis buffer (50 mM Tris, 150 correction for multiple comparisons was performed. Data mM NaCl, 0.05% Tween-20, 5 μl per mg of tissue) with were analyzed with Graph Pad Prism 8. Partial correlation 2× phosphatase (PhosStop, Roche) and protease and linear regression analyses were performed using R inhibitors (Complete Mini, Roche). Lysed samples were version 4.0.4. Significance was set at p < 0.05. centrifuged at 13,000 rpm for 45 min at 4 °C. To avoid repeated freeze and thaw cycles, supernatants were ali- Results quoted and stored at – 80 °C. Protein concentration was Demographic and neuropathological characteristics of measured using the DCTM Protein Assay Kit (BioRad). elderly individuals aged 66 to 79 years with or without Inflammatory protein expression was measured in brain Aβ deposition are illustrated in Table 1 and Additional homogenates using an electrochemiluminescence-linked files 1, 2 and 3, Supplemental Tables 1-3. Representative multiplex immunoassay (Meso-Scale Discovery). Three images of Aβ and tau histopathology are illustrated in different plates were customized to measure the follow- Fig. 1 and in Additional file 4, Supplemental Figure 1. ing inflammatory proteins: Interleukin (IL)-1β, IL-4, IL- No differences in global cognition, age, sex, years of 6, IL-8, IL-10, IL-13, Tumor necrosis factor (TNF)α, education, tau pathology, and post-mortem interval Interferon (IFN)γ, (plate 1), IL-1α, IL-5, IL-7, IL-12p40, were found between groups. The APOE e4 allele was IL-15, IL-16, Vascular endothelial growth factor (VEGF)- enriched in cognitively normal individuals with global Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 5 of 13 Fig. 1 Alzheimer’s disease neuropathology in the temporal, parietal, and frontal cortices of elderly individuals. Representative micrographs depicting amyloid-beta and tau pathology as revealed by immunohistochemistry (IHC) with 4G8 and AT8 antibodies, respectively. a Representative micrograph of a case without Aβ deposition in the temporal, parietal, and frontal cortices. b Representative micrograph of a case with Aβ deposition in the temporal, parietal, and frontal cortices. CV cresyl violet counterstaining. Scale bar = 100 μm Aβ deposition (p = 0.044). Individuals with cerebral Differential upregulation of inflammatory mediators Aβ met a neuropathological diagnosis of AD accord- across brain regions in elderly individuals with Aβ ing to the modified NIA-Reagan (p = 0.007) and pathology and its association with cognitive decline CERAD (p = 0.0003) criteria. Cognitive decline and Inflammatory protein expression was analyzed in the brain inflammatory protein expression were assessed temporal, parietal, and frontal cortices of individuals in elderly individuals with no symptomatic manifest- grouped by the presence or absence of Aβ pathology ation of AD. (Aβ+ and Aβ−, respectively) in each cortical brain area (Additional files 1-3, Supplemental Tables 1-3). Given that some cases displayed long post-mortem intervals, Elderly individuals with Aβ deposition display faster the correlation of cytokine expression in each brain re- cognitive decline gion with the corresponding post-mortem interval was Although elderly individuals harboring global Aβ depos- analyzed. IL-4 and TNF-α levels in the temporal cortex ition displayed a similar global cognitive score to those and eotaxin-3 and IL-4 expression in the parietal cortex, without cerebral Aβ (Fig. 2a–b), those with global Aβ negatively correlated with the post-mortem interval deposition displayed deficits in perceptual speed (p = (Additional file 6, Supplemental Table 5). No associations 0.012) (Fig. 2c). When assessing the person-specific rate between cytokine expression and post-mortem interval of change (i.e., cognitive decline slope), a higher rate of were found in the frontal cortex (Additional file 6, cognitive decline in perceptual speed (p = 0.04) was ob- Supplemental Table 5). served in individuals with Aβ deposition compared with In the temporal cortex, IL-1β (p = 0.024), IL-6 (p = those free of amyloid (Fig. 2d). No other cognitive do- 0.017) and eotaxin-3 (p = 0.001) were significantly ele- mains differed between groups (Fig. 2e–l). Notably, a vated in Aβ+ individuals when compared to those without faster decline in perceptual speed was associated with Aβ deposition (Fig. 3a, b, d). Furthermore, the expression higher Aβ deposition (p = 0.047), and no associations levels of IL-1β and IL-6 positively correlated to the extent were found among cognitive scores and global tau de- of Aβ deposition in this brain region (Additional file 7, position (Additional file 5, Supplemental Table 4). Supplemental Table 6). No associations between tau Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 6 of 13 Fig. 2 Elderly with Aβ deposition display cognitive deficits. a–l Cognitive scores and estimated slopes reflecting the person-specific rate of cognitive decline within a mean follow-up period of 5.7 ± 3.6 years (interquartile range: 3–9 years). c–d Elderly individuals with Aβ deposition display deficits in perceptual speed and a faster rate of cognitive decline in this domain, t (28) = 2.68, p = 0.012 and t (26) = 2.15, p = 0.04, respectively. a–b and e–l Global cognition, working memory, episodic memory, semantic memory, and visuospatial ability did not differ between elder individuals with or without Aβ deposition. Data are displayed in boxplots where the median is represented by the horizontal line and the whiskers represent the 25th and 75th percentiles; a, c–f, and h–l two-tailed Student t test; b, g two-tailed Mann–Whitney test; Aβ−, n = 14; Aβ+, n = 16; g global. Full circles represent outliers. *p < 0.05 deposition and cytokine expression were found in the In the parietal cortex, higher MIP-1β and TARC levels temporal cortex (Additional file 7, Supplemental Table 6). were associated with a lower rate of cognitive decline in Cognitive decline in episodic, semantic, and working working and episodic memory, respectively (Additional memory was associated with higher IL-1α, IL-13, file 8, Supplemental Table 7). and MDC levels in the temporal cortex, respectively Changes in inflammatory protein expression followed (Additional file 8, Supplemental Table 7). a similar pattern in the frontal cortex, revealing trends Individuals with Aβ positivity in the parietal cortex toward higher levels of IL-1β, IL-6, and MCP-1 in non- displayed higher levels of the pro-inflammatory demented individuals with Aβ deposition (p = 0.058, p = molecules IL-1β (p = 0.022) and MCP-1 (p = 0.042) 0.060, p =0.052, respectively) (Fig. 5 A–C). Moreover, (Figure 4a, c). Notably, Aβ deposition positively cor- IL-1β, IL-16, and IL-13 expression positively correlated related with IL-1β and MDC levels (Additional file 7, with Aβ deposition (Additional file 7, Supplemental Supplemental Table 6). Tau deposition in the par- Table 6). Cytokine expression levels did not correlate ietal cortex negatively correlated with IL-10 and with tau deposition in the frontal cortex (Additional file TNF-α levels and was positively associated with 7, Supplemental Table 6). MDC expression levels (Additional file 7, Supple- Finally, better cognitive scores in visuospatial orienta- mental Table 6). tion and perceptual speed were associated with higher Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 7 of 13 Fig. 3 Upregulation of inflammatory cytokines in the temporal cortex of Aβ+ individuals. a–s Cytokine expression assessed by multispot immunoassays revealed increased levels (dotted line) of a IL-1β, two-tailed Student t test; t (24) = 2.403, p = 0.024; Aβ−, n = 14; Aβ+, n = 12; b IL- 6, two-tailed Student t test; t (24) = 2.56, p = 0.017; Aβ−, n = 14; Aβ+, n = 12 and D eotaxin-3, two-tailed Student t test; t (22) = 3.594, p = 0.001; Aβ−, n = 14; Aβ+, n = 10, in the temporal cortex of non-demented individuals with Aβ deposition as compared with individuals without Aβ deposition in this brain region. c, e–s Expression of other cytokines (MCP-1, IL-1α, IL-4, IL-7, IL-8, IL-10, IL-12p40, IL-13, IL-15, IL-16, IFNγ, IP-10, MDC, MIP-1β, TARC, and TNFα) did not differ between Aβ− and Aβ+ individuals. Data are displayed in boxplots where the median is represented by the horizontal line, and the whiskers represent the 25th and 75th percentiles. *p < 0.05 IL-16 and IP-10 levels in the frontal cortex (Additional from elderly individuals with Aβ deposition (Fig. 7), re- file 8, Supplemental Table 7). inforcing the notion of an early, preclinical neuroinflam- matory process in the continuum of the AD pathology Neuropathological differences among cortical regions of [1, 3, 4]. elderly individuals with and without Aβ deposition Notably, non-demented individuals with Aβ pathology Given the elevation of key inflammatory cytokines ob- had an overall increase in IL-1β in the three brain re- served in the temporal, parietal, and frontal cortices ana- gions analyzed. Thus, raising the possibility that IL-1β lyzed, we examined whether there were differences in might be one of the earliest inflammatory markers asso- Aβ and p-tau deposition between these brain regions. ciated with the accumulation of Aβ neuropathology in No significant differences in Aβ deposition were ob- older adults. IL-1β overexpression might contribute to served between the temporal, frontal, and parietal corti- potentiate AD neuropathology [19–21]. Early involve- ces of Aβ+ individuals (p = 0.780) (Fig. 6a). In contrast, ment of IL-1β in the AD continuum is also suggested by the temporal cortex of elderly individuals with or with- its presence in diffuse amyloid plaques in AD brains [22] out Aβ deposition displayed higher levels of p-tau than and increased expression of caspase-1, which promotes the parietal and frontal cortices (p < .0001) (Fig. 6b). IL-1β maturation, at prodromal and clinical stages of AD [23]. Moreover, IL-1β upregulation has been re- Discussion ported before the establishment of an overt AD path- In this study, we show that non-demented individuals ology in Down syndrome [24, 25]. Increased neuronal with cortical Aβ deposition display cognitive impair- IL-1β expression is present in a transgenic rat model of ments in specific cognitive domains, such as perceptual amyloidosis before amyloid plaque deposition [5] . speed, suggesting that such cognitive domain might be IL-1β expression may lead to the translation of IL-6 disrupted at early stages in the AD continuum. This is in [26]. In our study, IL-6 levels were likewise increased in line with previous studies reporting the occurrence of the temporal cortex of non-demented individuals with cognitive impairments in NCI individuals with Aβ path- Aβ pathology. Similar to IL-1β, IL-6 is also present in ology [17, 18]. Our studies also underscore that global diffuse amyloid plaques in AD brains [27] and can en- cognitive scores might mask subtle cognitive domain- hance AD pathology and cognitive deficits in animal specific deficits. Furthermore, key inflammatory cyto- models [28–30]. Importantly, genetic variations of the kines were differentially elevated across brain regions IL-6 and IL-1β genes have been associated with an Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 8 of 13 Fig. 4 Increased expression of inflammatory cytokines in the parietal cortex of Aβ+ individuals. a–s Cytokine expression assessed by multispot immunoassays revealed increased levels (dotted line) of a IL-1β, two-tailed Student t test; t (22) = 2.46, p =0.022; Aβ−, n = 12; Aβ+, n = 12; c MCP-1, two-tailed Mann–Whitney test; U = 29, p = 0.042; Aβ−, n = 12; Aβ+, n = 10, in the parietal cortex of non-demented individuals with Aβ deposition as compared with individuals without Aβ deposition in this brain region. b, d–s Expression cytokine levels that did not differ between Aβ− and Aβ+ individuals (IL-6, eotaxin-3, IL-1α, IL-4, IL-7, IL-8, IL-10, IL-12p40, IL-13, IL-15, IL-16, IFNγ, IP-10, MDC, MIP-1β, TARC, and TNFα). Data are displayed in boxplots where the median is represented by the horizontal line and the whiskers represent the 25th and 75th percentiles. *p <0.05 increased risk of AD [31]. Of note, although IL-1β and and frontal cortices. Therefore, higher eotaxin-3 levels IL-6 expression is often accompanied by increased TNF- might reflect an immune response to pathological tau in α levels [32], we did not observe higher TNF-α levels in elders with amyloid deposition. brains with Aβ deposition. In addition, the relatively low Increased levels of MCP-1 were found in the parietal levels of TNF-α measured in this study, might mask sub- cortex of individuals with Aβ. MCP-1 is a potent mono- tle expression differences between brains with and with- cyte chemoattractant, and its upregulation in transgenic out Aβ deposition. Therefore, more sensitive assays are models of AD was suggested to contribute to peripheral needed to confirm whether TNF-α becomes dysregu- monocyte infiltration into the brain [36, 37]. MCP-1 up- lated at early AD stages. regulation has been reported in AD brains and body The chemokine eotaxin-3 was also elevated in the fluids [38, 39]. Notably, MCP-1 neuronal levels are temporal cortex of AD-asymptomatic individuals with upregulated before the development of amyloid plaques Aβ pathology. Increased eotaxin-3 levels have been re- in a transgenic rat model of the AD-like pathology and ported in the CSF of individuals with AD dementia and in Aβ-burdened neurons in the human cerebral cortex can distinguish AD from other neurodegenerative dis- [9], supporting its early involvement at incipient stages eases [33–35]. Moreover, eotaxin-3 levels are elevated of amyloid accumulation. before the establishment of overt AD pathology in the A trend towards higher levels of IL-1β, IL-6, and brains of people with Down syndrome [25]. It has also MCP-1 levels was observed in the frontal cortex of Aβ+ been reported that eotaxin-3 levels correlate with CSF individuals. This suggests that altered inflammatory pro- tau pathology in AD [33]. Interestingly, in our study, tein expression in the frontal cortex might occur after its eotaxin-3 was only upregulated in the temporal cortex exacerbation in the temporal and parietal cortices. It is of NCI individuals with Aβ pathology, and such region also possible that the small sample size posed a limita- displayed higher levels of tau pathology than the parietal tion to detect significant differences in this cortical Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 9 of 13 Fig. 5 Increased cytokine expression in the frontal cortex of Aβ+ individuals. Cytokine expression assessed by multispot immunoassays revealed a trend towards increased levels of a IL-1β, two-tailed Student t test; t (17) = 2.028, p = 0.058; Aβ−, n = 9; Aβ+, n = 10; b IL-6, two-tailed Student t test; t (22) = 1.981, p = 0.06; Aβ−, n = 11; Aβ+, n = 13; and c MCP-1, two-tailed Student t test; t (19) = 2.067, p = 0.052; Aβ−, n = 9; Aβ+, n = 12, in the frontal cortex of non-demented individuals with Aβ deposition as compared with individuals without Aβ deposition in this brain region. d–s Expression cytokine levels did not differ between Aβ− and Aβ+ individuals (Eotaxin-3, IL-1α, IL-4, IL-7, IL-8, IL-10, IL-12p40, IL-13, IL-15, IL-16, IFNγ, IP-10, MDC, MIP-1β, TARC, and TNFα). Data are displayed in boxplots where the median is represented by the horizontal line and the whiskers represent the 25th and 75th percentiles region. Therefore, these analyses remain exploratory and microglial activation, comparable to those found in await to be confirmed in a larger and independent young individuals [44]. Besides the expression of inflam- dataset. matory cytokines, other pathological events have been Our study also showed that although amyloid levels reported to occur in the brains of cognitively normal in- did not significantly differ among brain regions, the dividuals harboring AD neuropathology such as synaptic temporal cortex displayed higher tau deposition and in- deficits [45], oxidative stress [45], growth factor dysregu- flammatory cytokine expression than the parietal and lation [46], and epigenetic modifications [47]. Further frontal cortices. Therefore, it might be possible that the experiments are needed to understand whether these combined tau and amyloid pathologies make the and the previously mentioned inflammatory changes are temporal cortex a more vulnerable region to increased also present in brains with or without Aβ pathology. inflammatory protein expression. In this study, none of the upregulated cytokines in the In accordance with our results, cognitively normal in- temporal, parietal, or frontal cortices was associated with dividuals with CSF Aβ pathological values displayed an the rate of cognitive decline. It is important to note that upregulation of CSF inflammatory and cerebrovascular longitudinal cognitive measures and cytokine brain ex- markers compared with NCI individuals with normal pression analyses are disconnected in the time of assess- CSF Aβ values [40]. In addition, astroglial and microglial ment. Therefore, this type of analyses would have more activation are already present before the onset of demen- informative value if cytokine expression was measured tia in populations at genetic risk of AD [25, 41]. Import- in cerebrospinal fluid followed by post-mortem assess- antly, AD-resilient brains display decreased microglia ment of cytokine brain expression at the time of death. and astroglia activation and anti-inflammatory and pro- Surprisingly, our results also show that some cytokines resolution cytokine signature [42, 43]. Moreover, that display similar expression levels in brains with and “SuperAgers” brains also display lower levels of without Aβ deposition, correlate either positively or Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 10 of 13 Fig. 6 Aβ and tau deposition in the cortices of non-demented individuals with and without Aβ pathology. a Aβ deposition did not differ between the temporal, parietal, and frontal cortices of non-demented individuals harboring AD pathology; Kruskal–Wallis test, H = 0.496, p = 0.78. b The temporal cortex of Aβ− and Aβ+ individuals displayed increased tau pathology when compared with the parietal and frontal cortices; Kruskal–Wallis test, H = 14.94, p < 0.0001 followed by Dunn’s multiple comparisons test. Data are displayed in boxplots where the median is represented by the horizontal line and the whiskers represent the 25th and 75th percentiles. Full circles represent outliers according to the Tukey method. IR immunoreactivity. *p < 0.05, **p < 0.01 negatively with the rate of cognitive decline in specific cognitive domains. While higher cytokine expression (IL-1α, IL-13, and MDC) in the temporal cortex was associated with worsening of cognitive function, the op- posite was true for cytokine levels in the parietal (MIP- 1β and TARC) and frontal (IL-16 and IP-10) cortices. These observations might reflect the transient nature of cytokine expression and its relationship with AD pro- gression [48]. Moreover, the negative associations be- tween cytokine expression in the temporal cortex and cognitive deficits might be associated with the increased tau pathology that was observed in this region. However, these complex exploratory observations warrant further comprehensive longitudinal investigations to elucidate whether cytokine expression exerts transient effects on cognition at different stages of the AD continuum. This study has several limitations that must be consid- ered. First, although post-mortem pathological analysis Fig. 7 Brain cytokine expression, Aβ pathology, and tau deposition in non-demented individuals with cerebral Aβ pathology. The average of AD-asymptomatic individuals indicated the presence fold change of inflammatory molecule expression per brain region was of pathology meeting AD criteria, we cannot ascertain calculated. Briefly, NCI individuals with Aβ deposition displayed an whether these individuals would have progressed increase in IL-1β (~ 1.8 fold change), IL-6 (~ 3.4 fold change), and through the AD continuum had they lived longer. How- eotaxin-3 (~ 3 fold change) in the temporal cortex; an increase in IL-1β ever, the faster cognitive decline in elderly individuals (~ 2 fold change) and MCP-1 (~ 1.8 fold change) in the parietal cortex; and similar trends revealing higher levels of IL-1β (~ 1.5 fold change), with Aβ deposition, the higher proportion of APOE e4 IL-6 (~ 1.8 fold change), and MCP-1 (~ 1.7 fold change) in the frontal individuals in this population, and the neuropathological cortex as compared with individuals without Aβ deposition. Tau diagnosis of AD according to modified NIA-Reagan and pathology was highest in the temporal cortex as compared with the CERAD criteria, would suggest that these individuals parietal and frontal cortices from Aβ+ individuals. Individuals with Aβ might have indeed been at the earliest stages of AD. pathology also showed cognitive deficits in perceptual speed Second, to study preclinical stages of AD and to avoid Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 11 of 13 confounding factors (such as the presence of Super- Additional file 3: Supplemental Table 3. Demographics of the Agers), we conducted our investigations in NCI individ- studied population: frontal cortical samples from elderly individuals. uals under age 80. Such criteria limited the number of Additional file 4: Supplemental Figure 1. Amyloid-beta and tau samples from the ROS and ROSMAP cohorts available neuropathology in brains with Aβ deposition. Figure legend: Representa- tive micrographs depicting amyloid-beta and tau pathology as revealed for analysis. Although the groups in our study were by immunohistochemistry (IHC) with 4G8 and AT8 antibodies, respect- sex-balanced, the restricted sample size was not ively. (A-B) Representative micrographs of cases with Aβ deposition in sufficient to investigate the effects of sex on cytokine the temporal, parietal and frontal cortices. CV = Cresyl violet counterstain- ing. Scale bar = 100μm. expression and cognition. Moreover, the narrow sample Additional file 5: Supplemental Table 4. Association of cognitive analyzed in this study could have limited the detection scores with global amyloid and tau levels. of subtle cognitive deficiencies in other cognitive Additional file 6: Supplemental Table 5. Correlation of post-mortem domains. Ideally, this study should be confirmed in a interval with cytokine expression in the temporal, parietal and frontal larger sample size, when available. Finally, we have seg- cortices. regated our population by the extent of extracellular Additional file 7: Supplemental Table 6. Association of cytokine expression with cortical amyloid-beta and pathological tau burden in the Aβ deposition. Besides being a highly selected sample, temporal, parietal and frontal cortices of non-demented elderly we acknowledge that there is not a clear “cutoff” value individuals. between what represents healthy and AD-related Aβ Additional file 8: Supplemental Table 7. Association of cognitive deposition. Moreover, to fully understand the earliest decline with cytokine expression in the temporal, parietal and frontal cortices of non-demented elderly individuals. inflammatory changes associated with preclinical AD, it would be important to investigate the neuroinflamma- Acknowledgments tory response associated with early intraneuronal Aβ We would like to thank the participants that consented to brain donation. accumulation. Equally important would be to measure We acknowledge the staff at the Rush Alzheimer’s Disease Center, especially soluble Aβ levels in brains with and without Aβ depos- Gregory Klein who provided the neuropathological and clinical data utilized in this study. LFA was the holder of a PBEEE Doctoral Merit Scholarship from ition to define the Aβ concentration threshold associ- the Fonds de Recherche du Quebec–Nature et technologies and a doctoral ated with cytokine dysregulation. scholarship granted by CONACyT, Mexico. MFI acknowledges current support from the Jérôme Lejeune and Sisley D’Ornano Foundations and previous support from the Fonds de Recherche du Québec-Santé (2015-2019). ACC is Conclusions the holder of the McGill University Charles E. Frosst/Merck Chair in Pharmacology and a member of the Canadian Consortium of Our findings provide a link between amyloid deposition, Neurodegeneration in Aging. key brain inflammatory cytokines, and cognitive decline in non-demented elderly individuals with amyloid accu- Authors’ contributions LFA, MFI, and ACC contributed to the conception and design of the study. mulation. Our study also suggests that stratification by LFA, CO, and NT acquired the data. LFA, MFI, and JY contributed to the amyloid positivity should be considered when evaluating analysis of data. DAB was the principal collaborator of this project and anti-inflammatory approaches for AD. The investigation provided the brain samples along with neuropathological and cognitive data. LFA drafted the initial manuscript and prepared the figures. LFA, MFI, of cognitively healthy individuals with AD neuropathol- and ACC significantly edited the final manuscript. All authors revised and ogy offers the possibility of identifying the earliest approved the final manuscript. neuropathological changes associated with preclinical Funding AD, allowing for the identification of novel therapeutic ACC would like to acknowledge support from the Canadian Institutes of targets and biomarker candidates. Health Research (201603PJT 364544). DAB was supported by grants P30AG10161 and R01AG15819 from the National Institute on Aging. Abbreviations Availability of data and materials AD: Alzheimer’s disease; Aβ: Amyloid-beta; CERAD : Consortium to Establish a The datasets used and/or analyzed during the current study are available Registry for Alzheimer’s Disease; IFN: Interferon; IL: Interleukin; IP- from the corresponding author on reasonable request. 10: Interferon-inducible protein-10,; MAP: Rush Memory and Aging Project; MCP-1: Monocyte chemoattractant protein; MDC: Macrophage-derived chemokine; MIP-1β: Macrophage inflammatory protein; NCI: No cognitive Declarations impairment; NIA: National Institute on Aging; ROS: Religious Orders Study; Ethics approval and consent to participate TARC: Thymus and activation-regulated chemokine; TNF: Tumor necrosis This project has been approved by the McGill University Research Ethics factor; VEGF-A: Vascular endothelial growth factor A Board. Consent for publication Supplementary Information Not applicable The online version contains supplementary material available at https://doi. org/10.1186/s12974-021-02169-0. Competing interests The authors declare that they have no competing interests. Additional file 1: Supplemental Table 1. Demographics of the studied population: temporal cortical samples from elderly individuals. Author details 1 Department of Anatomy and Cell Biology, McGill University, Montreal, Additional file 2: Supplemental Table 2. Demographics of the studied population: parietal cortical samples from elderly individuals. Quebec, Canada. 2Sant Pau Memory Unit, Department of Neurology, Hospital de la Santa Creu i Sant Pau, Biomedical Research Institute Sant Pau, Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 12 of 13 Universitat Autònoma de Barcelona, Barcelona, Spain. 3Department of 16. De Jager PL, Shulman JM, Chibnik LB, Keenan BT, Raj T, Wilson RS, et al. A Neurology and Neurosurgery, McGill University, Montreal, Canada. genome-wide scan for common variants affecting the rate of age-related 4 Department of Pharmacology and Therapeutics, McGill University, Montreal, cognitive decline. Neurobiol Aging. 2012;33(5):1017 e1–15. Quebec, Canada. 5Rush Alzheimer’s Disease Center, Rush University Medical 17. Bennett DA, Wilson RS, Boyle PA, Buchman AS, Schneider JA. Relation Center, Chicago, IL, USA. 6Department of Neurological Sciences, Rush of neuropathology to cognition in persons without cognitive University Medical Center, Chicago, IL, USA. 7Department of Pharmacology, impairment. Ann Neurol. 2012;72(4):599–609. https://doi.org/10.1002/a Oxford University, Oxford, UK. na.23654. 18. Baker JE, Lim YY, Pietrzak RH, Hassenstab J, Snyder PJ, Masters CL, et al. Received: 4 November 2020 Accepted: 11 May 2021 Cognitive impairment and decline in cognitively normal older adults with high amyloid-beta: a meta-analysis. Alzheimers Dement (Amst). 2017;6(1): 108–21. https://doi.org/10.1016/j.dadm.2016.09.002. 19. Sheng JG, Ito K, Skinner RD, Mrak RE, Rovnaghi CR, Van Eldik LJ, et al. References In vivo and in vitro evidence supporting a role for the inflammatory 1. McGeer PL, McGeer EG. The amyloid cascade-inflammatory hypothesis of cytokine interleukin-1 as a driving force in Alzheimer pathogenesis. Alzheimer disease: implications for therapy. Acta Neuropathol. 2013;126(4): Neurobiol Aging. 1996;17(5):761–6. https://doi.org/10.1016/0197-4 479–97. https://doi.org/10.1007/s00401-013-1177-7. 580(96)00104-2. 2. Vlad SC, Miller DR, Kowall NW, Felson DT. Protective effects of NSAIDs on 20. Rogers JT, Leiter LM, McPhee J, Cahill CM, Zhan SS, Potter H, et al. the development of Alzheimer disease. Neurology. 2008;70(19):1672–7. Translation of the Alzheimer amyloid precursor protein mRNA is up- https://doi.org/10.1212/01.wnl.0000311269.57716.63. regulated by interleukin-1 through 5'-untranslated region sequences. J Biol 3. Rogers J. Principles for central nervous system inflammation research: a call Chem. 1999;274(10):6421–31. https://doi.org/10.1074/jbc.274.10.6421. for a consortium approach. Alzheimers Dement. 2018;14(11):1553–9. https:// 21. Griffin WS, Liu L, Li Y, Mrak RE, Barger SW. Interleukin-1 mediates Alzheimer doi.org/10.1016/j.jalz.2018.01.008. and Lewy body pathologies. J Neuroinflammation. 2006;3(1):5. https://doi. 4. Cuello AC. Early and late CNS inflammation in Alzheimer's disease: two org/10.1186/1742-2094-3-5. extremes of a continuum? Trends Pharmacol Sci. 2017;38(11):956–66. 22. Rozemuller JM, Eikelenboom P, Stam FC, Beyreuther K, Masters CL. A4 https://doi.org/10.1016/j.tips.2017.07.005. protein in Alzheimer's disease: primary and secondary cellular events in 5. Hanzel CE, Pichet-Binette A, Pimentel LS, Iulita MF, Allard S, Ducatenzeiler A, extracellular amyloid deposition. J Neuropathol Exp Neurol. 1989;48(6):674– et al. Neuronal driven pre-plaque inflammation in a transgenic rat model of 91. https://doi.org/10.1097/00005072-198911000-00009. Alzheimer's disease. Neurobiol Aging. 2014;35(10):2249–62. https://doi.org/1 23. Heneka MT, Kummer MP, Stutz A, Delekate A, Schwartz S, Vieira-Saecker A, 0.1016/j.neurobiolaging.2014.03.026. et al. NLRP3 is activated in Alzheimer's disease and contributes to 6. Ferretti MT, Bruno MA, Ducatenzeiler A, Klein WL, Cuello AC. Intracellular pathology in APP/PS1 mice. Nature. 2013;493(7434):674–8. https://doi.org/1 Abeta-oligomers and early inflammation in a model of Alzheimer's disease. 0.1038/nature11729. Neurobiol Aging. 2012;33(7):1329–42. https://doi.org/10.1016/j.neurobiola ging.2011.01.007. 24. Griffin WS, Stanley LC, Ling C, White L, MacLeod V, Perrot LJ, et al. Brain 7. McAlpine FE, Lee JK, Harms AS, Ruhn KA, Blurton-Jones M, Hong J, interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome et al. Inhibition of soluble TNF signaling in a mouse model of and Alzheimer disease. Proc Natl Acad Sci U S A. 1989;86(19):7611–5. Alzheimer's disease prevents pre-plaque amyloid-associated https://doi.org/10.1073/pnas.86.19.7611. neuropathology. Neurobiol Dis. 2009;34(1):163–77. https://doi.org/10.101 25. Flores-Aguilar L, Iulita MF, Kovecses O, Torres MD, Levi SM, Zhang Y, et al. 6/j.nbd.2009.01.006. Evolution of neuroinflammation across the lifespan of individuals with 8. Cavanagh C, Tse YC, Nguyen HB, Krantic S, Breitner JC, Quirion R, et al. Down syndrome. Brain. 2020;143(12):3653. Inhibiting tumor necrosis factor-alpha before amyloidosis prevents synaptic 26. Basu A, Krady JK, Levison SW. Interleukin-1: a master regulator of deficits in an Alzheimer's disease model. Neurobiol Aging. 2016;47:41–9. neuroinflammation. J Neurosci Res. 2004;78(2):151–6. https://doi.org/10.1 https://doi.org/10.1016/j.neurobiolaging.2016.07.009. 002/jnr.20266. 9. Welikovitch LA, Do Carmo S, Magloczky Z, Malcolm JC, Loke J, Klein 27. Hull M, Fiebich BL, Lieb K, Strauss S, Berger SS, Volk B, et al. Interleukin-6- WL, et al. Early intraneuronal amyloid triggers neuron-derived associated inflammatory processes in Alzheimer's disease: new therapeutic inflammatory signaling in APP transgenic rats and human brain. Proc options. Neurobiol Aging. 1996;17(5):795–800. https://doi.org/10.1016/01 Natl Acad Sci U S A. 2020;117(12):6844–54. https://doi.org/10.1073/pna 97-4580(96)00107-8. s.1914593117. 28. Ringheim GE, Szczepanik AM, Petko W, Burgher KL, Zhu SZ, Chao CC. 10. Bennett DA, Schneider JA, Arvanitakis Z, Kelly JF, Aggarwal NT, Shah RC, Enhancement of beta-amyloid precursor protein transcription and et al. Neuropathology of older persons without cognitive impairment from expression by the soluble interleukin-6 receptor/interleukin-6 complex. Brain two community-based studies. Neurology. 2006;66(12):1837–44. https://doi. Res Mol Brain Res. 1998;55(1):35–44. https://doi.org/10.1016/S0169-32 org/10.1212/01.wnl.0000219668.47116.e6. 8X(97)00356-2. 11. De Meyer G, Shapiro F, Vanderstichele H, Vanmechelen E, Engelborghs S, De 29. Quintanilla RA, Orellana DI, Gonzalez-Billault C, Maccioni RB. Interleukin-6 Deyn PP, et al. Diagnosis-independent Alzheimer disease biomarker induces Alzheimer-type phosphorylation of tau protein by deregulating the signature in cognitively normal elderly people. Arch Neurol. 2010;67(8):949– cdk5/p35 pathway. Exp Cell Res. 2004;295(1):245–57. https://doi.org/10.1016/ 56. https://doi.org/10.1001/archneurol.2010.179. j.yexcr.2004.01.002. 12. Fagan AM, Mintun MA, Shah AR, Aldea P, Roe CM, Mach RH, et al. 30. Campbell IL, Stalder AK, Chiang CS, Bellinger R, Heyser CJ, Steffensen S, Cerebrospinal fluid tau and ptau(181) increase with cortical amyloid et al. Transgenic models to assess the pathogenic actions of cytokines in deposition in cognitively normal individuals: implications for future clinical the central nervous system. Mol Psychiatry. 1997;2(2):125–9. https://doi.org/1 trials of Alzheimer's disease. EMBO Mol Med. 2009;1(8-9):371–80. https://doi. 0.1038/sj.mp.4000225. org/10.1002/emmm.200900048. 31. McGeer PL, McGeer EG. Polymorphisms in inflammatory genes and the risk 13. Morris JC, Storandt M, McKeel DW Jr, Rubin EH, Price JL, Grant EA, et al. of Alzheimer disease. Arch Neurol. 2001;58(11):1790–2. https://doi.org/10.1 Cerebral amyloid deposition and diffuse plaques in "normal" aging: 001/archneur.58.11.1790. evidence for presymptomatic and very mild Alzheimer's disease. Neurology. 32. Li R, Strohmeyer R, Liang Z, Lue LF, Rogers J. CCAAT/enhancer binding 1996;46(3):707–19. https://doi.org/10.1212/WNL.46.3.707. protein delta (C/EBPdelta) expression and elevation in Alzheimer's disease. 14. Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies Neurobiol Aging. 2004;25(8):991–9. https://doi.org/10.1016/j.neurobiola account for most dementia cases in community-dwelling older persons. ging.2003.10.016. Neurology. 2007;69(24):2197–204. https://doi.org/10.1212/01.wnl.0000271 33. Craig-Schapiro R, Kuhn M, Xiong C, Pickering EH, Liu J, Misko TP, et al. 090.28148.24. Multiplexed immunoassay panel identifies novel CSF biomarkers for 15. Rogalski EJ, Gefen T, Shi J, Samimi M, Bigio E, Weintraub S, et al. Youthful Alzheimer's disease diagnosis and prognosis. PLoS One. 2011;6(4):e18850. memory capacity in old brains: anatomic and genetic clues from the https://doi.org/10.1371/journal.pone.0018850. northwestern SuperAging project. J Cogn Neurosci. 2013;25(1):29–36. 34. Westin K, Buchhave P, Nielsen H, Minthon L, Janciauskiene S, Hansson O. https://doi.org/10.1162/jocn_a_00300. CCL2 is associated with a faster rate of cognitive decline during early stages Flores-Aguilar et al. Journal of Neuroinflammation (2021) 18:147 Page 13 of 13 of Alzheimer's disease. PLoS One. 2012;7(1):e30525. https://doi.org/10.1371/ journal.pone.0030525. 35. Hu WT, Chen-Plotkin A, Arnold SE, Grossman M, Clark CM, Shaw LM, et al. Novel CSF biomarkers for Alzheimer's disease and mild cognitive impairment. Acta Neuropathol. 2010;119(6):669–78. https://doi.org/10.1007/ s00401-010-0667-0. 36. Naert G, Rivest S. CC chemokine receptor 2 deficiency aggravates cognitive impairments and amyloid pathology in a transgenic mouse model of Alzheimer's disease. J Neurosci. 2011;31(16):6208–20. https://doi.org/10.1 523/JNEUROSCI.0299-11.2011. 37. El Khoury J, Toft M, Hickman SE, Means TK, Terada K, Geula C, et al. Ccr2 deficiency impairs microglial accumulation and accelerates progression of Alzheimer-like disease. Nat Med. 2007;13(4):432–8. https://doi.org/10.1038/nm1555. 38. Galimberti D, Schoonenboom N, Scheltens P, Fenoglio C, Bouwman F, Venturelli E, et al. Intrathecal chemokine synthesis in mild cognitive impairment and Alzheimer disease. Arch Neurol. 2006;63(4):538–43. https:// doi.org/10.1001/archneur.63.4.538. 39. Sokolova A, Hill MD, Rahimi F, Warden LA, Halliday GM, Shepherd CE. Monocyte chemoattractant protein-1 plays a dominant role in the chronic inflammation observed in Alzheimer's disease. Brain Pathol. 2009;19(3):392– 8. https://doi.org/10.1111/j.1750-3639.2008.00188.x. 40. Janelidze S, Mattsson N, Stomrud E, Lindberg O, Palmqvist S, Zetterberg H, et al. CSF biomarkers of neuroinflammation and cerebrovascular dysfunction in early Alzheimer disease. Neurology. 2018;91(9):e867–e77. https://doi.org/1 0.1212/WNL.0000000000006082. 41. Rodriguez-Vieitez E, Saint-Aubert L, Carter SF, Almkvist O, Farid K, Scholl M, et al. Diverging longitudinal changes in astrocytosis and amyloid PET in autosomal dominant Alzheimer's disease. Brain. 2016;139(Pt 3):922–36. https://doi.org/10.1093/brain/awv404. 42. Barroeta-Espar I, Weinstock LD, Perez-Nievas BG, Meltzer AC, Siao Tick Chong M, Amaral AC, et al. Distinct cytokine profiles in human brains resilient to Alzheimer's pathology. Neurobiol Dis. 2019;121:327–37. https:// doi.org/10.1016/j.nbd.2018.10.009. 43. Perez-Nievas BG, Stein TD, Tai HC, Dols-Icardo O, Scotton TC, Barroeta-Espar I, et al. Dissecting phenotypic traits linked to human resilience to Alzheimer's pathology. Brain. 2013;136(Pt 8):2510–26. https://doi.org/10.1 093/brain/awt171. 44. Gefen T, Kim G, Bolbolan K, Geoly A, Ohm D, Oboudiyat C, et al. Activated microglia in cortical white matter across cognitive aging trajectories. Front Aging Neurosci. 2019;11:94. https://doi.org/10.3389/fna gi.2019.00094. 45. Scheff SW, Ansari MA, Mufson EJ. Oxidative stress and hippocampal synaptic protein levels in elderly cognitively intact individuals with Alzheimer's disease pathology. Neurobiol Aging. 2016;42:1–12. https://doi.org/10.1016/j. neurobiolaging.2016.02.030. 46. Pentz R, Iulita MF, Ducatenzeiler A, AC BDAC. The human brain NGF metabolic pathway is impaired in the pre-clinical and clinical continuum of Alzheimers disease. Mol Psychiatry. 2020. https://doi.org/1 0.1038/s41380-020-0797-2. 47. De Jager PL, Srivastava G, Lunnon K, Burgess J, Schalkwyk LC, Yu L, et al. Alzheimer's disease: early alterations in brain DNA methylation at ANK1, BIN1, RHBDF2 and other loci. Nat Neurosci. 2014;17(9):1156–63. https://doi. org/10.1038/nn.3786. 48. Brosseron F, Krauthausen M, Kummer M, Heneka MT. Body fluid cytokine levels in mild cognitive impairment and Alzheimer's disease: a comparative overview. Mol Neurobiol. 2014;50(2):534–44. https://doi.org/10.1007/s12035- 014-8657-1. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.