YKL-40 (Chitinase 3-like I) is expressed in a subset of astrocytes in Alzheimer’s disease and other tauopathies

There is growing evidence that the immune system is involved early in the pathogenesis of Alzheimer’s disease (AD) and in other neurodegenerative diseases [1, 2]. The activation of the immune system in AD (often referred to as “neuroinflammation”) is known to be present at all stages of AD and is believed to play an active role in the disease process. The activation of microglia and astrocytes as a reaction to ongoing deposition of Aβ triggers the production of several proinflammatory signal molecules including cytokines, chemokines, complement molecules, growth factors and cell adhesion molecules [3]. The recent association of gene encoding inflammatory proteins, such as TREM2 and CD33 with AD [4‐6], has further supported the role of the innate immune response in the aetiology and progression of AD. In addition, the increased feasibility of measuring a wide range of inflammatory molecules in biofluids from patients at different AD stages has expanded our understanding about the type of immune responses observed in neurodegenerative diseases. A variety of cytokines, chemokines and other inflammatory mediators are increased in cerebrospinal fluid (CSF) or in plasma in AD [1]. One of the proteins that has been frequently measured in body fluids as a surrogate marker of neuroinflammation in AD and other neurological disorders is YKL-40 (also named Chitinase 3-like I) [7, 8]. It has been described that YKL-40 participates in connective tissue cell growth, endothelial cell migration and inhibition of mammary epithelial cell differentiation and promotes tumour angiogenesis [9]. However, its biological and physiological functions in the central nervous system remain unclear. YKL-40 is increased in the CSF of multiple sclerosis (MS) patients, and YKL-40 levels correlate well with disease progression [10]. In previous studies, we and others [11, 12] found elevated YKL-40 levels in the CSF of AD patients. Interestingly, increased levels of YKL-40 were found in the preclinical stages of AD [13, 14] indicating that the immune system activation occurs early in the disease. These studies have also shown that CSF levels of YKL-40 and tau strongly correlate [11, 12]. Additional studies have indicated that YKL-40 is also elevated in the CSF of patients with other tauopathies, such as frontotemporal dementia (FTD), corticobasal degeneration (CBD) and progressive supranuclear palsy (PSP) [11, 13, 15, 16].

Despite the wide use of YKL-40 as a biochemical marker in neurodegenerative diseases, its distribution and pattern of expression in the human brain remains unclear. A recent study found that expression levels of chitinase genes in the brain regions of late onset AD (LOAD) patients are increased compared with healthy controls [17], but the cellular source of expression of these proteins remains uncertain. Some studies have suggested that YKL-40 is expressed in astrocytes in a variety of acute neuroinflammatory conditions, such as traumatic brain injury or multiple sclerosis [18, 19]. Other studies, however, support the idea that YKL-40 is also expressed in macrophage/microglia cell types in these conditions [19, 20]. In AD, studies show a variable pattern of YKL-40 expression that includes astrocytes, microglia or, on rare occasions, neurons [12, 21, 22]. The aim of the present work was to determine the cellular pattern of YKL-40 expression in the human brain tissue in AD and other tauopathies. We also investigated the relationship between YKL-40 expression and tau aggregates in these disorders.

In the present study, we found that YKL-40 is expressed by astrocytes in human brain tissue in healthy controls and in different neurodegenerative diseases. The immunoreactivity pattern of YKL-40 was mainly cytoplasmic extending to proximal astrocytic processes. We also observed that this protein is expressed in a subset of astrocytes (<10%) that do not contain tau aggregates in non-AD tauopathies.

To date, the pattern of expression of YKL-40 in the central nervous system has remained controversial and has not been fully elucidated. Some studies have described that YKL-40 is expressed in microglia [33, 34] while others have found it in astrocytes [12, 35]. These discrepancies may be because YKL-40 expression seems to vary depending on the disease and the severity of the neuroinflammatory response [19, 35]. It has been shown that in multiple sclerosis, YKL-40 is expressed by macrophages/microglial cells (CD68+) in low and high inflammatory activity lesions [19]. However, YKL-40 expression was also expressed in the cytoplasm of astrocytes (GFAP+) in high inflammatory activity lesions [19]. Another study reported that in human brain infarction, YKL-40 astrocytic expression depends on the stage of underlying inflammation, increasing during the acute inflammation phase and diminishing as the inflammation resolves [35]. Moreover, it has been shown by others that YKL-40 is also expressed by peripheral cells including chondrocytes [36], synoviocytes [37], vascular smooth muscle cells [38], macrophages [39] and neutrophils [40]. For example, it has been reported that expression of YKL-40 in breast cancer tissue correlates with tumour grade [41] and that YKL-40 macrophage expression is upregulated in patients with chronic obstructive pulmonary disease and correlates with its severity [42]. In vitro studies have shown a dramatically increase of YKL-40 expression during astrocyte differentiation [43]. Other studies based on immunohistochemical techniques suggested that YKL-40 together with SSEA-4 marker expression represent an unexplored astrogenic lineage [44]. Here, we have confirmed that YKL-40 shows an astroglial cytoplasmic immunoreactivity pattern in postmortem human frontal cortex from healthy controls and AD patients. We did not detect YKL-40 expression in microglia or in neurons.

We also found YKL-40 in the cytoplasm of astrocytes in non-AD tauopathies, including PiD, CBD and PSP. Although YKL-40 was typically found in isolated astrocytes, we occasionally observed focal astrocytic YKL-40 immunoreactivity around blood vessels. Previous studies have suggested that YKL-40 transcription is induced in astrocytes by proinflammatory factors released from macrophages [21]. One possible explanation for this observation is that perivascular macrophages may induce YKL-40 expression in astrocytes that are in close proximity. The colocalization between YKL-40 and GFAP was around 80% indicating that GFAP immunoreactivity, typically extending towards the astrocyte membrane and processes, surrounds that of YKL-40. Interestingly, YKL-40 immmunoreactivity was independent of tau, indicating that in non-AD tauopathies, YKL-40 is expressed in a tau-negative subset of astrocytes. Whether YKL-40 could be a compensatory response to inhibit tau aggregation or, on the contrary, represents an initial event that facilities tau deposition requires further investigation. As expected, colocalization between tau and GFAP was found in non-AD tauopathies reflecting the distinctive glial tau aggregation in these disorders [32]. The 15% overlap observed in AD may be explained by neuropil threads crossing astrocyte processes in close proximity that are measured as colocalization due to the resolution limits.

Recent studies have demonstrated that YKL-40 levels are increased in the CSF of patients with AD and FTD compared with those in the healthy controls [11‐16, 45, 46]. In addition, a positive correlation between YKL-40, total tau and p-tau has been reported in CSF, suggesting that inflammation and tau-associated neurodegeneration are related pathophysiological processes. In agreement, animal models and co-culture studies have shown that activated glia-induced neuronal tau phosphorylation and aggregation [47]. In postmortem brains, our semi-automated method revealed that total YKL-40 levels were statistically increased in all tauopathies (except PiD) compared with healthy controls. These results are in agreement with studies that investigated YKL-40 levels in CSF of AD and FTD patients [11‐16, 45, 46]. It is important to note that only a limited proportion of astrocytes (always less than 10%) expressed YKL-40 in human frontal cortex. Interestingly, we found a positive correlation between YKL-40 and tau pathology burden (r = 0.447) suggesting that inflammation and neurodegeneration may be closely related processes in humans. The lack of correlation between YKL-40 and GFAP together with the positive correlation between tau and GFAP (r = 0.651) in our study also may indicate that YKL-40 expression is independent of astrocyte activation in neurodegenerative disease.

In conclusion, this is, to our knowledge, the first detailed neuropathologic characterization of YKL-40 expression in human brain tissue. Moreover, the study includes tissue samples from healthy controls and four neurodegenerative diseases. Combining confocal microscopy and the application of a semi-automated method to quantify pathology burden, we have shown that the immunoreactivity pattern of YKL-40 in AD and other tauopathies is astroglial. YKL-40 is expressed by a subset of astrocytes that do not contain tau aggregates in non-AD tauopathies. Finally, we have found that YKL-40 inflammatory marker is associated with tau pathology in neurodegenerative diseases that accumulate tau.