sTREM2 and GFAP Mediated the Association of IGF-1 Signaling Biomarkers with Alzheimer’s Disease Pathology

Abstract

Defects in insulin-like growth factor 1 (IGF-1) signaling is a key contributor to Alzheimer’s disease (AD). However, the mechanism of how IGF-1 signaling relates to AD remained unclear. Here, we investigated the association of IGF-1 signaling associated biomarkers with AD pathology, sTREM2, and GFAP. Finally, insulin-like growth factor binding protein 2 (IGFBP-2) was associated with AD pathology, and the association was partly medicated by sTREM2 (Aβ₄₂, β= 0.794, p = 0.016; T-tau, β= 0.291, p < 0.001; P-tau₁₈₁, β= 0.031, p < 0.001) and GFAP (T-tau, β= 0.427, p < 0.001; P-tau₁₈₁, β= 0.044, p < 0.001). It suggested that sTREM2 and GFAP mediated the relationship between IGF-1 signaling and AD pathology.

Keywords

AD pathology Alzheimer’s disease GFAP IGF-1 signaling IGFBP-2 sTREM2

INTRODUCTION

Alzheimer’s disease (AD) is characterized pathologically by aberrant senile plaques (amyloid deposition) and neurofibrillary tangles (tau phosphorylation), and clinically by cognitive decline and behavioral disorders [1]. AD is the common comorbidity of type 2 diabetes mellitus (T2DM) and T2DM is the crucial risk factor for dementia and AD [2, 3]. To date, emerging evidence identified that insulin resistance, the core mechanism underlying T2DM, also contributes to the onset of AD [4, 5]. However, the mechanism underlying the contribution of insulin resistance to AD pathogenesis was still unclear. Recently, it was observed that insulin resistance activated microglia and astrocytes to be involved in the biological process of AD pathology [amyloid-β (Aβ) and tau] in animal models [2]. Thus, it was possible that insulin resistance signaling, such as impairments in insulin signaling or insulin-like growth factor 1 (IGF-1) signaling, correlates with the activity of microglia and astrocytes and mediates the development of AD pathology in human brain.

Impaired IGF-1 signaling was the core feature of insulin resistance and was reported to be related to amyloid plaque, tau phosphorylation, and higher incidence of AD in previous studies [6, 7]. Insulin like growth factor binding proteins (IGFBPs) limit IGF-1 actions by altering the affinity of IGF-1 to their receptors. It was also observed that IGFBPs were correlated with the levels of Aβ and tau pathology previously [8, 9]. In addition, soluble triggering receptor expressed on myeloid cells 2 (sTREM2) is a soluble form of TREM2 and plays roles in modulating immune responses and activating microglia during the phagocytosis of Aβ pathology [10 –12]. Glial fibrillary acidic protein (GFAP) is an intermediate filament mainly expressed in astrocytes. GFAP is overexpressed in reactive astrocytes affecting neuroinflammatory and is associated with Aβ and tau pathology in AD brain [13, 14]. Based on the above findings, we proposed that IGF-1 signaling associated markers are related to AD pathology, and the associations may be mediated by microglia and astrocyte markers. Here, we aimed to extract the results of IGF-1 and IGFBPs from CSF proteomics data in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database to investigate the associations of IGF-1 signaling associated markers (IGF-1 and IGFBPs) with AD pathology, sTREM2 and GFAP, and to identify whether sTREM2 and GFAP mediated the associations of IGF-1 signaling with AD pathology in brain.

MATERIALS AND METHODS

Participants

All participants were from the ADNI database (https://adni.loni.usc.edu). The detailed full inclusion/exclusion criteria can be obtained on their website. Participants underwent systematic neuropsychological evaluations and were offered biological samples such as blood and cerebrospinal fluid (CSF) samples. The ADNI project was conducted in link with the Declaration of Helsinki. The study was approved by the Institutional Review Board of Qingdao Municipal Hospital.

CSF IGF-1 signaling associated markers

The results of IGF-1 and IGFBPs were from “Biomarkers Consortium ADNI CSF QC Multiplex data.csv” online (https://loni.usc.edu/). Given that the results of IGF-1 did not pass the quality control of data, only the results of IGFBP-2 were extracted to perform the next analysis. The measurements of CSF proteomics were based on Luminex immunoassay platform and improved by Rules Based Medicine to test inflammatory, lipid, and other molecules. Detailed information on CSF proteomics and quality control can be found in the content of “Biomarkers Consortium Data Primer” online.

AD core biomarkers

In this study, Aβ₄₂, T-tau, and P-tau₁₈₁ were obtained from the ADNI dataset “Upen_Biomk9.csv” online. The AD core biomarkers were analyzed by the electrochemiluminescence immunoassays (ECLIA) Elecsys total-tau CSF, the Elecsys phospho-tau(181P) CSF, and the Elecsys β-amyloid (1–42) CSF on an automatic Elecsys instrument. CSF sample collection, processing, and quality control of Aβ₄₂, T-tau, and P-tau₁₈₁ were described elsewhere [15]. Based on the previous study, we used CSF Aβ₄₂≤976.6 pg/mL, P-tau₁₈₁≥21.8 pg/ml, and T-tau≥245 g/ml as thresholds to identify abnormal levels [16].

CSF sTREM2 and GFAP

The measurement of CSF sTREM2 was performed with the MSD platform-based assay, which was previously reported and validated [17]. The sTREM2 data were from the file “CSF soluble triggering receptor expressed on myeloid cells 2 (sTREM2) and progranulin (PGRN)” online. The GFAP data were from the ADNI dataset “Biomarkers Consortium CSF Proteomics MRM Date” online. The methods of measurement were CSF MRM multiple panel tests developed by Caprion Proteomics in collaboration with the Biomarker Consortium Project Team. The information can be found on the website of ADNI.

Statistical analyses

Based on the National Institute on Aging-Alzheimer’s Association (NIA-AA) criteria [18], the participants with normal Aβ₄₂ (A), P-tau₁₈₁ (T), and T-tau (N) levels are classified as the stage 0; the participants with abnormal Aβ₄₂ (A), normal P-tau₁₈₁ (T) and normal T-tau (N) were classified as the stage 1; the participants with abnormal Aβ₄₂ (A), and abnormal P-tau₁₈₁ (T) or abnormal T-tau (N) levels as the stage 2; and participants with normal Aβ₄₂, but abnormal P-tau₁₈₁ or abnormal T-tau levels are classified as suspected non-Alzheimer’s pathophysiology (SNAP) group.

The differences among the four groups were computed using the one-way analysis of variance (ANOVA) for continuous variables and chi-square analysis for categorical variables. We assessed the association of IGFBP-2 (independent variable) with AD core biomarkers, sTREM2 and GFAP using a multiple linear regression model with age, sex, years of education, and APOE ɛ4 carrier status as covariates. Mediation analyses were fitted to investigate whether the association between IGFBP-2 and AD pathology was mediated by sTREM2 and GFAP, and the indirect effect was computed with 10,000 bootstrapped iterations. p value < 0.05 was considered significant.

RESULTS

Characteristics of participants

In the 233 participants who underwent the measurement of CSF IGFBP-2, AD core biomarkers (Aβ₄₂, T-tau, and P-tau181), sTREM2 and GFAP, 43 participants were in the stage 0, 33 participants were in the stage 1, 122 participants were in the stage 2, and 35 participants were in the SNAP group. The demographical and clinical characteristics are described in Table 1. There were no differences in terms of age, gender, and educational years between the four groups. As expected, the levels of AD core biomarkers (p < 0.01) showed significant differences between the four groups, and the proportion of APOE ɛ4 carriers (p < 0.01) was different among the four groups. In addition, the levels of sTREM2 (p < 0.01) and GFAP (p < 0.01) in CSF showed remarkable differences across the four groups. Finally, there is a significant difference on the level of IGFBP-2 in CSF between the four groups.

Table 1

The demographic and clinical characteristics of participants

	Stage 0	Stage 1	Stage 2	SNAP	p
Number of individuals	43	33	122	35
Age (y), mean (sd)	75.3 (5.35)	75.3 (5.52)	74.6 (7.27)	77.0 (7.00)	0.312
Female, n (%)	15 (34.9%)	9 (27.3%)	56 (45.9%)	13 (37.1%)	0.205
Education (y), mean (sd)	15.6 (2.97)	15.5 (3.80)	15.7 (3.02)	15.9 (2.88)	0.957
APOE ɛ4 carriers, n (%)	4 (9)	15 (46)	89 (73)	6 (17)	<0.001
Aβ₄₂ (pg/ml), mean (sd)	1445 (256)	642 (194)	606 (168)	1695 (613)	<0.001
T-tau (pg/ml), mean (sd)	192 (31.4)	177 (34.4)	368 (107)	332 (106)	<0.001
P-tau₁₈₁ (pg/ml), mean (sd)	16.9 (2.82)	16.2 (3.61)	37.4 (12.5)	30.6 (12.1)	<0.001
GFAP (pg/ml), mean (sd)	11.0 (0.37)	10.9 (0.56)	11.2 (0.54)	11.3 (0.46)	0.001
sTREM2 (pg/ml), mean (sd)	4,451 (2130)	3,074 (1292)	4,729 (2592)	5,658 (2016)	<0.001
IGFBP-2 (ng/ml), mean (sd)	102 (18.7)	95.8 (15.9)	109 (38.8)	118 (32.4)	0.027

Stage 0, normal Aβ₄₂ with normal P-tau₁₈₁ and T-tau level; Stage 1, abnormal Aβ₄₂ with normal P-tau₁₈₁ and normal T-tau level; Stage 2, abnormal Aβ₄₂ with abnormal P-tau₁₈₁ or abnormal T-tau level; SNAP, abnormal P-tau₁₈₁ or abnormal T-tau but with normal Aβ₄₂; sd, standard deviation; Aβ₄₂, amyloid-β_1–42; T-tau, total-tau; P-tau₁₈₁, phosphorylated-tau; sTREM2, soluble TREM2; GFAP, glial fibrillary acidic protein; IGFBP-2, insulin-like Growth Factor-Binding Protein 2; SNAP, suspected non-Alzheimer’s pathophysiology.

IGFBP-2 is associated with AD core biomarkers, sTREM2, and GFAP

The correlations between CSF IGFBP-2, AD core biomarkers, sTREM2, and GFAP were tested in 233 participants (Fig. 1A). We found that the elevated IGFBP-2 level was associated with the higher levels of GFAP (β= 0.005, 95% CI: 0.001–4.732, p < 0.001) (Fig. 1B) and sTREM2 (β= 24.408, 95% CI: 4.659–5.239, p < 0.001) (Fig. 1C). Increased IGFBP-2 level was remarkably correlated with higher levels of P-tau₁₈₁ (β= 0.122, 95% CI: 0.027–4.529, p < 0.001) (Fig. 1D) and T-tau (β= 1.186, 95% CI: 0.236–5.028, p < 0.001) (Fig. 1E). Further, these associations between these biomarkers still existed in sensitivity analyses with various covariates regressed out and in the mild cognitive impairment (MCI) group (Supplementary Table 1), which verified the robustness of our results. In addition, sTREM2 and GFAP were also markedly related to the levels of AD core biomarkers (Fig. 1A).

Fig. 1

Associations of IGFBP-2 with AD core biomarkers, sTREM2 and GFAP. A) The partial correlation between IGFBP-2 and AD core biomarkers, sTREM2 and GFAP with age, sex, years of education, and APOE ɛ4 status as control variables. B) IGFBP-2 was associated with the level of sTREM2 in a multiple regression model (β= 0.005, p < 0.001). C) IGFBP-2 was associated with the level of GFAP in a multiple regression model (β= 24.408, p < 0.001). D) IGFBP-2 was associated with T-tau in a multiple regression model (β= 0.122, p < 0.001). E) IGFBP-2 was associated with P-tau₁₈₁ in a multiple regression model (β= 0.186, p < 0.001). The multiple linear regression model was adjusted for age, sex, years of education, and APOE ɛ4 status. IGFBP-2, Insulin like growth factor binding protein 2; Aβ₄₂, amyloid-β_1–42; P-tau₁₈₁, phosphorylated tau protein; T-tau, total tau protein; sTREM2, soluble triggering receptor expressed on myeloid cells 2; GFAP, glial fibrillary acidic protein. *<0.05, **<0.01, ***<0.001.

sTREM2 and GFAP mediate the effects of IGFBP-2 on AD pathology

We investigated whether the connections between IGFBP-2 and AD core biomarkers were mediated by sTREM2 or GFAP. The association between IGFBP-2 and Aβ₄₂ was partly mediated by sTREM2 (β= 0.794, 95% CI: 0.124 –1.610, p = 0.016), but was not mediated by GFAP (β= 0.072, p = 0.815) (Fig. 2A). The connection between IGFBP-2 and T-tau was partly mediated by sTREM2 (β= 0.291, 95% CI: 0.122 –0.500, p < 0.001) and GFAP (β= 0.427, 95% CI: 0.222 –0.680, p < 0.001) (Fig. 2B). Likewise, the relationship between IGFBP-2 and P-tau₁₈₁ is partially mediated by sTREM2 (β= 0.031, 95% CI: 0.011 –0.060, p < 0.001) and GFAP (β= 0.044, 95% CI: 0.022 –0.070, p < 0.001) (Fig. 2C).

Fig. 2

Mediation effects of sTREM2 and GFAP on the link between IGFBP-2 and AD pathology. A) Mediation effects of sTREM2 (β= 0.794, p = 0.016) and GFAP (β= 0.072, p = 0.815) on the association between IGFBP-2 and Aβ₄₂. B) Mediation effects of sTREM2 (β= 0.291, p < 0.001) and GFAP (β= 0.427, p < 0.001) on the association between IGFBP-2 and T-tau. C) Mediation effects of sTREM2 (β= 0.031, p < 0.001) and GFAP (β= 0.044, p < 0.001) on the association between IGFBP-2 and P-tau₁₈₁. The effect sizes and p values were calculated in the mediation analyses with 10,000 bootstrapped iterations, which showed that cerebrospinal fluid sTREM2 or GFAP partly mediated the association between IGFBP-2 and AD core biomarkers. IGFBP-2, Insulin like growth factor binding protein 2; Aβ₄₂, amyloid-β_1–42; T-tau, total tau protein; P-tau₁₈₁, phosphorylated tau protein; sTREM2, soluble triggering receptor expressed on myeloid cells 2; GFAP, glial fibrillary acidic protein; IE, indirect effect.

DISCUSSION

Our study showed that IGFBP-2 was significantly correlated with AD core biomarkers, sTREM2 and GFAP. Besides, sTREM2 was significantly associated with the levels of Aβ₄₂, T-tau, and P-tau₁₈₁, and GFAP was markedly related to the levels of T-tau and P-tau₁₈₁. Next, we found that sTREM2 mediated the effects of IGFBP-2 on Aβ pathology (Aβ in CSF) and tau pathology (P-tau₁₈₁ and T-tau in CSF), and GFAP partially mediated the effect of IGFBP-2 on tau pathology (P-tau₁₈₁ and T-tau in CSF). These findings suggested that IGFBP-2 may correlate with the AD pathology as it is associated with the levels of sTREM2 and GFAP in CSF.

It has been documented that insulin resistance is the risk factor for incidence of AD [4, 5]. Deficits in IGF-1 signaling were the critical cause of insulin resistance [6, 19], and IGFBP-2 was involved in the IGF-1 signaling pathway by binding IGF-1 with high affinity. To date, epidemiological studies identified that both IGF-1 and IGFBP-2 were significantly associated with an increased risk of AD in Framingham prospective cohort [7, 20], and IGF-1 signaling associated molecules co-occurred with amyloid plaques in the hippocampus and cerebellar cortex in postmortem studies [5]. In addition, IGF-1 signaling was reported to regulate the change of AD pathology through altering the action of microglia and astrocytes in animal models [2]. Here, our study found that IGF-1 signaling associated biomarker (IGFBP-2) was related to AD pathology, and the relationship was mediated by sTREM2 and GFAP. These findings provided new evidence for the mechanisms of how IGF-1 signaling was related to the change of AD pathology.

The study was designed to test the association between IGF-1 signaling, AD pathology, and sTREM2 and GFAP. At first, we intended to extract the results of IGF-1 and IGFBPs from the CSF proteomics data to detect their associations with AD core biomarkers, whereas the results of IGF-1 failed to pass the quality control of proteomics data, and only IGFBP-2 was used in our study. Age, sex, and other factors were considered as covariates in our analyses due to their effects on IGFBP-2 and other markers. Moreover, it was possible that the level of IGFBP-2 was affected by special diseases (e.g., diabetes and hypertension) and medication history. We analysed the associations between IGFBP-2 and these diseases to test this possibility. Finally, neither diabetes (β= 0.990, p = 0.447) nor hypertension (β= 1.006, p = 0.172) was associated with the level of IGFBP-2 after controlling for age and sex. Additionally, due to limited information on medication history in the ADNI database, we did not investigate the effects of medication history on the content changes of these CSF biomarkers.

sTREM2 is the proteolytic product from TREM2, a transmembrane receptor expressed by microglia, and the level of sTREM2 in CSF could reflect the activity of microglia in central nervous system (CNS). It has been established that sTREM2 could regulate the amyloid pathology and neurodegeneration in CNS [2, 21]. Moreover, GFAP is generated from astrocytes and the level of GFAP in CSF was in line with the activity of astrocytes in brain. Similarly, GFAP was identified to be involved in the AD pathology (amyloid and tau pathology) in CNS [22, 23]. Further, sTREM2 and GFAP were recently identified to be potential diagnostic biomarkers for AD owning to their strong association with AD pathology [22 , 25]. Our study showed that sTREM2 and GFAP mediated the association between IGFBP-2 and AD pathology, which supported the hypotheses that insulin resistance could regulate the activation of microglia and astrocyte to participate in the AD pathology.

In summary, this study found that CSF IGFBP-2 was related to AD pathology, and the relationship was partly mediated by sTREM2 and GFAP. These findings elucidate that IGF-1 signaling correlates with AD pathology for it is related to sTREM2 and GFAP in brain. However, this investigation only focused on IGFBP-2, and it was necessary to explore the mechanisms of how other markers in the IGF-1/insulin signaling pathway were related to AD pathology in future.

Footnotes

ACKNOWLEDGMENTS

The authors thank all the researchers and participants in the ADNI initiative. Data used in preparation of this article were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database (https://adni.loni.usc.edu). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report. A complete listing of ADNI investigators can be found at:

FUNDING

This study was supported by grants from the National Natural Science Foundation of China (81971032), and the Taishan Scholars Program of Shandong Province (tsqn201812157).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The dataset supporting the conclusions of this article is available in the ADNI site, .

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

References

Knopman

, Amieva

, Petersen

, Chételat

, Holtzman

, Hyman

, Nixon

, Jones

(2021) Alzheimer disease. Nat RevDis Primers 7, 33.

Chen

, Cai

, Hoover

, Kahn

(2022) Insulin action in the brain: Cell types, circuits, and diseases. Trends Neurosci 45, 384–400.

, Xu

, Tan

, Andrieu

, Suckling

, Evangelou

, Pan

, Zhang

, Jia

, Feng

, Kua

, Wang

, Tan

, Li

, Hou

, Wan

, Tan

, Mok

, Tan

, Dong

, Touchon

, Gauthier

, Aisen

, Vellas

(2020) Evidence-based prevention of Alzheimer’s disease: Systematic review and meta-analysis of 243 observational prospective studies and 153 randomised controlled trials. J Neurol Neurosurg Psychiatry 91, 1201–1209.

Diehl

, Mullins

, Kapogiannis

(2017) Insulin resistance in Alzheimer’s disease. Transl Res 183, 26–40.

Talbot

, Wang

, Kazi

, Han

, Bakshi

, Stucky

, Fuino

, Kawaguchi

, Samoyedny

, Wilson

, Arvanitakis

, Schneider

, Wolf

, Bennett

, Trojanowski

, Arnold

(2012) Demonstrated brain insulin resistance in Alzheimer’s disease patients is associated with IGF-1 resistance, IRS-1 dysregulation, and cognitive decline. J Clin Invest 122, 1316–1338.

Carro

, Trejo

, Gomez-Isla

, LeRoith

, Torres-Aleman

(2002) Serum insulin-like growth factor I regulates brain amyloid-beta levels. Nat Med 8, 1390–1397.

Westwood

, Beiser

, DeCarli

, Harris

, Chen

, He

X-m

, Roubenoff

, Pikula

, Au

, Braverman

, Wolf

, Vasan

, Seshadri

(2014) Insulin-like growth factor-1 and risk of Alzheimer dementia and brain atrophy. Neurology 82, 1613–1619.

Lane

, Hohman

, Jefferson

, for the Alzheimer’s Disease Neuroimaging Initiative (2017) Insulin-like growth factor binding protein-2 interactions with Alzheimer’s disease biomarkers. Brain Imaging Behav 11, 1779–1786.

Bonham

, Geier

, Steele

NZR

, Holland

, Miller

, Dale

, Desikan

, Yokoyama

(2018) Insulin-like growth factor binding protein 2 is associated with biomarkers of Alzheimer’s disease pathology and shows differential expression in transgenic mice. Front Neurosci 12, 476.

10.

Zhang

, Hu

, Tan

, Yu

(2021) Microglia biomarkers in Alzheimer’s disease. Mol Neurobiol 58, 3388–3404.

11.

Yang

, Fu

, Zhang

, Xiong

, Meng

, Zhang

(2020) TREM2 ectodomain and its soluble form in Alzheimer’s disease. J Neuroinflammation 17, 204.

12.

Jiang

, Tan

, Zhu

, Zhang

, Cao

, Tan

, Gu

, Wang

, Ding

, Zhang

, Yu

(2014) Upregulation of TREM2 ameliorates neuropathology and rescues spatial cognitive impairment in a transgenic mouse model of Alzheimer’s disease. Neuropsychopharmacology 39, 2949–2962.

13.

Carter

, Herholz

, Rosa-Neto

, Pellerin

, Nordberg

, Zimmer

(2019) Astrocyte biomarkers in Alzheimer’s disease. Trends Mol Med 25, 77–95.

14.

Olsen

, Aguilar

, Sehlin

, Fang

, Antoni

, Erlandsson

, Syvänen

(2018) Astroglial responses to amyloid-beta progression in a mouse model of Alzheimer’s disease. Mol Imaging Biol 20, 605–614.

15.

Shaw

, Vanderstichele

, Knapik-Czajka

, Figurski

, Coart

, Blennow

, Soares

, Simon

, Lewczuk

, Dean

, Siemers

, Potter

, Lee

, Trojanowski

(2011) Qualification of the analytical and clinical performance of CSF biomarker analyses in ADNI. Acta Neuropathol 121, 597–609.

16.

Suárez-Calvet

, Morenas-Rodríguez

, Kleinberger

, Schlepckow

, Araque Caballero

, Franzmeier

, Capell

, Fellerer

, Nuscher

, Eren

, Levin

, Deming

, Piccio

, Karch

, Cruchaga

, Shaw

, Trojanowski

, Weiner

, Ewers

, Haass

(2019) Early increase of CSF sTREM2 in Alzheimer’s disease isassociated with tau related-neurodegeneration but not withamyloid-β pathology. Mol Neurodegener 14, 1.

17.

Kleinberger

, Yamanishi

, Suárez-Calvet

, Czirr

, Lohmann

, Cuyvers

, Struyfs

, Pettkus

, Wenninger-Weinzierl

, Mazaheri

, Tahirovic

, Lleó

, Alcolea

, Fortea

, Willem

, Lammich

, Molinuevo

, Sánchez-Valle

, Antonell

, Ramirez

, Heneka

, Sleegers

, van der Zee

, Martin

, Engelborghs

, Demirtas-Tatlidede

, Zetterberg

, Van Broeckhoven

, Gurvit

, Wyss-Coray

, Hardy

, Colonna

, Haass

(2014) TREM2 mutationsimplicated in neurodegeneration impair cell surface transport andphagocytosis. Sci Transl Med 6, 243ra286.

18.

Jack

Jr. , Bennett

, Blennow

, Carrillo

, Dunn

, Haeberlein

, Holtzman

, Jagust

, Jessen

, Karlawish

, Liu

, Molinuevo

, Montine

, Phelps

, Rankin

, Rowe

, Scheltens

, Siemers

, Snyder

, Sperling

(2018) NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 14, 535–562.

19.

Sui

, Yang

, Wang

, Xiong

, Guo

, Chen

, Wang

(2021) Exogenous IGF-1 improves tau pathology and neuronal pyroptosis in high-fat diet mice with cognitive dysfunction. Metab Brain Dis 36, 2079–2088.

20.

McGrath

, Himali

, Levy

, Conner

, DeCarli

, Pase

, Courchesne

, Satizabal

, Vasan

, Beiser

, Seshadri

(2019) Circulating IGFBP-2: A novel biomarker for incident dementia. Ann Clin Transl Neurol 6, 1659–1670.

21.

Jonsson

, Stefansson

, Steinberg

, Jonsdottir

, Jonsson

, Snaedal

, Bjornsson

, Huttenlocher

, Levey

, Lah

, Rujescu

, Hampel

, Giegling

, Andreassen

, Engedal

, Ulstein

, Djurovic

, Ibrahim-Verbaas

, Hofman

, Ikram

, van Duijn

, Thorsteinsdottir

, Kong

, Stefansson

(2013) Variant of TREM2 associated with the risk of Alzheimer’s disease. N Engl J Med 368, 107–116.

22.

Pereira

, Janelidze

, Smith

, Mattsson-Carlgren

, Palmqvist

, Teunissen

, Zetterberg

, Stomrud

, Ashton

, Blennow

, Hansson

(2021) Plasma GFAP is an early marker of amyloid-β but not tau pathology in Alzheimer’s disease. Brain 144, 3505–3516.

23.

Elahi

, Casaletto

, La Joie

, Walters

, Harvey

, Wolf

, Edwards

, Rivera-Contreras

, Karydas

, Cobigo

, Rosen

, DeCarli

, Miller

, Rabinovici

, Kramer

(2020) Plasma biomarkers of astrocytic and neuronal dysfunction in early- and late-onset Alzheimer’s disease. Alzheimers Dement 16, 681–695.

24.

Suárez-Calvet

, Kleinberger

, Araque Caballero

, Brendel

, Rominger

, Alcolea

, Fortea

, Lleó

, Blesa

, Gispert

, Sánchez-Valle

, Antonell

, Rami

, Molinuevo

, Brosseron

, Traschütz

, Heneka

, Struyfs

, Engelborghs

, Sleegers

, Van Broeckhoven

, Zetterberg

, Nellgård

, Blennow

, Crispin

, Ewers

, Haass

(2016) sTREM2 cerebrospinal fluid levels are apotential biomarker for microglia activity in early-stageAlzheimer’s disease and associate with neuronal injury markers. EMBO Mol Med 8, 466–476.

25.

, Tan

, Bi

, Shen

, Xu

, Ma

, Li

, Dong

, Yu

(2020) Dynamic changes of CSF sTREM2 in preclinical Alzheimer’s disease: The CABLE study. Mol Neurodegener 15, 25.