Fig 2: Quantification of YKL-40, tau pathology burden and astrogliosis in different tauopathies. a–o Representative images of YKL-40, tau and GFAP immunoreactivity of controls and the four tauopathies under study. a–e YKL-40 expression pattern. f–j Main tau deposits and aggregates of each condition. k–o Astrogliosis among the tauopathies and controls. Scale bar = 20 µm. p YKL-40 immunoreactivity (objects/µm2) measured in all conditions. q–s Correlation between different markers. Levels of YKL-40 and GFAP correlated positively with tau aggregation. Solid lines indicate the linear regression, and dotted lines indicate 95% CI. RS, Spearman rho coefficient.

Fig 3: Biomarker candidates in tissue and plasma.(A) Immunohistochemistry staining for LGALS3BP, S100A8/A9, CHI3L1, and MMP9 plus SMC (SMA, CNN1) and macrophage (CD68) markers. Scale bars: 500 µm. (B) Association between the biomarker candidates and atherosclerosis (n = 560) and CVD (n = 685) in the Bruneck study. Direction and strength of associations are coded by color and significance levels by size of frames. ORs are derived from logistic regression analysis, and hazard ratios from Cox models. ORs and hazard ratios are expressed for a 1-SD higher loge-transformed level of the given marker. Full details are presented in Supplemental Tables 6 and 7. Model 1 (M1) was adjusted for age (years) and sex. M2 was additionally controlled for standard vascular risk factors, including LDL cholesterol (mg/dl), HDL cholesterol (mg/dl), high-sensitivity CRP (mg/l), diabetes mellitus (0 vs. 1), hypertension (0 vs. 1), smoking (pack-years), and body mass index (kg/m2). *Additionally adjusted for atherosclerosis score (mm).

Fig 4: YKL-40 immunoreactivity and protein levels remain similar in post-mortem frontal cortex from FTLD and non-demented controls. A Representative images of paraffin sections from the frontal cortex stained with the anti-YKL-40 antibody. YKL-40 immunoreactivity is present in FTLD and control cases in glial and neuronal cells. Scale bars represent 50µM. Semi-quantitation of YKL-40 immunoreactivity was performed by grouping cases into either negative (i.e., zero or 1 positive cell) or positive (i.e., 2 or more positive cell groups). Stacked bar plots represent the percentage of cases with either negative (-, white area) or positive (+, black area) YKL-40 immunoreactivity in NDC (n = 7) and FTLD cases (n = 24; FTLD-Tau = 19 and FTLD-TDP = 5). B Representative immunoblot of YKL-40 in frontal cortex lysates. Actin was used as a loading control. Box-dot plots depict YKL-40 immunoblot reactivity corrected for actin in NDC (n = 14) and FTLD cases (n = 67; FTLD-Tau = 33, FTLD-TDP = 34). C YKL-40 levels were quantified by ELISA and corrected for total protein concentration in NDC (n = 14) and FTLD cases (n = 67; FTLD-Tau = 33 and FTLD-TDP = 34). Overall, no significant differences across groups were identified. Box represents median ± interquartile range with bars showing the lowest to highest points. Abbreviations: NDC, non-demented control; FTLD, frontal temporal lobar degeneration; n.s., non-significant difference between groups

Fig 5: YKL-40 expression pattern in human brain tissue from an AD patient and a healthy control. Representative images of double immunofluorescence performed with YKL-40 (red) and three different cellular markers, GFAP (astroglial, green), MAP2 (neuronal, green) and IBA-1 (microglial, green). Nuclei are marked in blue. a–h YKL-40 immunoreactivity was detected in the cytoplasm of GFAP+ cells (asterisk), indicating an astroglial origin. i–p No colocalization was observed between YKL-40 and the neuronal marker, MAP2 or (q–x) with the microglial marker, IBA-1. Scale bar = 20 µm