Cerebrospinal Fluid Concentrations of Extracellular Matrix Proteins in Alzheimer's Disease

Soluble Proteoglycans and Proteoglycan Fragments as Biomarkers of Pathological Extracellular Matrix Remodeling

Proteoglycans and their proteolytic fragments diffuse into biological fluids such as plasma, serum, urine, or synovial fluid, where they can be detected by antibodies or mass-spectrometry. Neopeptides generated by the proteolysis of proteoglycans are recognized by specific neoepitope antibodies and can act as a proxy for the activity of certain proteases. Proteoglycan and proteoglycan fragments can be potentially used as prognostic, diagnostic, or theragnostic biomarkers for several diseases characterized by dysregulated extracellular matrix remodeling such as osteoarthritis, rheumatoid arthritis, atherosclerosis, thoracic aortic aneurysms, central nervous system disorders, viral infections, and cancer. Here, we review the main mechanisms accounting for the presence of soluble proteoglycans and their fragments in biological fluids, their potential application as diagnostic, prognostic, or theragnostic biomarkers, and highlight challenges and opportunities ahead of their clinical translation.

Challenges and Future Perspectives in Modeling Neurodegenerative Diseases Using Organ-on-a-Chip Technology

Neurodegenerative diseases (NDDs) affect more than 50 million people worldwide, posing a significant global health challenge as well as a high socioeconomic burden. With aging constituting one of the main risk factors for some NDDs such as Alzheimer's disease (AD) and Parkinson's disease (PD), this societal toll is expected to rise considering the predicted increase in the aging population as well as the limited progress in the development of effective therapeutics. To address the high failure rates in clinical trials, legislative changes permitting the use of alternatives to traditional pre-clinical in vivo models are implemented. In this regard, microphysiological systems (MPS) such as organ-on-a-chip (OoC) platforms constitute a promising tool, due to their ability to mimic complex and human-specific tissue niches in vitro. This review summarizes the current progress in modeling NDDs using OoC technology and discusses five critical aspects still insufficiently addressed in OoC models to date. Taking these aspects into consideration in the future MPS will advance the modeling of NDDs in vitro and increase their translational value in the clinical setting.

Serum Brevican as a Biomarker of Cerebrovascular Disease in an Elderly Cognitively Impaired Cohort

In the brain, the extracellular matrix (ECM) composition shapes the neuronal microenvironment and can undergo substantial changes with cerebral pathology. Brevican is integral to the formation of the ECM's neuroprotective perineuronal nets (PNNs). Decreased brevican levels were reported in vascular dementia (VaD) but not in Alzheimer's disease (AD). However, the status of brevican in clinical cohorts with high concomitance of AD pathological burden and cerebrovascular disease (CeVD) is unclear. In this study, 32 non-cognitively impaired (NCI), 97 cognitively impaired no dementia (CIND), 46 AD, and 23 VaD participants recruited from memory clinics based in Singapore underwent neuropsychological and neuroimaging assessments, together with measurements of serum brevican. Association analyses were performed between serum brevican and neuroimaging measures of CeVDs, including white matter hyperintensities (WMHs), lacunes, cortical infarcts, and cerebral microbleeds. Using an aggregated score for CeVD burden, only CIND participants showed lower brevican levels with higher CeVD compared to those with lower CeVD burden (p = 0.006). Among the CeVD subtypes assessed, only elevated WMH burden was associated with lower brevican levels (OR = 2.7; 95% CI = 1.3-5.5). Our findings suggest that brevican deficits may play a role in early cerebrovascular damage in participants at risk of developing dementia.

Extracellular Matrix Changes in Subcellular Brain Fractions and Cerebrospinal Fluid of Alzheimer’s Disease Patients

The brain’s extracellular matrix (ECM) is assumed to undergo rearrangements in Alzheimer’s disease (AD). Here, we investigated changes of key components of the hyaluronan-based ECM in independent samples of post-mortem brains (N = 19), cerebrospinal fluids (CSF; N = 70), and RNAseq data (N = 107; from The Aging, Dementia and TBI Study) of AD patients and non-demented controls. Group comparisons and correlation analyses of major ECM components in soluble and synaptosomal fractions from frontal, temporal cortex, and hippocampus of control, low-grade, and high-grade AD brains revealed a reduction in brevican in temporal cortex soluble and frontal cortex synaptosomal fractions in AD. In contrast, neurocan, aggrecan and the link protein HAPLN1 were up-regulated in soluble cortical fractions. In comparison, RNAseq data showed no correlation between aggrecan and brevican expression levels and Braak or CERAD stages, but for hippocampal expression of HAPLN1, neurocan and the brevican-interaction partner tenascin-R negative correlations with Braak stages were detected. CSF levels of brevican and neurocan in patients positively correlated with age, total tau, p-Tau, neurofilament-L and Aβ1-40. Negative correlations were detected with the Aβ ratio and the IgG index. Altogether, our study reveals spatially segregated molecular rearrangements of the ECM in AD brains at RNA or protein levels, which may contribute to the pathogenic process.

The Role of Extracellular Matrix in Human Neurodegenerative Diseases

The dense neuropil of the central nervous system leaves only limited space for extracellular substances free. The advent of immunohistochemistry, soon followed by advanced diagnostic tools, enabled us to explore the biochemical heterogeneity and compartmentalization of the brain extracellular matrix in exploratory and clinical research alike. The composition of the extracellular matrix is critical to shape neuronal function; changes in its assembly trigger or reflect brain/spinal cord malfunction. In this study, we focus on extracellular matrix changes in neurodegenerative disorders. We summarize its phenotypic appearance and biochemical characteristics, as well as the major enzymes which regulate and remodel matrix establishment in disease. The specifically built basement membrane of the central nervous system, perineuronal nets and perisynaptic axonal coats can protect neurons from toxic agents, and biochemical analysis revealed how the individual glycosaminoglycan and proteoglycan components interact with these molecules. Depending on the site, type and progress of the disease, select matrix components can either proactively trigger the formation of disease-specific harmful products, or reactively accumulate, likely to reduce tissue breakdown and neuronal loss. We review the diagnostic use and the increasing importance of medical screening of extracellular matrix components, especially enzymes, which informs us about disease status and, better yet, allows us to forecast illness.

Potential neurotoxic activity of diverse molecules released by astrocytes

Astrocytes are the main support cells of the central nervous system. They also participate in neuroimmune reactions. In response to pathological and immune stimuli, astrocytes transform to reactive states characterized by increased release of inflammatory mediators. Some of these molecules are neuroprotective and inflammation resolving while others, including reactive oxygen species (ROS), nitric oxide (NO), matrix metalloproteinase (MMP)- 9, L-glutamate, and tumor necrosis factor α (TNF), are well-established toxins known to cause damage to surrounding cells and tissues. We hypothesized that similar to microglia, the brain immune cells, reactive astrocytes can release a broader set of diverse molecules that are potentially neurotoxic. A literature search was conducted to identify such molecules using the following two criteria: 1) evidence of their expression and secretion by astrocytes and 2) direct neurotoxic action. This review describes 14 structurally diverse molecules as less-established astrocyte neurotoxins, including C-X-C motif chemokine ligand (CXCL)10, CXCL12/CXCL12(5-67), FS-7-associated surface antigen ligand (FasL), macrophage inflammatory protein (MIP)- 2α, TNF-related apoptosis inducing ligand (TRAIL), pro-nerve growth factor (proNGF), pro-brain-derived neurotrophic factor (proBDNF), chondroitin sulfate proteoglycans (CSPGs), cathepsin (Cat)B, group IIA secretory phospholipase A₂ (sPLA₂-IIA), amyloid beta peptides (Aβ), high mobility group box (HMGB)1, ceramides, and lipocalin (LCN)2. For some of these molecules, further studies are required to establish either their direct neurotoxic effects or the full spectrum of stimuli that induce their release by astrocytes. Only limited studies with human-derived astrocytes and neurons are available for most of these potential neurotoxins, which is a knowledge gap that should be addressed in the future. We also summarize available evidence of the role these molecules play in select neuropathologies where reactive astrocytes are a key feature. A comprehensive understanding of the full spectrum of neurotoxins released by reactive astrocytes is key to understanding neuroinflammatory diseases characterized by the adverse activation of these cells and may guide the development of novel treatment strategies.

A Novel Panel of Plasma Proteins Predicts Progression in Prodromal Alzheimer's Disease

BACKGROUND

A cheap and minimum-invasive method for early identification of Alzheimer's disease (AD) pathogenesis is key to disease management and the success of emerging treatments targeting the prodromal phases of the disease.

OBJECTIVE

To develop a machine learning-based blood panel to predict the progression from mild cognitive impairment (MCI) to dementia due to AD within a four-year time-to-conversion horizon.

METHODS

We created over one billion models to predict the probability of conversion from MCI to dementia due to AD and chose the best-performing one. We used Alzheimer's Disease Neuroimaging Initiative (ADNI) data of 379 MCI individuals in the baseline visit, from which 176 converted to AD dementia.

RESULTS

We developed a machine learning-based panel composed of 12 plasma proteins (ApoB, Calcitonin, C-peptide, CRP, IGFBP-2, Interleukin-3, Interleukin-8, PARC, Serotransferrin, THP, TLSP 1-309, and TN-C), and which yielded an AUC of 0.91, accuracy of 0.91, sensitivity of 0.84, and specificity of 0.98 for predicting the risk of MCI patients converting to dementia due to AD in a horizon of up to four years.

CONCLUSION

The proposed machine learning model was able to accurately predict the risk of MCI patients converting to dementia due to AD in a horizon of up to four years, suggesting that this model could be used as a minimum-invasive tool for clinical decision support. Further studies are needed to better clarify the possible pathophysiological links with the reported proteins.

Brevican and Neurocan Cleavage Products in the Cerebrospinal Fluid - Differential Occurrence in ALS, Epilepsy and Small Vessel Disease

The neural extracellular matrix (ECM) composition shapes the neuronal microenvironment and undergoes substantial changes upon development and aging, but also due to cerebral pathologies. In search for potential biomarkers, cerebrospinal fluid (CSF) and serum concentrations of brain ECM molecules have been determined recently to assess ECM changes during neurological conditions including Alzheimer’s disease or vascular dementia. Here, we measured the levels of two signature proteoglycans of brain ECM, neurocan and brevican, in the CSF and serum of 96 neurological patients currently understudied regarding ECM alterations: 16 cases with amyotrophic lateral sclerosis (ALS), 26 epilepsy cases, 23 cerebral small vessel disease (CSVD) patients and 31 controls. Analysis of total brevican and neurocan was performed via sandwich Enzyme-linked immunosorbent assays (ELISAs). Major brevican and neurocan cleavage products were measured in the CSF using semiquantitative immunoblotting. Total brevican and neurocan concentrations in serum and CSF did not differ between groups. The 60 kDa brevican fragment resulting from cleavage by the protease ADAMTS-4 was also found unchanged among groups. The presumably intracellularly generated 150 kDa C-terminal neurocan fragment, however, was significantly increased in ALS as compared to all other groups. This group also shows the highest correlation between cleaved and total neurocan in the CSF. Brevican and neurocan levels strongly correlated with each other across all groups, arguing for a joint but yet unknown transport mechanism from the brain parenchyma into CSF. Conclusively our findings suggest an ALS-specific pattern of brain ECM remodeling and may thus contribute to new diagnostic approaches for this disorder.

Brevican and Neurocan Peptides as Potential Cerebrospinal Fluid Biomarkers for Differentiation Between Vascular Dementia and Alzheimer's Disease

BACKGROUND

Brevican and neurocan are central nervous system-specific extracellular matrix proteoglycans. They are degraded by extracellular enzymes, such as metalloproteinases. However, their degradation profile is largely unexplored in cerebrospinal fluid (CSF).

OBJECTIVE

The study aim was to quantify proteolytic peptides derived from brevican and neurocan in human CSF of patients with Alzheimer's disease (AD) and vascular dementia (VaD) compared with controls.

METHODS

The first cohort consisted of 75 individuals including 25 patients with AD, 7 with mild cognitive impairment (MCI) diagnosed with AD upon follow-up, 10 patients with VaD or MCI diagnosed with VaD upon follow-up, and 33 healthy controls and cognitively stable MCI patients. In the second cohort, 31 individuals were included (5 AD patients, 14 VaD patients and 12 healthy controls). Twenty proteolytic peptides derived from brevican (n = 9) and neurocan (n = 11) were quantified using high-resolution parallel reaction monitoring mass spectrometry.

RESULTS

In the first cohort, the majority of CSF concentrations of brevican and neurocan peptides were significantly decreased inVaDas compared withADpatients (AUC = 0.83.0.93, p≤0.05) and as compared with the control group (AUC = 0.79.0.87, p ≤ 0.05). In the second cohort, CSF concentrations of two brevican peptides (B87, B156) were significantly decreased in VaD compared with AD (AUC = 0.86.0.91, p ≤ 0.05) and to controls (AUC = 0.80.0.82, p ≤ 0.05), while other brevican and neurocan peptides showed a clear trend to be decreased in VaD compared with AD (AUC = 0.64.80, p > 0.05). No peptides differed between AD and controls.

CONCLUSION

Brevican and neurocan peptides are potential diagnostic biomarkers for VaD, with ability to separate VaD from AD.

Neuroligin-1 in brain and CSF of neurodegenerative disorders: investigation for synaptic biomarkers

Synaptic pathology is a central event in Alzheimer’s disease (AD) and other neurodegenerative conditions, and investigation of synaptic proteins can provide valuable tools to follow synaptic dysfunction and loss in these diseases. Neuroligin-1 (Nlgn1) is a postsynaptic cell adhesion protein, important for synapse stabilization and formation. Nlgn1 has been connected to cognitive disorders, and specifically to AD, as target of the synaptotoxic effect of amyloid-β (Aβ) oligomers and Aβ fibrils. To address changes in Nlgn1 expression in human brain, brain regions in different neurological disorders were examined by Western blot and mass spectrometry. Brain specimens from AD (n = 23), progressive supranuclear palsy (PSP, n = 11), corticobasal degeneration (CBD, n = 10), and Pick’s disease (PiD, n = 9) were included. Additionally, cerebrospinal fluid (CSF) samples of AD patients (n = 43) and non-demented controls (n = 42) were analysed. We found decreased levels of Nlgn1 in temporal and parietal cortex (~ 50–60% reductions) in AD brains compared with controls. In frontal grey matter the reduction was not seen for AD patients; however, in the same region, marked reduction was found for PiD (~ 77%), CBD (~ 66%) and to a lesser extent for PSP (~ 43%), which could clearly separate these tauopathies from controls. The Nlgn1 level was reduced in CSF from AD patients compared to controls, but with considerable overlap. The dramatic reduction of Nlgn1 seen in the brain extracts of tauopathies warrants further investigation regarding the potential use of Nlgn1 as a biomarker for these neurodegenerative diseases.

Disruption of rat deep cerebellar perineuronal net alters eyeblink conditioning and neuronal electrophysiology

The perineuronal net (PNN) is a specialized type of extracellular matrix found in the central nervous system. The PNN forms on fast spiking neurons during postnatal development but the ontogeny of PNN development has yet to be elucidated. By studying the development and prevalence of the PNN in the juvenile and adult rat brain, we may be able to understand the PNN's role in development and learning and memory. We show that the PNN is fully developed in the deep cerebellar nuclei (DCN) of rats by P18. By using enzymatic digestion of the PNN with chondroitinase ABC (ChABC), we are able to study how digestion of the PNN affects cerebellar-dependent eyeblink conditioning in vivo and perform electrophysiological recordings from DCN neurons in vitro. In vivo degradation of the PNN resulted in significant differences in eyeblink conditioning amplitude and area. Female animals in the vehicle group demonstrated higher levels of conditioning as well as significantly higher post-probe conditioned responses compared to males in that group, differences not present in the ChABC group. In vitro, we found that DCN neurons with a disrupted PNN following exposure to ChABC had altered membrane properties, fewer rebound spikes, and decreased intrinsic excitability. Together, this study further elucidates the role of the PNN in cerebellar learning in the DCN and is the first to demonstrate PNN degradation may erase sex differences in delay conditioning.

Downregulation of synapse-associated protein expression and loss of homeostatic microglial control in cerebrospinal fluid of infectious patients with delirium and patients with Alzheimer's disease

Delirium is a complex and multifactorial condition associated with long-term cognitive decline. Due to the strong links between systemic inflammation, delirium and dementia we hypothesized that responses within the brain in patients who develop delirium could show biochemical overlap with patients with Alzheimer's disease (AD). In this observational study we analyzed protein expression signatures in cerebrospinal fluid (CSF) from 15 patients with infectious delirium and compared these to 29 patients with AD, 30 infectious patients without delirium and 15 non-infectious controls free of neurological disease. A proximity extension assay was performed measuring a total of 184 inflammatory and neurology-related proteins. Eight inflammatory proteins (4%), including the key neuron-microglia communication marker CX3CL1 (fractalkine), were significantly upregulated in both delirium and AD, compared to infectious patients without delirium. Likewise, 23 proteins (13%) showed downregulation in both delirium and AD, relative to infectious patients without delirium, which interestingly included CD200R1, another neuron-microglia communication marker, as well as a cluster of proteins related to synapse formation and function. Synaptopathy is an early event in AD and correlates strongly with cognitive dysfunction. These results were partially mediated by aging, which is an important predisposing risk factor among many others for both conditions. Within this study we report the first in vivo human evidence suggesting that synapse pathology and loss of homeostatic microglial control is involved in the pathophysiology of both infectious delirium and AD and thus may provide a link for the association between infections, delirium and long-term cognitive decline.

The CNS-specific proteoglycan, brevican, and its ADAMTS4-cleaved fragment show differential serological levels in Alzheimer's disease, other types of dementia and non-demented controls: A cross-sectional study

Evidence indicate that the brain-specific protein, brevican, is proteolytically cleaved during neurodegeneration, hence positioning fragments of brevican as potential blood biomarkers of neurodegenerative diseases, such as dementia. We aimed to develop two assays capable of detecting the brevican N-terminal (N-Brev) and the ADAMTS4-generated fragment (Brev-A), cleaved at Ser401, in serum and to perform a preliminary assessment of their diagnostic potential in dementias. Monoclonal antibodies against N-Brev and Brev-A were used to develop two ELISAs detecting each epitope. A comparison of brevican fragments in serum from individuals with AD (n = 28), other dementia (OD) (n = 41), and non-dementia-related memory complaints (NDCs) (n = 48) was conducted. Anti-N-Brev and anti-Brev-A antibodies selectively recognized their targets and dilution and spike recoveries were within limits of ±20%. Intra- and inter-assay CVs were below limits of 10% and 15%, respectively. For the N-Brev biomarker, serum from patients with OD showed significantly lower levels than those with AD (p = 0.05) and NDCs (p < 0.01). The opposite pattern was evident for Brev-A: serum levels in patients with OD were significantly higher than for AD (p = 0.04) and NDCs (p = 0.01). For both N-Brev and Brev-A, levels did not differ between AD and NDCs. The ratio of N-Brev/Brev-A resulted in increased significant differences between OD and AD (p < 0.01) and between OD and NDCs (p < 0.0001). The ratio discriminated between NDCs and OD (AUC: 0.75, 95% CI: 0.65-0.85, p < 0.0001) and between OD and AD (AUC: 0.72, 95% CI: 0.59-0.85, p < 0.01). In conclusion, we developed the first assays detecting the N-terminal of brevican as well as an ADAMTS4-cleaved fragment of brevican in blood. Differential levels of N-Brev and Brev-A between AD and OD allow for these biomarkers to possibly distinguish between different forms of dementias.