Novel method

Abstract

The invention relates to methods of analysing one or more protein complexes in body fluid samples using ultra-sensitive techniques such as single molecule pulldown. The methods find particular use in detecting protein complex biomarkers for the detection or diagnosis of neurodegenerative disorders. The invention also relates to novel combination biomarkers.

Description

[0001] NOVEL METHOD

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to methods of analysing protein complexes, particularly for use in methods of detecting protein complex biomarkers for use in the detection or diagnosis of neurodegenerative disorders. The invention also relates to novel combination biomarkers.

[0004] BACKGROUND OF THE INVENTION

[0005] Parkinson’s disease and Alzheimer’s disease (PD and AD) are the most prevalent neurodegenerative disorders, affecting 55 million people worldwide (Hampel, H. et al. (2021); Aarsland, D. et al. (2021)). Although diagnosis of well-established disease can be accurate, based on clinical features and supportive diagnostic imaging, there is a pressing need for quantitative biological tools that work at the early stage disease, allowing diagnostics and recruitment for early phase clinical trials (Porsteinsson, A. P. et al. (2021)). There are diseasespecific protein aggregates of a-synuclein (a-syn) in PD and amyloid-p (Ap) and tau in AD, and both diseases are also characterised by chronic inflammation, which precedes and predicts clinical progression. The development of sensitive biomarkers of inflammation has potential to improve early diagnosis particularly if used in conjunction with protein aggregation markers, and also to support stratification and monitoring in immunotherapeutic clinical trials (Angiulli, F. et al. (2021); Zimmermann, M. & Brockmann, K. (2022)). Inflammatory cytokines show low grade elevation in the blood in mild cognitive impairment (MCI), AD and PD and may predict faster disease progression, but cannot distinguish reliably between the disease state and controls (Zimmermann, M. & Brockmann, K. (2022); King, E. et al.; Chatterjee, P. et al. (2021); Williams-Gray, C. H. et al. (2016)). However specific inflammatory molecules may be of more utility, such as the complement system component regulator, clusterin, which is elevated in prodromal AD ahead of conversion to clinically manifest disease (Hakobyan, S. et al. (2016)). Biomarker identification remains challenging due to a combination of extremely low concentration (e.g. plasma tumour necrosis factor (TNF)-a ~7 pg / mL, plasma Ap42 ~20 pg / mL (Julian, A. et al. (2015); Janelidze, S. et al. (2016)) and heterogeneity in both size and structure of candidate protein biomarkers; many of which form aggregates in disease rather than monomers.

[0006] New methods are required which are sensitive and specific enough to identify the pathological species from the population of aggregates present in samples. ELISA and single-molecule array assays have been developed and utilised for detecting picomolar concentrations of monomeric proteins in the blood and cerebrospinal fluid (CSF) (Chong, J. R. et al. (2021)). However, these techniques do not distinguish aggregates over monomers or aggregates with different sizes and shapes.

[0007] There is a need in the art to provide improved methods and biomarkers for the detection of neurodegenerative diseases.

[0008] SUMMARY OF THE INVENTION

[0009] According to a first aspect of the invention, there is provided a method of analysing one or more protein complexes in a blood, serum or plasma sample comprising:

[0010] (a) performing an assay to quantify the level of the one or more protein complexes present in the sample, wherein the assay comprises a capture binding agent and a detection binding agent which specifically bind to the protein complex, wherein the same binding agent is used as the capture binding agent and the detection binding agent; and

[0011] (b) performing a method for assessing one or more morphological features of the protein complexes present in the sample, wherein the combination of information obtained from performing steps (a) and (b) is used to analyse the protein complex.

[0012] According to a further aspect of the invention, there is provided a method for determining whether a subject has, or is at risk of developing, a neurodegenerative disorder, the method comprising performing the method as defined herein to analyse one or more protein complexes in a blood, serum or plasma sample obtained from the subject.

[0013] According to a further aspect of the invention, there is provided a combined biomarker for use in the detection or prognosis of a neurodegenerative disorder, wherein the combined biomarker comprises two types of protein complexes present in a sample, wherein one of the protein complexes is ASC speck.

[0014] According to a further aspect of the invention, there is provided a combined biomarker for use in the detection or prognosis of a neurodegenerative disorder, wherein the combined biomarker comprises the sum of the proportion of two or more protein aggregates that form during the development of neurodegenerative disease.

[0015] According to a further aspect of the invention, there is provided a method of diagnosing a neurodegenerative disorder in a subject, comprising: (a) measuring the combined biomarker as defined herein in a sample obtained from the subject; and

[0016] (b) using the measurement detected for the combined biomarker to determine if the subject has, or is at risk of developing, a neurodegenerative disorder.

[0017] According to a further aspect of the invention, there is provided a method of monitoring the efficacy of a therapy in a subject having, suspected of having, or being predisposed to a neurodegenerative disorder in a subject, comprising:

[0018] (a) measuring the combined biomarker as defined herein in a sample obtained from the subject; and

[0019] (b) using the measurement detected for the combined biomarker to determine if the subject has, or is at risk of developing, a neurodegenerative disorder.

[0020] BRIEF DESCRIPTION OF THE FIGURES

[0021] Figure 1 : Single ASC speck imaging detection assay: ASC-SiMPull. (A) Workflow of the SiM Pull method for detecting ASC specks in human biofluids. The same anti-ASC (AL177) antibody was used for capture and detection. (D, E) Time course quantification of ASC specks detected in the lysates (D) and media (E) from inflammasome-activated (LPS+Nigericin for 3.75, 7.5, 15, 30, 60 mins) THP-1 macrophages vs non-stimulated controls (LPS only). Dots in D and E represent independent replicates (n = 3). The data are shown as mean ± SD obtained from triplicate measurements, and the lower and upper boundaries of the box indicate the 25th and 75th percentiles, respectively. Student’s t-test: * p < 0.05, ** p < 0.01. (B, C) Example super-resolution (dSTORM) images of ASC specks of different sizes and shapes detected in the THP-1 cell lysates (B) and media (C). Scale bar, 0.2 pm. Comparison in the fraction of individual ASC aggregates with (F) area > 0.03 pm2and (G) circularity > 0.9 detected in the conditioned lysates vs media with dSTORM.

[0022] Figure 2: Detection of ASC specks in human serum, CSF and soaked brain from people with PD, AD and FTD compared to age-matched controls using SiMPull assay. Quantification of the number of ASC specks per field of view (FOV) detected in serum (A,B) and CSF (F) from PD patients vs controls (cohort 1 : n = 10 PD vs 10 HC serum, cohort 2: n = 8 PD vs 5 HC serum, n = 6 PD vs 6 HC CSF) and (G) in the amygdala of soaked post-mortem brains from 10 patients with PD versus 5 non-demented controls. (D) Quantification of the number of ASC specks per FOV detected in serum from early AD (n = 20), AD dementia (n = 20) and FTD (n = 10) patients compared to age-matched controls (n = 30 HC) and (H) in the frontal cortex (FC) vs hippocampus (HPC) vs the visual association cortex (VAC) of soaked brains from 3 Braak stage III patients with AD. The data are shown as mean ± SD, and the lower and upper boundaries of the box indicate the 25th and 75th percentiles, respectively. Permutation (exact) test: * p < 0.05, ** p < 0.01 **** p < 0.0005. (C, E) ROC curve analysis for quantity detection of ASC specks in human serum as a promising marker of inflammation discriminating controls from people with neurodegenerative disease and different dementia types (AD dementia vs FTD) with high specificity and sensitivity: AUC = 83% in PD vs HC serum cohort 1 (C), AUC = 100% in PD vs HC serum cohort 2 (C), 64% in early AD vs HC serum (E), 88% in AD dementia vs HC serum (E), and 81 % in AD dementia vs FTD serum (E).

[0023] (I) Denaturation curve of ASC specks present in PD human brain homogenate and serum sample when treated with increasing concentration of Gdn HCI varying from 0.05 to 4 M.

[0024] (J) Specificity controls for ASC speck detection in SiMPull assay using a no capture and correct detection antibody (no capture control) and a correct capture and non-target IgG isotype control detection antibody (detection control) for PD serum, CSF and soaked brain samples. Error bars are mean ± STD from two or three PD patients shown in Iog2 scale. PD = Parkinson’s disease, AD = Alzheimer’s disease, FTD = Frontotemporal dementia, HC = controls.

[0025] Figure 3: Combination of single-aggregate measurements of ASC specks with Ap aggregates and p-tau aggregates in early AD (n = 20), AD dementia (n = 20), FTD (n = 10) and HC (n = 30) serum samples and with a-syn aggregates and Ap aggregates in the two early-stage PD serum cohorts (cohort 1 : n = 10 PD and 10 HC and cohort 2: n = 8 PD and 5 HC). (A) ASC / Ap and (D) (ASC + p-tau-AT8) / Ap ratios as candidate composite biomarkers between early AD serum vs controls, and between AD dementia serum vs controls; (A) ASC / Ap, (C) (ASC / ptau-AT8) and (D) (ASC + p-tau-AT8) / Ap ratios as differential composite biomarkers distinguishing AD dementia from FTD serum; (G, H) ASC / Ap and (J, K) (ASC + cr- syn) / Ap ratios as candidate composite biomarkers between early-stage PD serum and controls. References to “ptau-AT8” represents p-tau aggregates phosphorylated at positions 202 / 205. The data in (G,H) are plotted in Iog10 scale. The data are shown as mean ± SD, and the lower and upper boundaries of the box indicate the 25th and 75th percentiles, respectively. Permutation (exact) test: ** p < 0.01 , *** p < 0.005,**** p < 0.0005,***** p < 0.00005. (B,E,F,I, L) ROC curve analysis using the corresponding metrics.

[0026] Figure 4: Morphology analysis of ASC specks extracted from the amygdala of human PD

[0027] (n = 9) and non-demented control (n = 5) post-mortem soaked brain using STORM. (A) Comparison of the fraction of individual ASC aggregates that are smaller (area < 0.018 pm2) and rounder (circularity > 0.8) than the defined threshold in the brain of people with PD vs controls. We identified the area (size) and circularity (shape) threshold (maximum statistically significant difference in the size and shape of ASC speck histograms between diagnostic groups) giving us the morphological phenotype of ASC specks which is increased in PD brains. (B) ROC curve analysis of the identified phenotype. (C,D) Examples of super-resolved ASC aggregates in PD (C) and control (D) brain samples. Cumulative size (E) and shape (G) distributions of ASC specks for PD vs control brains. Difference between PD and control cumulative size (F) and shape (H) distributions retrieved from (E and G). The dotted line indicates 99% confidence using the Kolmogorov-Smirnov statistical test.

[0028] Figure 5: Morphology analysis of ASC specks detected in the control (n = 8) and PD (n = 8) serum from people with early-stage disease (within 6 months of diagnosis) using dSTORM. (A) Comparison in the fraction of individual ASC aggregates that are smaller (area < 0.05 pm2) and rounder (circularity > 0.5) than the defined threshold in the serum of people with PD vs controls. (B) ROC curve analysis of the identified phenotype. (D, E) The (morphologically distinctive ASC fraction + total ASC) / A / 3 as a candidate composite biomarker in PD serum and its ROC curve analysis. The data are shown as mean ± SD, and the lower and upper boundaries of the box indicate the 25th and 75th percentiles, respectively. Permutation (exact) test: *** p < 0.005. (C, F) Examples of super-resolved ASC aggregates in PD (C) and control (F) serum samples. Cumulative size (D) and shape (I) distributions of ASC specks for PD vs control serum. Difference between PD and control cumulative size (H) and shape (J) distributions retrieved from D and I. The dotted line indicates 99% confidence using the Kolmogorov-Smirnov statistical test.

[0029] Figure 6: Morphology analysis of ASC specks detected in the control (n = 9) and AD (n = 9) serum using dSTORM. (A) Comparison in the fraction of individual ASC aggregates that are smaller (area < 0.04 pm2) and rounder (circularity > 0.75) than the defined threshold in the serum of people with AD vs controls. (B) ROC curve analysis of the identified phenotype. (D, E) The total ASC divided by the morphologically distinct fraction of ASC specks as a candidate composite biomarker in AD serum and its ROC curve analysis. The data are shown as mean ± SD, and the lower and upper boundaries of the box indicate the 25th and 75th percentiles, respectively. Permutation (exact) test: ** p < 0.01 , **** p < 0.0005. (C, F) Examples of superresolved ASC aggregates in AD (C) and control (F) serum samples. Cumulative size (G) and shape (I) distributions of ASC specks for AD vs control serum. Difference between AD and control cumulative size (H) and shape (J) distributions retrieved from G and I. The dotted line indicates 99% confidence using the Kolmogorov-Smirnov statistical test. Figure ?: Diffraction limited images of coupled silica nanoparticles showing no significant aggregation. (A) a-syn calibrators at 10 nM detected using SC211-SiMPull. (B) Ap42 calibrators at 1nM detected using 6E10-SiMPull. (C) tau calibrators at 1 nM detected using HT7-SiMPull. Scale bar = 10 pm.

[0030] Figure 8: Single-Molecule Array for the detection of protein aggregates. (A) Standard curve of a-syn calibrator using the 4B12-4B12 antibody pair. (B) Standard curve of Ap calibrator using the 6E10-6E10 antibody pair. (C) Standard curve of tau lysate calibrator using the HT7-HT7 antibody pair. (D) Standard curve of tau lysate calibrator using the AT8- AT8 antibody pair. Panels show the mean of n = 4 technical replicates.

[0031] Figure 9: Validation of SIMOA assays for the detection of protein aggregates. (A) SIMOA assays were tested for cross-reactivity against other protein aggregates to ensure specificity for the respective protein aggregate type. (B) Denaturation of the protein aggregates with increasing concentrations of guanidinium chloride (0-6 M) to test the specificity of the SIMOA assays for protein aggregates as opposed to monomers. (C) Accuracy of the aggregate assays across the working range for a-synuclein aggregates (4B12 antibody pair), Ap aggregates (6E10 antibody pair), tau aggregates (HT7 antibody pair), p-tau aggregates (AT8 antibody pair). Panel A shows the mean of n = 3 technical replicates. Statistical analysis was conducted using one-way ANOVA and Tukey’s multiple comparisons test. Panel A shows the mean ± S.D. of n = 3 technical replicates, B shows the mean of n = 2 technical replicates, C shows the mean ± S.D. of n = 4 technical replicates, ns: p > 0.05, ****: p < 0.0001.

[0032] Figure 10: Aggregate levels in brain homogenate from AD and control patients. (A) Schematic of post-mortem brain tissue samples used in the study. (B-G) Aggregate levels detected in brain homogenate from AD (frontal cortex, Braak Stage VI,) and control patients (frontal cortex, Braak Stage 0) using aggregate SIMOA assays for (B) a-synuclein aggregates (SC211 antibody pair), (C) a-synuclein aggregates (4B12 antibody pair), (D) Ap aggregates (6E10 antibody pair), (E) tau aggregates (HT7 antibody pair), (F) p-tau aggregates (AT8 antibody pair). (G) The ratio of p-tau to total tau aggregates was determined showing improved separation between AD and control patients. Each data point in the plot represents the mean of 2 technical replicates. Panel A-G show the mean of n = 5 AD and n = 5 control patients from n = 2 technical replicates. Statistical analysis was conducted using Welch’s t-test. ns: p > 0.05, *: p < 0.05, **** p < 0.0001. Figure 11 : Quantification of protein aggregates in human serum. Aggregate levels detected in the serum of early AD and control patients using aggregate SIMOA assays for (A) a-synuclein aggregates (SC211 antibody pair, 15 data points not shown on graph due to values being zero), (B) a-synuclein aggregates (4B12 antibody pair), (C) Ap aggregates (6E10 antibody pair), (D) tau aggregates (HT7 antibody pair), (E) p-tau aggregates (AT8 antibody pair). (F) Ratio of p-tau to total tau aggregates. Each data point in the plot represents the mean of 2 technical replicates. Panels show the mean ± S.D. of n = 20 AD and n = 20 control patients. Statistical analysis was conducted using a t-test. ns: p > 0.05, **: p < 0.01.

[0033] Figure 12: Detection of brain-derived tau aggregates in plasma from AD and non-AD patients on the SiMPull platform.

[0034] Figure 13: The size of ASC specks as a sensitive biomarker in early-stage AD CSF. Morphology analysis of ASC specks detected in the control (n = 14) and early-stage AD (n = 14) CSF using dSTORM. (A,B) Comparison in the number (A) and fraction (B) of individual ASC aggregates that are smaller (area < 0.03 pm2) and rounder (circularity > 0.5) than the defined threshold in the CSF of people with early AD vs age-matched controls. (E,F) ROC curve analysis of the identified phenotype. Data are presented as box plots (centre line at the median, upper bound at 75th percentile, lower bound at 25th percentile) with whiskers at minimum and maximum values. Each dot represents one participant. Permutation (exact) test: *** p < 0.001 , **** p < 0.0001. (C) Examples of super-resolved ASC aggregates in AD and control CSF samples. The examples selected here are representatives of all three replicates. (D) Correlation in the number of ASC specks in paired serum versus CSF samples from 20 early AD patients and 20 HC participants.

[0035] Figure 14: The size of ASC specks as a candidate differential biomarker between FTD and AD. Morphology analysis of ASC specks detected in the FTD (n = 10) and AD (n = 10) serum using dSTORM. (A,D) Comparison in the fraction of individual ASC aggregates with area > 0.0122 pm2 (A) or circularity > 0.88 (D) in the serum of people with FTD syndromes vs AD. (B,E) ROC curve analysis of the identified phenotypes allowing to distinguish FTD from AD with AUC = 86% using the area metric and AUC = 81% using the circularity metric of ASC specks. Data are presented as box plots (centre line at the median, upper bound at 75th percentile, lower bound at 25th percentile) with whiskers at minimum and maximum values. Each dot represents one participant. Permutation (exact) test: * p < 0.05. Cumulative size (C) and shape (F) distributions of ASC specks for FTD vs AD serum. Figure 15: ASC / ptau ratio improves differential diagnosis between FTD and AD serum whereas (ASC + ptau) / A better distinguish different FTD syndromes. Single-molecule detection of ASC specks (A) and p-tau aggregates phosphorylated at positions 202 / 205 (AT8) (B) in frontotemporal dementia (FTD) (n = 10, cohort described in Example 8) and AD (n = 10 AD dementia) serum using SiMPull assay. (C) (ASC + p-tau-AT8) / Ap ratio as candidate differential biomarker between different FTD syndromes and a corresponding ROC curve; (D) ASC / p-tau-AT8 ratio as candidate differential biomarker between FTD and AD and a corresponding ROC curve. Data are presented as box plots (centre line at the median, upper bound at 75th percentile, lower bound at 25th percentile) with whiskers indicating STD. Each dot represents one participant. Permutation (exact) test: * p < 0.05, ** p < 0.01. FTLD = frontotemporal lobar degeneration, which is also known as FTD.

[0036] DETAILED DESCRIPTION OF THE INVENTION

[0037] The present invention provides an optimised method and combination biomarker for detecting neurodegenerative diseases, particularly at early stages. The inventors have used ultrasensitive assay methods for single-aggregate detection, in particular of inflammasome ASC specks in human biofluids. Using the same ASC antibody for capture and detection, only species larger than dimers are detected which enables inflammasome-active aggregated ASC complexes / specks, rather than inactive monomers present in the cell-free supernatant (the amount of ASC specks released by the cells), to be specifically measured. Combining ASC speck detection with analysis of Ap, p-tau and a-syn aggregates using the optimised method revealed the improved diagnostic accuracy of their different combinations (e.g.: ASC / Ap, ASC / p-tau, (p-tau + ASC) / Ap, a-syn + ASC, (a-syn + ASC) / Ap, etc.). The results establish the utility of ASC as a sensitive and specific biomarker at identifying disease stage when combined with other biomarkers of protein aggregates.

[0038] METHODS OF ANALYSING PROTEIN COMPLEXES

[0039] The misfolding and aggregation of proteins and peptides plays an important role in the pathogenesis of neurodegenerative diseases. For example, in Alzheimer’s disease (AD), betaamyloid (Ap) aggregates form extra-cellular plaques and hyperphosphorylated tau aggregates form intra-cellular tangles. In Parkinson’s disease (PD), alpha-synuclein (a-syn) aggregates form intra-cellular inclusions called Lewy Bodies. While these larger structures have been studied in AD and PD, a body of work suggests that the smaller, sub-micron aggregates formed during the aggregation process are the cytotoxic species, initiating and promoting the pathology through variety of mechanisms. It is therefore important to develop methods to detect these small aggregates in accessible biofluids for diagnosis of disease. However, this is challenging due to the low concentration of these aggregates in biofluids and the fact that significant post-translational modification can also occur during the aggregation process such a truncation and phosphorylation, which may hinder immunolabelling.

[0040] According to one aspect of the invention, there is provided a method of analysing one or more protein complexes in a body fluid sample, in particular a blood, serum or plasma sample, comprising:

[0041] (a) performing an assay to quantify the level of the one or more protein complexes present in the sample, wherein the assay comprises a capture binding agent and a detection binding agent which specifically bind to the protein complex, wherein the same binding agent is used as the capture binding agent and the detection binding agent; and

[0042] (b) performing a method for assessing one or more morphological features of the protein complexes present in the sample, wherein the combination of information obtained from performing steps (a) and (b) is used to analyse the protein complex.

[0043] It will be understood that references herein to “protein complex” refer to a protein assembly formed by binding of two or more monomeric proteins. This includes whether the monomeric proteins are the same (i.e. a protein aggregate) or different. As such, a protein complex can contain a protein aggregate. Methods of the invention are particularly suited to detecting protein aggregates because the same binding agent is used as the capture and detection binding agent. This means a pair of the same binding agents is used as first the capture binding agent, and then the detection binding agent. Therefore, the capture and detection binding agents bind to the same epitope in the protein complex. This ensures to minimise the signal from any monomers.

[0044] Some protein complexes contain several components so they can be captured and detected by using two different binding agents directed to different components of the same complex. Therefore, in an alternative aspect of the invention, the same binding agent is not used as the capture and detection binding agent, i.e. the capture and detection binding agents comprise different binding agents which specifically bind to different components (e.g. different epitopes or different proteins) of the protein complex.

[0045] The disadvantage of using conventional ELISA for aggregate quantification is that the assays can only measure in terms of total mass-weight, not number of entities. Thus, they cannot distinguish between a large number of small aggregates or a small number of large aggregates. To overcome the limitation of conventional ELISA, single-molecule immunoassays, including surface-based Fluorescence Intensity Distribution Analysis (sFIDA) and Single Molecule Array (SIMOA) assay, may be used. The analytes are captured and form immunocomplexes on a surface, before single-molecule microscopy is used to report on the number of targets.

[0046] In one embodiment, the assay of step (a) is a single molecule pulldown assay. A singlemolecule pull-down assay (SiMPull) is a powerful technique that allows proteins to be selectively immobilized on a microscope coverslip via immunoprecipitation and labelled with fluorophores for imaging in a fluorescence microscope. SiMPull provides an optimal platform for highly sensitive and specific single-molecule imaging of various protein complexes present in brain samples from neurodegenerative diseases (Je, G. et al. (2017); Sideris, D. I. et al. (2021); Emin, D. et al. (2022)).

[0047] SiMPull requires at least the following steps, i.e. anchoring of capture binding agents to a surface, passivation of the surface, applying the sample which contains the protein of interest, applying labelled detection binding agents and detecting the presence of the detection binding agent by single-molecule microscopy.

[0048] Therefore, in one embodiment, the SiMPull assay comprises:

[0049] (i) contacting the body fluid sample with a solid phase comprising the capture binding agent which specifically binds to the protein complex;

[0050] (ii) contacting the protein complex bound by the capture binding agent with a detection binding agent; and

[0051] (iii) detecting the presence of the detection binding agent using single-molecule microscopy.

[0052] In one embodiment, the single-molecule microscopy is performed using a total internal fluorescence (TIRF) microscope. TIRF microscopy is known in the art, such as described in Narayan, P. et al. (2013), Fish, K. N. (2009), and Moerner, W. E. et al. (2003).

[0053] In one embodiment, the solid phase (also referred to herein as a “substrate”) is selected from a bead, a particle, a filter, a fibre, a glass substrate, a microtiter plate and a microfluidic device. In a further embodiment, the solid phase is a bead or glass substrate. In one embodiment, the solid phase is passivated. A well-passivated surface minimises the non-specific binding of biomolecules to a surface. The term “passivate” refers to a process whereby the surface of a particle is rendered relatively inert biologically. In particular, a passivated surface exhibits reduced non-specific binding of biological molecules or cells to the surface. Therefore, in one embodiment, the SiMPull assay comprises applying a blocking agent to the solid phase. Blocking agents are known in the art, for example hydrophilic polymers such as polysaccharides, synthetic oligosaccharides, proteins and synthetic peptides. In one embodiment, the blocking agent is a pluronic surfactant such as PLURONIC F127 (an ethylene oxide / propylene oxide block co-polymer).

[0054] In one embodiment, the one or more morphological features are selected from: size of the protein complex and shape of the protein complex. Determining the size and shape of the protein complex (in particular, a protein aggregate) can be established by a person skilled in the art using the techniques described herein. For example, cumulative histograms of the area, perimeter and circularity distributions and their relative differences can be generated to investigate the morphological differences of the protein complexes present in a sample. The area and perimeter may be measured collectively to establish the size of the complex.

[0055] The super-resolution method described herein may be used for measuring the size and shape of biomarkers, in particular ASC specks in human biofluids. The results provided herein show that this information provides value as a diagnostic biomarker, particularly for early dementia. Therefore, according to one aspect of the invention, there is provided a method (as described herein) for measuring the size and / or shape of ASC specks in body fluid samples. Further, there is provided the use of the size and / or shape of ASC specks as a biomarker for dementia diagnosis, in particular AD, PD or FTLD diagnosis.

[0056] Super-resolution imaging of individual complexes from the blood or CSF using stochastic optical reconstruction microscopy (dSTORM) or DNA-PAINT allows one to characterise indetail the size and shape of single complexes down to 30 nm resolution (Rust, M. J. et al. (2006); Lobanova, E. et al. (2022)). This resolution, and the statistics of distribution of complex size and shape, provides additional metrics to distinguish healthy controls from people with early disease (De, S. et al. (2019)).

[0057] In one embodiment, step (b) is performed using an imaging method. In a further embodiment, the imaging method is Single-Molecule Localization Microscopy (SMLM). In one embodiment, the SMLM is selected from: stochastic optical reconstruction microscopy (STORM), photo-activated localization microscopy (PALM), or point accumulation for imaging in nanoscale topography (PAINT). In a further embodiment, the SMLM is selected from: direct stochastic optical reconstruction microscopy (dSTORM) and DNA-Point Accumulation for Imaging in Nanoscale Topology (DNA-PAINT).

[0058] Methods of single molecule localization microscopy (SMLM) involve imaging small random subsets of fluorophores in many consecutive diffraction-limited images, which are computationally detected and localized with high precision. It uses the photochemical properties of certain fluorophores to switch them between a dark and an emissive state. The combined fluorophore localizations are used to generate a dense super-resolution image (or dense image) defined for example as a 2D histogram of independent localizations. STORM, PALM and PAINT are all SMLM techniques; they differ only in how they achieve the random activation of a subset of molecules during the acquisition.

[0059] In one embodiment, step (b) is performed by measuring brightness distribution of the protein complexes present in the sample. This embodiment is based on the principle that larger protein complexes will bind more antibodies and hence be brighter and produce a larger fluorescence signal. It is therefore possible to use the brightness of the capture complexes as a proxy for size and measure the fraction of complexes brighter than a certain threshold.

[0060] Methods of the invention are highly sensitive. In one embodiment, the method has a limit of detection which is less than about 10, 5 or 1 pg / ml, such as less than about 0.5 pg / ml. In particular, the limit of detection is about 0.4 pg / ml. As used herein, the term "limit of detection" or "LOD" is the point at which the measured value is larger than the uncertainty associated with it.

[0061] In one embodiment, two or more types of protein complexes are quantified. In one embodiment, the protein complex comprises more than one type of protein. In this embodiment, a second detection binding agent may be used to detect another protein present in the protein complex. For example, a different second detector antibody with a different coloured dye label (to the capture antibody-matched detection antibody) can be used to study co-aggregate quantity and size by intensity with diffraction-limited co-localisation imaging. BINDING AGENTS

[0062] Binding agents as used herein may comprise a ligand or binder specific for the desired biomarker, e.g. target protein complexes and / or aggregates. It will be clear to those skilled in the art that the terms “antibody”, “binder” or “ligand” as used herein are intended to include any binder capable of binding to particular molecules or entities and that any suitable binder can be used in the method of the invention. In one embodiment, the binding agent, such as the antibody, specifically binds to the target biomarker. The specificity of an antibody is the ability of the antibody to recognize a particular antigen as a unique molecular entity and distinguish it from another. An antibody that “specifically binds” to an antigen or an epitope is a term well understood in the art. A molecule is said to exhibit “specific binding” if it reacts more frequently, more rapidly, with greater duration and / or with greater affinity with a particular target antigen or epitope, than it does with alternative targets. An antibody “specifically binds” to a target antigen or epitope if it binds with greater affinity, avidity, more readily, and / or with greater duration than it binds to other substances.

[0063] Ligands or binders, for example, naturally occurring or chemically synthesised compounds, are capable of specific binding to the desired target. A ligand or binder may comprise a peptide, an antibody or a fragment thereof, or a synthetic ligand such as a plastic antibody, or an aptamer or oligonucleotide, capable of specific binding to the desired target. Non-limiting exemplary binding agents include aptamers, antibodies, adnectins, ankyrins, other antibody mimetics and other protein scaffolds, autoantibodies, chimeras, small molecules, nucleic acids, lectins, ligand-binding receptors, imprinted polymers, avimers, peptidomimetics, hormone receptors, cytokine receptors, synthetic receptors, and modifications and fragments thereof. In some embodiments, a binding agent is selected from an aptamer and an antibody. The antibody can be a monoclonal antibody or a fragment thereof. It will be understood that if an antibody fragment is used then it retains the ability to bind the biomarker so that the biomarker may be detected (in accordance with the present invention).

[0064] In a preferred embodiment, the capture binding agent and detection binding agent are antibodies. In a further preferred embodiment, the capture binding agent and detection binding agent are monoclonal antibodies.

[0065] References herein to “detection binding agent” refers to an binding agent that is capable of being detected. For example, the detection binding agents used herein are labelled to enable detection. Therefore, the detection binding agent may be labelled with a detectable marker, such as a luminescent, fluorescent, enzyme or radioactive marker. In one embodiment, the detection binding agent is labelled with a fluorescent marker. In a further embodiment, the fluorescent marker is a fluorescent dye, in particular a fluorescent dye with photophysical properties.

[0066] In another embodiment, the detection binding agent comprises a nucleic acid reporter. In this embodiment, a nucleic acid sequence acts as a label (a “nucleic acid reporter”) to indicate that the detection binding agent is present in the sample. The nucleic acid reporter may then be detected by qPCR, digital PCR, or next generating sequencing (NGS). In a further embodiment, the nucleic acid reporter comprises a target ID or barcode sequence which may be used to identify the presence of the reporter.

[0067] Alternatively, or additionally, a binding agent may be labelled with an affinity tag, e.g. a biotin, avidin, streptavidin or His (e.g. hexa-His) tag.

[0068] In one embodiment, the capture binding agent is indirectly attached to the solid phase such as through biotin and streptavidin, neutravidin or avidin interactions.

[0069] In one embodiment, the binding agent is selected from the monoclonal antibody clones: 4B12, SC211 , 6E10, HT7, AT8, 8B9, 5H3, Syn202, Syn-O2, MJFR-14 and LB509. In a further embodiment, the binding agent is selected from the monoclonal antibody clones: 4B12, SC211 , 6E10, HT7, AT8, 8B9 and 5H3. In a yet further embodiment, the binding agent is selected from the monoclonal antibody clones: 4B12, SC211 , 6E10, HT7 and AT8. In one embodiment, the binding agent is selected from the antibody clones: AL177 and TMS1 , in particular AL177.

[0070] PROTEIN COMPLEXES

[0071] Methods and biomarkers of the invention detect protein complexes, in particular protein aggregates. As used herein, the term “protein aggregate” refers to a collection of proteins that grouped together. Therefore, protein aggregates comprise two or more protein monomers (e.g. wherein the monomers are the same protein, i.e. a homo-multimeric protein complex). The terms “protein complex”, “protein aggregate” or “protein assembly” are used interchangeably herein. It will be understood that protein aggregates or complexes are not necessarily all one protein but will contain the protein targeted by the binding agent. Methods of the invention use the same binding agent for capture and detection, thereby ensuring that only species larger than dimers are specifically measured, rather than inactive monomers which may be present in the sample.

[0072] Methods of the invention may be run multiple times, optionally in parallel, to detect more than one type of protein complex. In one embodiment, two or more types of protein complexes are analysed. Examples of proteins complexes, such as protein complexes associated with inflammation or protein aggregates that form during the development of neurodegenerative disease, are described herein.

[0073] As described herein, the protein complex may comprise more than one type of protein. For example, a co-aggregate that forms during the development of neurodegenerative disease.

[0074] In one embodiment, the protein complexes analysed are protein aggregates that form during the development of neurodegenerative disease (e.g. an amyloidogenic protein aggregate) and / or a protein complex associated with inflammation.

[0075] In one embodiment, the assay is performed to quantify a total level of the protein complex and a level of the protein complex which has been post-translationally modified or truncated, to produce a combined biomarker. References herein to the “level” of a protein complex may refer to the molar concentration or the number of protein complexes present in a sample.

[0076] In a further embodiment, the combined biomarker is the ratio of the total level of the protein complex to the level of protein complex which has been post-translationally modified or truncated.

[0077] PROTEIN COMPLEXES ASSOCIATED WITH INFLAMMATION

[0078] Protein complexes detected by methods of the invention may be associated with inflammation.

[0079] Inflammasomes are part of the innate immune system and activate inflammatory responses. An inflammasome is a multiprotein oligomer of caspase 1 , PYCARD / ASC, and a member of the NOD-like receptor (NLR) family. NLRP1 , NLRP3 and NLRC4 are subsets of the NLR family and thus have two common features: the first is a nucleotide-binding domain (NBD) which is bound to by ribonucleotide-phosphates (rNTP) and is important for self-oligomerization. The second is a C-terminus leucine-rich repeat (LRR), which serves as a ligand-recognition domain for other receptors (e.g. TLR) or microbial ligands. In some embodiments, the protein complex is a protein aggregate present in an inflammasome, such as an NLRP3 inflammasome.

[0080] One such protein complex is the adapter protein apoptosis associated speck-like protein containing a CARD (ASC) speck which is present in inflammasomes. Therefore, in one embodiment, the protein complex analysed is an ASC speck.

[0081] The term “ASC” refers to the human adapter protein apoptosis associated speck-like protein containing a CARD (UniProt Acc. No. Q9LILZ3) encoded by the PYCARD gene or an allelic variant or ortholog thereof. It may also be referred to as “CARD5” or “TMS1”.

[0082] The results presented herein show both early-stage PD and AD serum cohorts had an 1.5-2 fold increase in the number of ASC specks in people with disease compared to controls. The results show that the number of ASC specks in serum distinguished people with early PD (disease duration 0.5+ / -0.3 years) from controls with AUC = 83% in cohort 1 and 100% in cohort 2, as well as AD from controls with AUC = 86%.

[0083] PROTEIN COMPLEXES ASSOCIATED WITH NEURODEGENERATION

[0084] Protein complexes detected by methods of the invention may be protein aggregates that form during the development of neurodegenerative disease.

[0085] Protein complexes detected by methods of the invention may be amyloidogenic protein aggregates. Amyloidogenic proteins are typically soluble monomeric precursors, which undergo conformation changes associated with the polymerization into 8- to 10-nm wide fibrils, which culminate in the formation of amyloid aggregates. Some amyloidogenic aggregates are extracellular, such as senile plaques of Alzheimer's disease, which are composed of amyloid beta (Ap) peptides. Intracytoplasmic amyloid aggregates, such as neurofibrillary tangles in Alzheimer's disease and Lewy bodies in Parkinson's disease, are composed of the proteins tau and alpha-synuclein, respectively.

[0086] In one embodiment, the protein aggregates that form during the development of neurodegenerative disease are brain-derived. Aggregates that form during neurodegenerative diseases may not be specific to the brain and therefore aggregates of the same protein deriving from peripheral, non-brain sources may also be detected in body fluids. Methods of the invention may use binding agents that selectively bind brain-derived aggregates (e.g. brain-derived tau) and avoid isoforms which are associated with peripheral sources. In one embodiment, the protein complexes analysed are selected from: Amyloid-p (Ap) aggregates, tau aggregates, a-synuclein (a-syn) aggregates, TAR DNA-binding protein 43 (TDP-43) aggregates or co-aggregates thereof. In particular, the protein complexes analysed are protein aggregates comprising Ap, p-tau, a-syn or TDP-43.

[0087] Amyloid-beta (A-beta; Ap) refers to 39 to 43 amino acid peptides which are derived by proteolytic cleavage from the amyloid precursor protein (APP). Amyloid-beta is a major component of proteinaceous plaques found in the brain tissue of Alzheimer's disease patients. Initially it was thought that amyloid-beta becomes cytotoxic when it forms large insoluble fibrillar aggregates but more recent studies have shown that small oligomers of amyloid-beta may be the major cytotoxic species. It has been hypothesized that these small oligomers bind to the neuronal membrane, leading to cell death, possibly by membrane permeabilization.

[0088] Tau is a microtubule binding protein which can aggregate to form paired helical filaments (PHFs), which are amyloid in nature (based on cross p-sheet structure). The accumulation of hyperphosphorylated tau (p-tau) in neurons is associated with neurofibrillary degeneration. Recent work shows that oligomeric tau aggregation intermediates are the most toxic compounds formed during the process of tau fibril formation. These oligomers effectively decrease cell viability and increase phospholipid vesicle leakage. Tau oligomers have also been identified as an acutely toxic tau species in vivo, and induce neurodegeneration by affecting mitochondrial and synaptic function, both of which are early hallmarks in AD and other tauopathies.

[0089] Alpha-synuclein (a-synuclein; a-syn) is a small protein expressed at high levels in neuronal cells. The central role of a-synuclein in the pathogenesis of Parkinson's disease is well established by genetic and pathological data. Point mutations and multiplications of the SNCA gene are sufficient to cause an autosomal dominant form of Parkinson's disease, suggesting that dysfunction of the alpha-synuclein protein is a primary step in disease pathogenesis.

[0090] TAR DNA-binding protein 43 (TDP-43) is a highly conserved RNA / DNA-binding protein involved in the regulation of RNA processing. TDP-43 is an intranuclear protein encoded by the TARDBP gene that is involved in RNA splicing, trafficking, stabilization, and therefore affects the regulation of gene expression. TDP-43 species have been shown to colocalize with senile plaques and neurofibrillary tangles. Furthermore, TDP-43 has been associated with the severity of AD pathology. Phosphorylated and truncated forms of TDP-43 have been associated with amyotrophic lateral sclerosis (ALS) and a subset of frontotemporal lobar degeneration (FTLD).

[0091] Protein complexes that have been modified, in particular post-translationally modified, are of particular interest in the present invention because modified proteins are commonly associated with disease pathology. Therefore, in one embodiment, the protein complex comprises a post-translationally modified protein or a truncated protein. It will be understood that references to “protein” include references to “peptides”.

[0092] In one embodiment, the assay is performed to produce a combined biomarker. In a further embodiment, the combined biomarker is the ratio of the total level of the protein complex to the level of protein complex which has been post-translationally modified or truncated.

[0093] As used herein, the term “post-translational modification” or “PTM” refers to a reaction wherein a chemical moiety is covalently added to or non-covalently binds to a protein, after the initial synthesis (i.e. translation) of the polypeptide chain. The post translational modification may include phosphorylation, ubiquitination, acetylation, methylation, which may be mono-, di-or tri-methylation, ribosylation, citrullination, hydroxylation, glycosylation, nitrosylation, glutamination, pyroglutamination and / or isomerisation. In a preferred embodiment, the post- translational modification is phosphorylation or ubiquitination.

[0094] For example, tau can be phosphorylated at one or more amino acid residues including tyrosine at amino acid positions 18, 29, 97, 310, and 394 serine at amino acid positions 184, 185, 198, 199, 202, 208, 214, 235, 237, 238, 262, 293, 324, 356, 396, 400, 404, 409, 412, 413, and 422; and threonine at amino acids positions 175, 181 , 205, 212, 217, 231 , and 403. In one embodiment, p-tau is p-tau205, p-tau202, p-tau181 , p-tau217 or p-tau231. In one embodiment, p-tau is p-tau202 and / or p-tau205.

[0095] In one embodiment, the protein complex comprises a protein with an N-terminal or a C- terminal truncation.

[0096] In one embodiment, the protein complex comprises phosphorylated a-syn, truncated a-syn, ubiquitinated a-syn, ubiquitinated tau, phosphorylated tau, pyroglutamate-modified Amyloid- P, truncated tau or truncated Amyloid-p. COMBINED BIOMARKER

[0097] Aggregates of a-syn, Ap and tau are sensed by the immune system via pattern recognition receptors (PRRs) (Kouli, A. et al. (2018); Tahara, K. et al. (2006); Meng, J. X. et al. (2022)). These aggregates trigger a pro-inflammatory cascade through Toll-like receptors and NLRP3 inflammasome formation in immune cells such as microglia, astrocytes and blood monocytes which, if not resolved, may lead to neuronal damage and ultimately neurodegeneration (Latz, E. et al. (2013); Gordon, R. et al. (2018); Halle, A. et al. (2008)). Inflammasome complexes contain an NLR (nucleotide-binding domain, leucine-rich repeat containing) protein, the adapter protein apoptosis associated speck-like protein containing a CARD (ASC) and an effector protein (caspase 1). Activation of inflammasomes results in cleavage of pro-l L-1 and pro-IL-18 (to form active inflammatory cytokines) and gasdermin D to trigger lytic cell death. During inflammasome activation ASC rapidly assembles into a large protein complex, up to 1 pm in size, termed the ‘speck1which can, ultimately, be lost from the cell when it ruptures (Venegas, C. et al. (2017)). Inflammasome activation contributes to AD and PD pathology in disease models (Gordon, R. et al. (2018); Dempsey, C. et al. (2017)). There is increased expression of NLRP3 inflammasome in human AD and PD brain and blood (Holbrook, J. A. et al. (2021); Jewell, S. et al. (2022)).

[0098] In one embodiment, the method is used to measure two or more types of protein complexes to produce a combined biomarker. In one embodiment, the method is used to measure a total level of a protein complex and a level of the protein complex which has been post- translationally modified or truncated, to produce a combined biomarker.

[0099] The present inventors have developed assays for both total aggregates and post-translational modified aggregates which enables a ratio to be determined (i.e. as a combined biomarker). Aggregate concentration during assay runs due to the aggregates sticking to the walls of containers and sample handling which is a big problem in the field. The present inventors have found that measuring the changes in the proportion of aggregates, rather than concentration (i.e. total monomer concentration / weight, as currently measured), reduces the effect of adsorption and sample handling to provide a more accurate diagnostic test on a sample. This may be done by taking ratios of the concentration of one type of aggregate to another.

[0100] In one embodiment, the combined biomarker is determined as a ratio. In particular, the combined biomarker may be the ratio of two or more protein aggregates that form during the development of neurodegenerative disease. In one embodiment, the combined biomarker comprises the sum of the level of two or more protein aggregates that form during the development of neurodegenerative disease.

[0101] In one embodiment, the biomarker comprises:

[0102] (i) the sum of the ratio of the level of a-synuclein aggregates to the level of Amyloid-p aggregates, and the level of tau aggregates to the level of p-tau aggregates (i.e. (a-syn I Ap) + (tau I p-tau));

[0103] (ii) the ratio of the level of p-tau aggregates to the level of tau aggregates (i.e. p-tau I tau);

[0104] (iii) the ratio of the level of Amyloid-p aggregates to the sum of the level of Amyloid- P, a-synuclein and tau aggregates (i.e. Ap aggregates / (Ap + a-synuclein + tau aggregates));

[0105] (iv) the ratio of the sum of the level of Amyloid-p and p-tau aggregates to the level of a-synuclein aggregates (i.e. (Ap + p-tau aggregates) I a-syn aggregates); or

[0106] (v) the ratio of the sum of the level of Amyloid-p and p-tau aggregates to the sum of the level of a-synuclein aggregates and the ratio of the level of p-tau aggregates to the level of tau aggregates (i.e. (Ap + p-tau aggregates) I (a-syn aggregates + (p-tau aggregates / tau aggregates))).

[0107] In one embodiment, the combined biomarker is the ratio of the level of p-tau aggregates to the level of tau aggregates (i.e. p-tau I tau). The present inventors have surprisingly found that the proportion of phosphorylated tau aggregates increases in AD.

[0108] There are a number of small changes that occur in the aggregates described herein with the development of disease. The present inventors have found that combining several of these small changes together in a combined biomarker indicates a bigger change with disease and provides much higher diagnostic accuracy.

[0109] According to one aspect of the invention, there is provided a combined biomarker for use in the detection or prognosis of a neurodegenerative disorder, wherein the combined biomarker comprises the sum of the proportion of two or more protein aggregates that form during the development of neurodegenerative disease.

[0110] In one embodiment, the combined biomarker comprises the sum of the ratio of the level of a- synuclein aggregates to the level of Amyloid-p aggregates, and the level of tau aggregates to the level of p-tau aggregates (i.e. (a-syn / Ap) + (tau / p-tau)). References herein to a "combined biomarker" or "composite biomarker" refer to a biomarker comprising two or more discrete biomarkers. The combined biomarker may comprise a post- translationally modified protein or a truncated protein. Therefore, the combined biomarker may comprise the same protein in two or more discrete formats, e.g. the protein in its natural state and post-translationally modified and / or truncated, or the protein with different post- translational modifications.

[0111] The results presented herein show that the detection of ASC specks in patient samples is a useful marker of neuroinflammation that could help monitor disease progression, especially when combined with measurement of protein aggregates associated with neurodegeneration, such as a-syn, Ap and tau aggregates.

[0112] According to a further aspect of the invention, there is provided a combined biomarker for use in the detection or prognosis of a neurodegenerative disorder, wherein the combined biomarker comprises two types of protein complexes present in a sample, wherein one of the protein complexes is ASC speck.

[0113] In one embodiment, the combined biomarker comprises two biomarkers, i.e. a level or score related to ASC speck (e.g. number or morphology) and a second level or score related to another type of protein complex (e.g. Ap, tau or a-syn aggregates).

[0114] References herein to "biomarker level" or "level" refer to a measurement that is made using any analytical method for detecting the biomarker in a biological sample and that indicates the presence, absence, absolute amount or concentration, relative amount (i.e. number) or concentration, titer, a level, an expression level, a ratio of measured levels, or the like, of, for, or corresponding to the biomarker in the biological sample. The exact nature of the "level" depends on the specific design and components of the particular analytical method employed to detect the biomarker.

[0115] The components of the combined biomarker can be elevated (i.e. they can be "higher") or they can be reduced (i.e. they can be "lower") with respect to a reference level. The term "reference level" as used herein refers to the level of one or more biomarkers disclosed herein, or a score derived from the measurement of one or more biomarkers disclosed herein, obtained from a reference / control group (e.g. healthy individuals, individuals associated with a particular group, such as a prognostic group, for example recurrence of the disease or non-response of the disease to a particular therapeutic agent or combination thereof).

[0116] In one embodiment, the other type of protein complex is a protein aggregate that forms during the development of neurodegenerative disease, such as protein complexes selected from Amyloid-p aggregates, tau aggregates, a-synuclein aggregates and TDP-43 aggregates.

[0117] In one embodiment, the combined biomarker comprises the level of ASC speck and / or the level of one or more protein aggregates that form during the development of neurodegenerative disease.

[0118] In one embodiment, the combined biomarker comprises the sum of the level of two or more protein aggregates that form during the development of neurodegenerative disease.

[0119] In one embodiment, the combined biomarker is the sum of the level of ASC speck and the level of one or more protein aggregates that form during the development of neurodegenerative disease.

[0120] In one embodiment, the combined biomarker comprises a score for one or more morphological features for the ASC speck and / or a score for one or more morphological features for the protein aggregates that form during the development of neurodegenerative disease.

[0121] In a further embodiment, the one or more morphological features is selected from: the size of the ASC speck or the protein aggregate; or shape of the ASC speck or the protein aggregate. The shape and size of the aggregates can be detected as described hereinbefore.

[0122] In one embodiment, the combined biomarker is determined as a ratio. In some embodiments, the combined biomarker is determined as the ratio of two or more biomarkers, e.g. where the ratio is the level of ASC speck to the level of one or more protein aggregates that form during the development of neurodegenerative disease. This may be compared with a biomarker control ratio reference value or control ratio reference value range, and biomarker ratio values greater than the value of the control ratio reference value are indicative of the condition (e.g. a neurodegenerative disorder).

[0123] The results presented herein show that the ratio of the number of ASC specks to Ap aggregates detected in serum (ASC / A ) is 10.3 times larger in AD than the control serum and the discrimination accuracy of the assay improves from 66% (using Ap aggregates only) to 78%. Therefore, in one embodiment, the combined biomarker comprises the ratio of the level of ASC speck to the level of Amyloid-p aggregates.

[0124] The results presented herein show that by combining the measures of serum Ap aggregates, p-tau202 / 205 aggregates and ASC specks together, a composite biomarker profile defined as a ratio of the sum of the number of p-tau202 / 205 aggregates and ASC specks to the number Ap aggregates ((p-tau202 / 205 + ASC) I Ap) increases 6 fold in early disease and could distinguish AD from control serum with an accuracy of 92%. Therefore, in one embodiment, the combined biomarker comprises the ratio of the sum of the level of ASC speck and the level of p-tau aggregates to the level of Amyloid-p aggregates.

[0125] The results presented herein show that by taking the ratio of the absolute number of ASC specks to the number of p-tau202 / 205 aggregates in serum (ASC I p-tau202 / 205) can improve the discrimination accuracy between AD dementia and FTD from 82% (using a single p- tau202 / 205 aggregate biomarker) to 91%. Therefore, in one embodiment, the combined differential biomarker comprises the ratio of the level of ASC speck to the level of p-tau aggregates.

[0126] The results presented herein show that the serum (a-syn + ASC) / Ap ratio is increased 3.1 - 4.9-fold in disease and accurately differentiated people with early PD from controls (AUC = 97% in cohort 1 and 93% in cohort 2). Interestingly, the ASC / Ap ratio alone also gave an AUC of 96% and 3.4-fold increase in cohort 1 and AUC of 98% and 3-fold increase in cohort 2. Therefore, in one embodiment, the combined biomarker comprises the ratio of the sum of the level of ASC speck and the level of a-synuclein aggregates to the level of Amyloid-p aggregates. In another embodiment, the combined biomarker comprises the ratio of the level of ASC speck to the level of Amyloid-p aggregates.

[0127] The results presented herein show that the total number of ASC specks divided by the morphologically distinct fraction of ASC specks (total ASC / morphologically distinct fraction ASC ratio) increased 4.7-fold in disease and worked best at differentiating AD from controls (AUC = 100%). In PD (cohort 1), the ratio of the sum of the morphologically distinct fraction of ASC specks and total number of ASC speck to the number of Ap aggregates ((morphologically distinct fraction ASC + total ASC) / AP) was the most promising at discriminating early PD from control serum, providing an AUC of 97% and increased 1.7-fold in PD. Therefore, in one embodiment, the ratio of the sum of the level of ASC speck and the score for one or more morphological features for the ASC speck to the level of Amyloid-p aggregates.

[0128] In one embodiment, the combined biomarker comprises:

[0129] (i) the ratio of the level of ASC speck to the level of Amyloid-p aggregates;

[0130] (ii) the ratio of the level of ASC speck to the level of p-tau aggregates;

[0131] (iii) the ratio of the sum of the level of ASC speck and the level of p-tau aggregates to the level of Amyloid-p aggregates;

[0132] (iv) the ratio of the sum of the level of ASC speck and the level of a-synuclein aggregates to the level of Amyloid-p aggregates;

[0133] (v) the ratio of the sum of the level of ASC speck and the score for one or more morphological features for the ASC speck to the level of Amyloid-p aggregates.

[0134] In one embodiment, the combined biomarker comprises the ratio of the sum of the level of ASC speck and the level of a-synuclein aggregates to the level of Amyloid-p aggregates for use in the detection of Parkinson’s disease. In another embodiment, the combined biomarker comprises the ratio of the sum of the level of ASC speck and the score for one or more morphological features for the ASC speck to the level of Amyloid-p aggregates for use in the detection of Parkinson’s disease.

[0135] In one embodiment, the combined biomarker comprises the ratio of the sum of the level of ASC speck and the level of p-tau aggregates to the level of Amyloid-p aggregates for use in the detection of Alzheimer’s disease.

[0136] In one embodiment, the combined biomarker comprises the ratio of the level of ASC speck to the level of p-tau aggregates for use in the differential diagnosis between Alzheimer’s disease dementia and frontotemporal dementia.

[0137] According to a further aspect of the invention, there is provided the use of the combined biomarker defined herein as a biomarker for the manufacture of a diagnostic kit for diagnosing or detecting a neurodegenerative disorder.

[0138] According to a further aspect of the invention, there is provided the use of the combined biomarker defined herein as a biomarker for the manufacture of a diagnostic kit for determining the prognosis of a subject with a neurodegenerative disorder. According to a further aspect of the invention, there is provided the use of the combined biomarker defined herein as a biomarker for the manufacture of a diagnostic kit for monitoring the efficacy of a therapy in a subject having, suspected of having, or being predisposed to a neurodegenerative disorder.

[0139] NEURODEGENERATIVE DISORDERS

[0140] The neurodegenerative disorder is preferably characterised by protein aggregation. Protein aggregation is known to underlie a number of neurodegenerative diseases, including Parkinson's disease, Alzheimer's disease, Huntington's disease, dementia with Lewy bodies (DLB), pure autonomic failure (PAF), multiple system atrophy (MSA), progressive supranuclear palsy (PSP), corticobasal syndrome (CBS), Hallervorden-Spatz disease, dementia such as frontotemporal dementia, frontal temporal dementia with Parkinsonism linked to chromosome 17 (FTDP-17), tauopathies, Pick's disease, traumatic brain injury (TBI), corticobasal degeneration, transmissible spongiform encephalopathy, Amyotrophic lateral sclerosis and amyloidosis.

[0141] In one embodiment, the neurodegenerative disorder is selected from Parkinson’s disease (PD) and other Parkinsonian disorders (such as MSA, PSP, CBS), Alzheimer’s disease (AD), frontotemporal dementia (FTD), tauopathies (such as PSP and Pick’s disease), and traumatic brain injury.

[0142] In one embodiment, the neurodegenerative disorder is a proteinopathy (i.e. a class of diseases which are caused by structurally abnormal proteins). In one embodiment, the neurodegenerative disorder is preferably a synucleinopathy, most preferably Parkinson's disease.

[0143] Parkinson's Disease (PD) is characterized by muscle rigidity, tremor, bradykinesia progressing to akinesia, and is caused by progressive loss of dopaminergic neurons in the substantia nigra. PD is primarily a disease of an aging population and with a rapidly increasing aging population, the burden of PD will reach even broader proportions in the near future. Neurodegenerative disorders also include other Parkinsonian disorders. Atypical Parkinsonian disorders are progressive diseases that present with some of the signs and symptoms of Parkinson’s disease, but that generally do not respond well to drug treatment with levodopa. They are also associated with abnormal protein build-up within brain cells. Disorders considered a Parkinsonian disorder include dementia with Lewy bodies, PSP, MSA and CBS. In one embodiment, the Parkinson's disease is prodromal Parkinson’s disease or early Parkinson's disease.

[0144] Prodromal Parkinson’s disease refers to the stage at which individuals do not fulfil diagnostic criteria for PD (i.e. bradykinesia and at least 1 other motor sign) but do exhibit signs and symptoms that indicate a higher than average risk of developing motor symptoms and a diagnosis of PD in the future. A number of prodromal symptoms in patients with PD have been identified. The most common symptoms associated with prodromal-PD include hyposmia, constipation, mood disorders, and REM sleep behaviour disorder (RBD).

[0145] In one embodiment, the neurodegenerative disorder is preferably a tauopathy, most preferably Alzheimer's disease.

[0146] In one embodiment, the Alzheimer’s disease is mild cognitive impairment (MCI) or early AD. There are several types of neurological disorders related to Alzheimer’s disease. As cognitive dysfunction appears gradually in dementia including Alzheimer’s disease, there is a disease status of pre-stage of dementia. This stage is called as mild cognitive impairment (MCI). MCI is defined as a condition characterized by newly acquired cognitive decline to an extent that is beyond that expected for age or educational background, yet not causing significant functional impairment, and not showing disturbance in daily life.

[0147] METHODS OF DIAGNOSIS

[0148] Uses and methods of detecting, monitoring and of diagnosis and prognosis according to the invention described herein are useful to confirm the existence of a disease, to monitor development of the disease by assessing onset and progression, or to assess amelioration or regression of the disease. Uses and methods of detecting, monitoring and of diagnosis are also useful in methods for assessment of clinical screening, prognosis, choice of therapy, evaluation of therapeutic benefit, i.e. for drug screening and drug development.

[0149] According to a further aspect of the invention, there is provided a method for determining whether a subject has, or is at risk of developing, a neurodegenerative disorder, the method comprising performing a method as defined herein to analyse one or more protein complexes in a body fluid sample obtained from the subject. This technology has been shown in the Examples presented herein to be particularly well suited for the detection of protein complexes for the diagnosis of early neurodegenerative disease. Disease pathology, particularly at the early stage, involves small changes in multiple aggregates / protein assemblies and the inventors have found that detecting several small changes together can form a combined biomarker which shows clear changes with disease and hence has high diagnostic accuracy. This method is for use in body fluids, in particular serum or plasma, as will be described herein.

[0150] The terms “detecting”, “determining” or “diagnosing” as used herein encompasses identification, confirmation, and / or characterisation of a disease state. Methods of determining, monitoring and of diagnosis and prognosis according to the invention are useful to confirm the existence of a disease, to monitor development of the disease by assessing onset and progression, or to assess amelioration or regression of the disease. Methods of determining, monitoring and of diagnosis are also useful in methods for assessment of clinical screening, prognosis, choice of therapy, evaluation of therapeutic benefit, i.e. for drug screening and drug development. Methods of the invention find particular use in the detection of disease at an early stage (e.g. before symptoms develop) and therefore find use in prognosis, i.e. identifying patients at risk of developing disease.

[0151] According to a further aspect of the invention, there is provided a method of diagnosing a neurodegenerative disorder in a subject, comprising:

[0152] (a) measuring the combined biomarker described herein in a sample obtained from the subject; and

[0153] (b) using the measurement detected for the combined biomarker to determine if the subject has, or is at risk of developing, a neurodegenerative disorder.

[0154] In one aspect, the combined biomarker is measured using the method of analysing protein complexes as described herein.

[0155] Detecting and / or quantifying may be performed directly on the purified or enriched sample, or indirectly on an extract therefrom, or on a dilution thereof. Quantifying the amount of the biomarker present in a sample may include determining the concentration of the biomarker present in the sample.

[0156] In one embodiment, the sample is a body fluid (which is used interchangeably with the term “biological fluid” herein). Any body fluid sample type may be used for the invention including without limitation blood, plasma, serum, cerebrospinal fluid (CSF), nasal discharge, faeces, urine, saliva, mucous, semen and breath, e.g. as condensed breath, or an extract or purification therefrom, or dilution thereof. Biological samples also include specimens from a live subject or taken post-mortem. The samples can be prepared, for example where appropriate diluted or concentrated, and stored in the usual manner.

[0157] In one embodiment, the body fluid sample is selected from: blood, serum, plasma, cerebrospinal fluid and saliva. In a preferred embodiment, the body fluid sample is selected from: blood, serum or plasma. It will be clear to those skilled in the art that the detection of chromatin fragments in a body fluid has the advantage of being a minimally invasive method that does not require biopsy.

[0158] Detecting and / or quantifying may be compared to a cut-off level. Cut-off values can be predetermined by analysing results from multiple patients and controls, and determining a suitable value for classifying a subject as with or without the disease. For example, for diseases where the level of biomarker is higher in patients suffering from the disease, then if the level detected is higher than the cut-off, the patient is indicated to suffer from the disease. Alternatively, for diseases where the level of biomarker is lower in patients suffering from the disease, then if the level detected is lower than the cut-off, the patient is indicated to suffer from the disease. The advantages of using simple cut-off values include the ease with which clinicians are able to understand the test and the elimination of any need for software or other aids in the interpretation of the test results. Cut-off levels can be determined using methods in the art.

[0159] Detecting and / or quantifying may also be compared to a control. Therefore, in one embodiment, the method comprises comparing the measurement / information obtained for the protein complexes to one or more controls. It will be clear to those skilled in the art that the control subjects may be selected on a variety of basis which may include, for example, subjects known to be free of the disease or may be subjects with a different disease (for example, for the investigation of differential diagnosis). Comparison with a control is well known in the field of diagnostics.

[0160] The “control” may comprise a healthy subject, a non-diseased subject and / or a subject without neurodegeneration. In a further embodiment, the control is a healthy subject, optionally a healthy aged subject. In one embodiment, the level of protein complex increases compared to the control. In an alternative embodiment, the level of protein complex decreases compared to the control.

[0161] In the case of differential diagnosis, the “control” may comprise a subject with a different neurodegenerative disorder. Therefore, according to one aspect of the invention, there is provided a differential diagnosis method for determining whether a subject has, or is at risk of developing, a neurodegenerative disorder, the method comprising performing the method as defined herein to analyse one or more protein complexes in a body fluid sample obtained from the subject and comparing the results to one or more controls with a different neurodegenerative disorder. In one embodiment, the method is used to distinguish between different forms of dementia. In one embodiment, the method is used to distinguish between different tauopathies. In a further embodiment, the method provides a differential diagnosis of a subject with AD from a subject with FTD.

[0162] In one embodiment, the level or score of the biomarker is elevated compared to the control.

[0163] It will be understood that it is not necessary to measure controls levels for comparative purposes on every occasion. For example, for healthy / non-diseased controls, once the ‘normal range’ is established it can be used as a benchmark for all subsequent tests. A normal range can be established by obtaining samples from multiple control subjects without neurodegeneration and testing for the level of biomarker. Results (i.e. biomarker levels) for subjects suspected to have neurodegeneration can then be examined to see if they fall within, or outside of, the respective normal range. Use of a ‘normal range’ is standard practice for the detection of disease.

[0164] In one embodiment, the method described herein is repeated on multiple occasions. This embodiment provides the advantage of allowing the detection results to be monitored over a time period. Such an arrangement will provide the benefit of monitoring or assessing the efficacy of treatment of a disease state. Such monitoring methods of the invention can be used to monitor onset, progression, stabilisation, amelioration, relapse and / or remission.

[0165] In monitoring methods, test samples may be taken on two or more occasions. The method may further comprise comparing the level of the biomarker(s) present in the test sample with one or more control(s) and / or with one or more previous test sample(s) taken earlier from the same test subject, e.g. prior to commencement of therapy, and / or from the same test subject at an earlier stage of therapy. The method may comprise detecting a change in the nature or amount of the biomarker(s) in test samples taken on different occasions.

[0166] A change in the level of the biomarker in the test sample relative to the level in a previous test sample taken earlier from the same test subject may be indicative of a beneficial effect, e.g. stabilisation or improvement, of said therapy on the disorder or suspected disorder. Furthermore, once treatment has been completed, the method of the invention may be periodically repeated in order to monitor for the recurrence of a disease.

[0167] Methods for monitoring efficacy of a therapy can be used to monitor the therapeutic effectiveness of existing therapies and new therapies in human subjects and in non-human animals (e.g. in animal models). Current treatments are largely aimed at reducing the level of target aggregates or inflammation, therefore biomarkers and methods of the present invention can be used to monitor the efficacy of a treatment. Monitoring methods can be used to optimise the dose, increasing or lowering it depending on how the patient responds.

[0168] According to a further aspect of the invention, there is provided monitoring the efficacy of a therapy in a subject having, suspected of having, or being predisposed to a neurodegenerative disorder in a subject, comprising:

[0169] (a) measuring the combined biomarker as defined herein in a sample obtained from the subject; and

[0170] (b) using the measurement detected for the combined biomarker to determine if the subject has, or is at risk of developing, a neurodegenerative disorder.

[0171] According to a further aspect of the invention, there is provided monitoring the efficacy of a therapy in a subject having, suspected of having, or being predisposed to a neurodegenerative disorder in a subject, comprising:

[0172] (a) performing a method as defined herein to analyse one or more protein complexes in a body fluid sample obtained from the subject; and

[0173] (b) using the analysis of the one or more protein complexes to determine if the subject has, or is at risk of developing, a neurodegenerative disorder.

[0174] These monitoring methods can also be incorporated into screens for new drug substances and combinations of substances. The identification of biomarkers for a disease state permits integration of diagnostic procedures and therapeutic regimes. The biomarker of the invention provides the means to indicate therapeutic response, failure to respond, unfavourable side-effect profile, degree of medication compliance and achievement of adequate serum drug levels. The biomarkers may be used to provide warning of adverse drug response. Biomarkers are useful in development of personalized therapies, as assessment of response can be used to fine-tune dosage, minimise the number of prescribed medications, reduce the delay in attaining effective therapy and avoid adverse drug reactions. Thus by monitoring a biomarker of the invention, subject care can be tailored precisely to match the needs determined by the disorder and the pharmacogenomic profile of the subject, the biomarker can thus be used to titrate the optimal dose, predict a positive therapeutic response and identify those subjects at high risk of severe side effects.

[0175] Methods and biomarkers described herein may be used to identify if a patient is in need of further tests to confirm diagnosis of a neurodegenerative disorder. Therefore, according to a further aspect of the invention there is provided a method of identifying a patient in need of further diagnostic tests comprising obtaining a body fluid sample from said patient, detecting the combined biomarker described herein in the body fluid sample, and using the results obtained to identify whether the patient is in need of further diagnostic tests. Such tests may include brain imaging (such as magnetic resonance imaging, computed tomography (CT), a positron emission tomography (PET) scan, a dopamine transporter scan (DaTscan), a single photon emission computed tomography (SPECT) scan), referral to a neurologist, a neurological scan and / or a biopsy.

[0176] References to “subject”, “individual” or “patient” are used interchangeably herein. The subject may be a human or an animal subject. Preferably, the subject is preferably a human. The subject may be at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, at least 60, at least 65, or at least 70 years of age. Preferably, the subject is at least 60, or at least 65 years of age. Most preferably, the subject is at least 60 years of age. Alternatively, the subject may be up to 30, up to 35, up to 40, up to 45, up to 50, up to 55, up to 60, up to 65, or up to 70 years of age.

[0177] In one embodiment, the subject has not been diagnosed with a neurodegenerative disorder. In addition, the subject may display no clinical symptoms associated with a neurodegenerative disorder (e.g. PD or AD). Methods of the invention may be used to confirm a diagnosis of a neurodegenerative disorder. Therefore, in one embodiment, the subject is suspected of suffering from a neurodegenerative disorder. In addition, the subject may display one or more clinical symptoms associated with a neurodegenerative disorder (e.g. PD or AD).

[0178] Clinical symptoms of Parkinson's disease include motor and non-motor symptoms and are well known to the skilled practitioner. Primary motor symptoms include bradykinesia (slowness of movement), rigidity, resting tremor (for example in the hand, foot, jaw or face), or postural instability. Secondary motor symptoms include stooped posture, dystonia, fatigue, impaired fine motor dexterity and motor coordination, impaired gross motor coordination, poverty of movement, akathisia, speech problems, loss of facial expression, micrographia, difficulty swallowing, sexual dysfunction and drooling. Non-motor symptoms of Parkinson's disease include pain, sleep disturbances, hyposmia, constipation, skin problems, depression, fear or anxiety, memory difficulties and slowed thinking, urinary problems, fatigue and aching, loss of energy, compulsive behaviour and cramping.

[0179] Clinical symptoms of Alzheimer's disease are also well-known in the art and include forgetfulness, confusion, depression, social withdrawal, mood swings, distrust in others, irritability and aggressiveness, changes in sleeping habits, wandering, loss of inhibitions, and delusions.

[0180] In addition, or alternatively, the subject may have a family history of a neurodegenerative disorder, such as a first degree relative with a neurodegenerative disorder. Where the neurodegenerative disorder is Parkinson's disease, the subject may have been determined to have a mutations or polymorphism in the SNCA gene, which encodes alpha-synuclein.

[0181] The uses and methods described herein may be performed in vitro, in vivo or ex vivo. The methods described herein are preferably performed in vitro. References to acts carried out on a body fluid sample “obtained” from a subject are intended to encompass acts carried only a body fluid sample already obtained of “obtainable” from a subject and vice versa.

[0182] METHODS OF TREATMENT

[0183] According to a further aspect, there is provided a method of treating a neurodegenerative disorder in a subject, which comprises the following steps: (i) detecting or measuring the combined biomarker described herein in a body fluid sample obtained from the subject;

[0184] (ii) using the measurement in step (i) as indicative of the presence of said neurodegenerative disorder in the subject; and

[0185] (iii) administering a treatment if the subject is determined to have said neurodegenerative disorder in step (ii).

[0186] According to a further aspect, there is provided a method of treating a neurodegenerative disorder in a subject in need thereof, which comprises the step of administering a treatment to a subject identified as having a differing level or score of the combined biomarker described herein in a body fluid sample obtained from said subject, when compared to the level or score of the combined biomarker in a body fluid sample obtained from a control subject.

[0187] Treatments may be selected by those skilled in the art depending on the neurodegenerative disorder. In one embodiment, the treatment is selected from medication, supportive therapy (e.g. physiotherapy or speech and language therapy) or surgery.

[0188] Suitable PD treatments will be readily recognised by the skilled person, and include those in development or that are yet to be developed but which will treat or ameliorate symptoms of PD. Suitable treatments include those that improve outcome and / or symptoms for patients. For example, the treatment being administered may be levodopa (for example, co-beneldopa or co-careldopa), dopamine agonists (for example, pramipexole, ropinirole or rotigotine), monoamine oxidase-B (MAO-B) inhibitors (for example, rasagiline, selegiline or safinamide), catechol-o-methyltransferase (COMT) inhibitors (for example, entacapone or opicapone), amantadine or anticholinergics (for example, procyclidine or trihexyphenidyl). Specific forms of these treatments may be required if symptoms are severe, such as non-oral therapies (for example apomorphine or co-careldopa) or skin patches (for example a rotigotine skin patch).

[0189] Suitable AD treatments will be readily recognised by the skilled person, and include those in development or that are yet to be developed but which will treat or ameliorate symptoms of AD. Suitable treatments include those that improve outcome and / or symptoms for patients. For example, the treatment being administered may be a cholinesterase inhibitor (e.g. galantamine, rivastigmine or donepezil) which improve cognitive function, an N-methyl-D- aspartate (NMDA) antagonist (e.g. memantine) which ameliorate symptoms of cognitive impairment, an orexin antagonist (e.g. suvorexant) used to treat insomnia, some antipsychotics (e.g. brexpiprazole), or may be an immunotherapy which targets amyloid p to reduce amyloid plaques (e.g. lecanemab or aducanumab).

[0190] The methods may comprise:

[0191] (i) measuring the level or score of the combined biomarker described herein in a body fluid (e.g., blood, serum or plasma) sample obtained from the subject, optionally using the method of analysing protein complexes as described herein;

[0192] (ii) identifying the subject as suffering from a neurodegenerative disorder based on a higher level or score of the combined biomarker compared to a control; and

[0193] (iii) administering a treatment to the subject.

[0194] It will be understood that all embodiments described herein may be applied to all aspects of the invention. Other features and advantages of the present invention will be apparent from the description provided herein. It should be understood, however, that the description and the specific examples while indicating preferred embodiments of the invention are given by way of illustration only, since various changes and modifications will become apparent to those skilled in the art.

[0195] The present invention will now be illustrated by the following examples.

[0196] EXAMPLES

[0197] Materials and Methods:

[0198] Participants

[0199] Patients with idiopathic Parkinson’s disease (within 1-2 years of diagnosis according to UK PD Brain Bank Criteria) and controls without neurological disease were enrolled from the Parkinson’s Disease Research Clinic at the John Van Geest Centre for Brain Repair, University of Cambridge / Cambridge University Hospitals NHS Trust, UK. Age and sex- matched participants without neurological disease were also recruited from the NIHR Cambridge Bioresource (http: / / www.cambridgebioresource.org.uk). Participants with Parkinson’s disease were assessed using the Movement Disorder Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS), and the Addenbrooke’s Cognitive Examination (ACE- Ill or ACE-R). Parkinson’s disease stage was determined using the Hoehn and Yahr scale (Table 1). Patients with Alzheimer’s disease (AD, including the prodromal state of mild cognitive impairment) and frontotemporal dementia (FTD) as well as participants without neurological disease were recruited in the GOLDeN Study (Genetics of Leucopathology, Dementia and Neurodegeneration) at the Cambridge University Hospitals NHS Trust), with amnestic mild cognitive impairment or Alzheimer’s disease (Table 2). Participants with mild cognitive impairment were followed clinically to confirm progression and / or have biomarker evidence of underlying Alzheimer’s disease pathology. The participants will also underwent cognitive testing such as the ACE-R and Mini-mental state examination (MMSE) to indicate the progression of disease to AD dementia stage. Ethical approval was obtained from the East of England - Essex Research Ethics Committee (16 / EE / 0445), and East of England - Cambridge Central Research Ethics Committee (03 / 303, and 15 / EE / 0270). Informed consents was provided by all participants. Table 1 . Demographic and clinical characteristics participants included in the PD versus control serum and CSF sample comparison.

[0200] Table 2. Demographic and clinical characteristics of participants included in the AD versus control and

[0201] AD dementia versus FTP serum sample comparison. Values represent the mean ± SD. Variables were compared using the permutation (exact) test except the sample size for which the binomial test was used (*p < 0.05). *Early AD cohort includes the patients who sought medical advice at a memory clinic for the first time and were diagnosed for AD based on the positive AD CSF biomarker profile. AD CSF biomarker profile was defined by a CSF Ap42 / p-tau181 ratio < 10.25 (as measured by Lumipulse G600II, Fujirebio).

[0202] Serum and CSF sampling

[0203] Blood from participants was collected by venepuncture using 7.5 ml S-Monovette tubes, samples were left to clot at room temperature for 15 minutes before centrifuging at 2000 rpm for 15 minutes at room temperature. The supernatant (serum) was collected and stored at -80°C until use. In order to minimise the impact of repeat frozen-thawed cycles, upon collection of samples, single-use 11 pL aliquots were prepared into protein low bind tubes (Protein LoBind® Tubes 0.5mL, Eppendorf). Serum samples from two independent cohorts of 10 PD patients and 10 age-matched controls (PD cohort 1 , Table 1), and 8 PD and 5 controls (PD cohort 2, Table 1) were assayed. Serum from a cohort of 20 early AD, 30 AD dementia and 10 FTD patients as well as 30 non-demented controls was also used in the study.

[0204] Lumbar puncture was performed using standard sterile procedures and with 1 % lidocaine local anaesthetic in 6 PD and 6 control subjects (Table 1). CSF (5 ml) was collected and centrifuged for 10 min at 300g at 4°C. The supernatant was collected and stored at -80°C.

[0205] Extracting soluble aggregates from brain tissue

[0206] Post-mortem brain tissue from 9 PD and 3 AD donors as well as 5 controls with no known history of neurological or neuropsychiatric symptoms was acquired from the Cambridge Brain Bank (under the Neuropathology Research in Dementia protocol, approved by London — Bloomsbury Research Ethics Committee; 16 / LO / 0508, Table 3). Brain tissue was flash-frozen and stored at -80 °C. Tissue was available from the amygdala for PD and control brains, and from the hippocampus, frontal cortex and visual association cortex for AD brains. For extraction of soluble aggregates, a previous published protocol was adapted (Hong, W. et al. (2018)). 300 mg tissue samples were cut into smaller pieces and placed into an eppendorf containing 1.5 mL of artificial cerebrospinal fluid buffer (aCSF, 124mM NaCI, 2.8mM KCI, 1.25mM NaH2PC>4, 26mM NaHCCh; pH 7.4, supplemented with 5mM EDTA, 1 mM EGTA, 5 pg / mL leupeptin, 5 pg / mL aprotinin, 2 pg / mL pepstatin, 20 pg / mL Pefabloc, 5mM NaF) for 30 min at 4°C. Afterwards the samples were centrifuged at 2000 x g for 10min and the upper 90% of the supernatant was transferred into a fresh tube. This solution was then centrifuged at 14,000 x g for 2 h and again the upper 90% supernatant was collected and dialysed for 72 h using Slide-A-Lyzer cassettes (MKCO 2 kDa, Thermo Scientific, Cat. 66330) with three buffer exchanges against aCSF buffer at 4 °C. The samples were aliquoted and stored at -80°C and each aliquot was used for just one experiment in order to avoid unnecessary freezing / thawing cycles.

[0207] Table 3. Demographic and clinical characteristics of the PD and AD brain cohort.

[0208] Preparation of inflammasome-activated THP-1 cell samples

[0209] The immortalized human monocyte cell line THP-1 from American Type Culture Collection was maintained in RPMI 1640 (Invitrogen) containing 10% HI-FBS, 2mM L-Glutamine, and 1 % penicillin and streptomycin. Monocytes were differentiated into macrophages with 10nM Phorbol 12-myristate 13-actetate (PMA) at a density of 21052.6 / cm2in a 6 well plate for 24 hours. Macrophages were then recovered in fresh complete media for another 24 hours. Ultra pure lipopolysaccharide from E. coli 0111 :B4 at 200 ng / ml was added to all macrophages for 3 hours to prime the cells. Nigericin was added at 10pM and incubated for increasing time points before conditioned media was collected and cells lysed in 50pl per well NP-40 buffer (150 mM NaCI, 50 mM Tris-HCI pH 7.4, 1 mM EDTA, 1 % NP-40). All samples were snap frozen in dry ice before SiMPull analysis. Conditioned media was also assayed for IL-1 by ELISA to confirm inflammasome activation. Coverslip preparation

[0210] Coverslips were coated according to our recently-published protocol (Zhang, Y. P. et al. (2023)). Each glass coverslip (26x76 mm, thickness #1.5, VWR, Cat. No. MENZBC026076AC40) was plasma cleaned for 15 mins. 50-well PDMS chamber gasket (cut from a CultureWell chambered coverglass, Sigma, cat. no. GBL103350-20EA) was attached to the cleaned coverslip. To prepare a non-sticky glass surface, we exploited a hydrophobic self-assembling monolayer approach using a Rain-X Rain Repellent solution. Rain-X and isopropanol were mixed in 1 :1 proportion, filtered and added to each well (10 pl). The coated coverslip was left at room temperature until full evaporation of liquid. The wells were then washed twice with 10 pl of PBS by pipetting the liquid in and out to remove RAIN-X residuals. Coverslips were prepared freshly and used in the same day as a platform for the SiMPull assay.

[0211] ASC speck Sim Pul I assay

[0212] A Rain-X hydrophobic glass coverslip was prepared according to the previous section. Then, 10 pl of 0.1 mg / ml Neutravidin (Thermo Scientific, 31000) was added to each well (last 4 lines, 40 wells and first line were unused due to its closeness to the edge) for 10 mins and then washed twice with PBS twice by pipetting the liquid in and out. In order to minimise unspecific binding, the wells were blocked with 1 % filtered solution of pluronic f127 diluted in PBS for 60 mins followed by washing with PBS twice and blocking again with 3 mg / ml of BSA (Molecular Biology Grade, New England Biolabs) in PBS-T (0.05% Tween-20 in PBS) for 30mins. 10 pl of 10nM anti-ASC antibody in PBS-T was added to each well and incubated for 15 min and then washed off the unbound Ab with PBS-T twice. Then, 10 pl of the undiluted sample (or 3- fold diluted in PBS for serum and THP-1 cell lysates only) was added to each well for 90 mins following washing with PBS-T twice. For detection of captured ASC specks, 10 pl of 5 nM detection anti-ASC antibody in PBS-T was added to the well for 15 mins and washed with PBS-T twice. Finally, each well was covered with 10ul of PBS for diffraction-limited imaging. For reliability of results, 2 quality control wells were included in each experiment: (1) buffer control, when PBS was added instead of sample and (2) no capture control, when there was no capture antibody but the correct detection antibody, to check the amount of non-specific absorption of sample on the imaging surface. The SiMPull assay for detection of a-syn and Ap aggregates in the sample was assessed in parallel using the same protocol as for ASC specks, using the same protein-specific antibody for capture and detection (see Antibody Section for details). For dSTORM imaging, another 3 layers of PDMS chambers were stacked onto the coverslip to increase the well capacity. 16.5 pl dSTORM buffer (50 mM PBS-Tris, 0.5 mM glucose, 1.3 pM glucose oxidase, 2.2 pM catalase, and 50 mM mercaptoethylamine (MEA)) was then added to each well. MEA was added to the buffer immediately before imaging. To maintain the pH during imaging, the top of the chambered coverslip was sealed using a second cleaned coverslip. To further reduce oxygen penetration, the edges of the integrated coverslip complex were further coated with nail polish and parafilm.

[0213] Antibodies and aptamer

[0214] Unconjugated rabbit anti-ASC (clone AL177, 1 mg / ml) antibody against aa at the N-terminal human ASC was purchased from AdipoGen. The Alexa 647 conjugated anti-ASC antibody was generated using Zip Alexa Fluor™ Rapid Antibody Labeling Kit (Cat. No. Z11235) and the biotinylated anti-ASC and 211 antibodies were generated using a Pierce™ FITC Antibody Labeling Kit (Cat. No. 53027). Alexa 647 anti-Ap antibody (6E10) targeting A amino acids 1- 16 was purchased from Biolegend (Cat. No. 803020 and No. 803013). The biotinylated 6E10 antibody was purchased from Biolegend (Cat. No. 803007). Unconjugated 211 antibody was purchased from Santa Cruz biotech (Cat.# SC-12767). Alexa 647 Mouse lgG1 (Cat.# MA5- 18168; PRID: AB_2539542, Invitrogen), Alexa 647 Rabbit IgG (Cat.# 3452S; RRID: AB_10695811 , Cell Signalling Technology), Bovine Serum Albumin (B9000S, New England biolabs) were also purchased commercially. Biotinylated T-SO508 aptamer (GCCTGTGGTGTTGGGGCGGGTGCG (SEQ ID NO: 1)) was purchased from ATDBio (Southampton, UK) and purified by high-performance liquid chromatography (HPLC). The aptamer recognises p-sheet structure and is specific to both a-syn and Ap oligomers (Tsukakoshi, K. et al. (2012)). Alexa 647 conjugated anti-a-syn antibody (211) was purchased from Santa Cruz biotech (Cat. SC-12767 AF647) and recognises amino acid 121-125 of human a-synuclein.

[0215] Aggregate stability assay for ASC specks

[0216] Prior adding to SiMPull assay, samples underwent a denaturation treatment with different concentrations of guanidine hydrochloride (Gdn HCI) which affects the secondary and tertiary structure of protein aggregates without affecting the primary structure of monomers. First we optimised our protocol to determine the molar concentration of Gdn HCI at which the quantity of ASC speck aggregates decreased in value to 1 / e ~ 0.4 relative to the untreated sample. For this, 10 pl aliquots of sample (undiluted PD brain homogenate or PD serum) were individually treated with decreasing concentrations of 8M Gdn HCI diluted in TRIS-buffered saline (TBS), pH 8 (Thermo Scientific) so that the final concentration in the sample varied from 0.8 M to 4 M with increment 0.5 M. After 60 min incubation at room temperature, individual samples were diluted back to the diminished concentrations of Gdn HCI in TRIS-buffered saline (TBS), pH 7.4 at the final concentration of 0.4 M in all samples and immediately added the SiMPull coverslip for ASC speck detection. For both serum and homogenate sample there was an exponential step decrease in the number of ASC specks compared to untreated sample with 0.8 M Gdn HCI treatment. To measure the range 0.72M - 0.05 M Gdn HCI with a finer increment, 10 pl aliquots of sample (3-fold diluted PD brain homogenate or undiluted PD serum) were individually treated with 1 pl of decreasing concentrations of 8M Gdn HCI diluted in TRIS-buffered saline (TBS), pH 8 (Thermo Scientific) so that the final concentration in the sample varied from 0.72M to 0.05 M. After 60 min incubation at room temperature, each condition sample was immediately added to the SiMPull coverslip for ASC speck analysis.

[0217] Imaging

[0218] Imaging was performed using a custom-build total internal reflection fluorescence (TIRF) microscope. A Nikon Ti2 Eclipse inverted microscope is integrated with 100x 1.49 NA oilimmersion objective (UPLSAPO, 100x, TIRF, Olympus) and a perfect focus system. An excitation laser beam (Oxxius, 635nm) was circularly polarised by a quarter-wave plate (WPQ05M-405, Thorlabs) and focused onto the back focal plane of the objective. The fluorescence emission was collected using the same objective and separated by a dichroic beamsplitter (Di01-R405 / 488 / 561 / 635, Semrock), with filtering performed by a long-pass emitter (BLP01-635R-25, Laser 2000). Emission is imaged onto an air-cooled EMCCD camera (Photometries Evolve, EVO-512-M-FW-16-AC-110) with frame transfer mode (electronmultiplying Gain of 11.5 e-1 / ADU and 250 ADU / photon). The open-source software MicroManager 1.4 was employed to automate image acquisition. 638 nm laser (Cobolt MLD 638, Cobalt) was used to excite Alexa 647 dyes. For diffraction-limited imaging, 1 ,5mW of laser power was applied and images were acquired with an exposure time of 50 ms and frame number of 50. For STORM imaging, 150 mW of laser power was applied and images were acquired with an exposure time of 15 ms and frame number of 8000. Continuous illumination by 405 nm laser (LBX-405-50-CIR-PP, Oxxius) at 10 mW was applied. The pixel size of the camera was measured 103.5nm.

[0219] Data analysis and statistical testing

[0220] Data were analysed using MATLAB (R2020b) unless otherwise stated. The diffraction-limited data were analysed using in-house software called Path-Connected Aggregate Recognition (PCAR) which is freely available from https: / / github.com / LobanovaEG-LobanovSV / PCAR. dSTORM data were analysed using established Imaged plug-ins. The drift correction, image reconstruction, and morphology analysis was performed by mean shift algorithm (Fazekas, F. J. et al. (2021)), ThunderSTORM (Ovesny, M. et al. (2014)), and morphology library (Legland, D. (2016)), respectively. A custom-written Matlab code was used to integrate mentioned plug- ins and automate data analysing. The code is available from https: / / github.com / YPZ858 / Super-res-code / issues. For normally distributed data, two-tailed t- test was employed. Otherwise, the permutation (exact) test was used except for the sample size, for which the binomial test was employed. Statistical significance was indicated when p < 0.05. The cumulative histograms of the area, perimeter and circularity distributions and their relative differences are generated to investigate the morphological differences in the ASC specks between diagnostic groups. To relate area and perimeter to the size of ASC speck, the size was estimated as a circle with a diameter determined as two square roots of the area divided by TT (d=2 (A rr)) or the perimeter divided by TT (d=F rr).

[0221] Diagnostic performance of each candidate biomarker was assessed using Receiver operating characteristic (ROC) curve analysis and its accuracy was evaluated by area under the curve (AUC). To identify the most promising biomarker combination, we ran a loop in MATLAB which calculates various simple combinations of measured biomarkers and their ALICs. We then ranked the output combinations of biomarkers based on their AUC values from most (highest AUC) to least (lowest AUC) promising.

[0222] Example 1 : Establishment of the ASC-SimPull assay for ASC specks detection in human biofluids

[0223] The workflow of the SiMPull method for detecting ASC specks in human biofluids is demonstrated in the Fig. 1. Briefly, by immobilising single ASC specks present in a sample with a biotinylated antibody against the protein on a glass surface, we selectively capture ASC specks. Fluorescently labelled antibody is added for single-molecule detection. Blocking the surface with polymer (polymer passivation) significantly reduces the fluorescence signal caused by non-specific antibody binding. If no sample is presented, the blocked surface will not bind any detection antibodies. By using the same ASC antibody for capture and detection, we can only detect species larger than dimers which allows us to specifically measure the activated inflammasome ASC complexes or specks rather than inactive monomers present in the sample. By combining the SiMPull assay with dSTORM we can precisely measure the size, shape and number of ASC aggregates in biofluids and use this morphology information as additional metric to determine whether there are differences between healthy controls and people with PD and AD.

[0224] We first proved that our method allows sensitive detection of the inflammatory response through measurement of ASC speck aggregates generated in THP-1 cells during inflammasome activation (Fig. 1B-G). The SiMPull assay is capable of detecting ASC specks in both the NLRP3 inflammasome-activated THP-1 cell lysates (LPS-primed nigericin- activated) and those secreted into the media compared to untreated negative controls (Fig. 1 B-E). There was a gradual increase in the number of ASC specks detected in the THP-1 media when treated with LPS / nigericin over time with the maximum number of specks detected at 30 min nigericin (Fig. 1 D & E). Significantly more specks were seen at 15 mins nigericin incubation vs det IgG CTRL media and capture / stimulation media (p<0.01 , p<0.05 respectively). There was also an increasing number of ASC specks found within the lysates when treated with nigericin over time. Lysate specks were present at 7.5 min and markedly increased by 30 min (p < 0.05), but there were also two significant reductions in the number of lysate ASC specks at 15 min (-29% , p < 0.05) and 60 min (-31 %) nigericin (Fig. 1 D). The -29% decrease in the 15 min nigericin lysate specks matched to the 3-fold increase in the corresponding media (Fig. 1 D & E). The assay was also validated with two specificity controls. The first one was when we applied the correct capture antibody for specific pulldown of ASC specks to the surface and detected the level of non-specific single-molecule fluorescence using a non-target Alexa647 IgG detection control antibody. As our second specificity control we did not use a capture antibody and detected the non-specific binding with the correct fluorescent ASC antibody. We confirmed that both our specificity controls gave us significantly lower non-specific fluorescence compared to the sample (15 min nigericin-treated THP-1 lysates or media) with the correct capture and detection antibody indicating that the measured signals are ASC specks and not non-specific (Fig. 1 D & E). After successfully validating the assay, we applied this method for super-resolution imaging of ASC speck aggregates in the same THP-1 samples. We detected ASC aggregates ranging in size between 0.035 pm (area = 0.001 pm2, see the data analysis and statistical testing section for details) to 1 pm (area = 0.745 pm2) in both cell lysates and secreted media although the media has a trend for higher proportion of smaller and rounder ASC specks than the cell lysates ( Fig. 1 F, G). At 7.5 min nigericin ASC formed specks of up to 1 pm in size (Fig. 1 B) in the cell lysates which did not increase over time (Fig. 1F). There was a significant difference in size distributions of the ASC specks between the 7.5 min and 15 min nigericin in both THP-1 cell lysates and media. The shape of ASC specks was also significantly different between the 7.5 min and 15 min nigericin lysates but not for those media. We started to observe significantly more circular ASC specks released into the media at 30 min compared to 15 min nigericin. If one considers only ASC aggregates bigger than 0.2 pm (area > 0.03 pm2), cell lysates and those media treated with 7.5 min nigericin had about 30% and 29% fractions of these aggregates compared to only 15% and 14% for the 15 min nigericin lysates and media (Fig. 1 F). Both the 15 min nigericin cell lysates and media contained the smallest and most circular ASC specks. Overall, these experiments showed that we could detect ASC speck formation in LPS-primed nigericin-activated THP-1 cells and that the ASC specks formed early in the inflammatory response (7.5 min nigericin) are as big as those treated with inflammasome activators for 8 times longer (60 min nigericin). We also observed a critical time point of 15 min with nigericin at which ASC formed the smallest and most circular specks.

[0225] Example 2: ASC specks in human biofluids as a biomarker of inflammation and inflammasome formation

[0226] Using samples from people with PD, AD, AD dementia and FTD and people without neurological disease (HC), we then measured the number of ASC specks in CSF (n = 6 PD vs 6 HCs) and serum (PD cohort 1 : n = 10 PD vs 10 HC, PD cohort 2: n = 8 PD vs 5 HC; AD plus FTD cohort: n = 30 HC vs 20 early AD, 20 AD dementia vs FTD). We also assessed soluble aggregates in samples derived from soaking brain tissue. Amygdala samples from PD cases (n = 9) with Lewy Braak stages from 3 to 5 were compared to amygdala samples from controls without known neurological disease during life (n = 5, Lewy body Braak stage 0, tau Braak stage 0 or 1). Soaked brain samples from AD cases (n = 3, tau Braak stage 3) across three different brain regions (hippocampus, frontal cortex and visual association cortex) were also assessed.

[0227] In both our early-stage PD and AD serum cohorts, we found an 1.5-3 fold increase in the number of ASC specks in people with disease compared to controls (Fig. 2A, B, D). To establish the utility of ASC specks as a potential fluid biomarker of inflammation, inflammasome formation and cell apoptosis, we performed a ROC curve analysis. We showed that the number of ASC specks in serum distinguished people with early PD (disease duration 0.5+ / -0.3 years) from controls with AUC = 83% in cohort 1 and 100% in cohort 2 (Fig. 2C) as well as AD from controls with AUC = 86% (Fig. 2I). Although the total number of detected ASC specks in CSF (Fig. 2E) was only half of those detected in serum, the use of CSF allowed us to fully discriminate PD from age-matched controls (no overlap between PD and control data) (Fig. 2F).

[0228] The ASC specks levels were also detectable in human brain tissue. We found that the number of ASC specks in the soaked brain fluid samples were similar between controls and PD. Comparing the amount of ASC speck in different brain regions in AD brains with Braak stage 3 tau pathology, we observed about a 12-fold and 8-fold increase in the number of extracellular ASC specks detected in the frontal cortex and hippocampus (regions of severe amyloid-p and tau pathology) compared to the visual association cortex (mild pathology) from the same subjects (Fig. 2H).

[0229] We also performed a series of control experiments to confirm that we were detecting ASC speck aggregates. As demonstrated in the Fig 2J, the non-specific signal due to the sample sticking to the surface without capture antibody but with correct detection antibody is negligible in comparison to the sample signal in the presence of the correct capture and detection antibody as verified using PD serum and brain samples. We also confirmed that our ASC- SimPull assay has low non-specific fluorescence compared to a positive control sample (PD serum and CSF) indicating that the majority of detected aggregates are ASC specks and not something else (Fig. 2J). As a cross-validation biochemical assay to prove our ASC SiMPull method is aggregate specific, we did a conformation stability assay on serum and brain homogenate samples. Prior to the SiMPull assay, samples underwent a denaturation treatment with different concentrations of guanidine hydrochloride (Gdn HCI) which affects the secondary and tertiary structure of protein aggregates without affecting the primary structure of monomers. All samples were then diluted back to the same low concentration of Gdn HCI (see the Material and Methods section for details) and analysed using SiMPull for ASC speck detection. Since only aggregates in the sample can be denatured by treatment we observed an exponential decay in the number of detected ASC aggregates with increasing concentrations of Gdn HCI (Fig. 21). This confirmed that we are specifically detecting aggregates.

[0230] Example 3: ASC speck composite serum biomarkers for AD and PD

[0231] Given that AD and PD are complex multifactoral disorders with different protein aggregates contributing to disease, ASC specks may have a better diagnostic value when combined with disease-specific protein aggregate markers of AD / PD. To test the potential of our ASC speck blood assay for improving discrimination between AD samples and controls, we combined measurement of ASC specks with measurements of total Ap aggregates, and phosphorylated tau aggregates in the same cohort of serum samples from 20 early AD patients and 30 controls using our established SiMPull assay with identical 6E10, AT8 (p-Ser202, p-Thr205) antibodies for capture and detection (Fig. 3). We demonstrated that the ratio of the number of ASC specks (ASC) to A aggregates (Ap) detected in serum ASC / Ap is 10.3 and 12.5 times larger in early AD and late AD (AD dementia) than the control serum and the discrimination accuracy of the assay improves from 66% (AUC = 66% using Ap aggregates only) to 78% in early AD (AUC = 78%, Fig. 3B) and to 89% in AD dementia (AUC = 89%, Fig. 3B). Furthermore, we showed that by taking the ratio of the absolute number of ASC specks to the number of p- tau202 / 205 aggregates in serum (ASC I p-tau202 / 205) we can improve the discrimination accuracy between AD dementia and FTD serum from 82% (AUG = 82% using a single p- tau205 aggregate biomarker) to 91% (AUG = 91%, Fig. 3F). By combining the measures of serum Ap aggregates, p-tau202 / 205 aggregates and ASC specks together we built a composite biomarker profile defined as a ratio of the sum of the number of p-tau202 / 205 aggregates and ASC specks to the number Ap aggregates (p-tau202 / 205 + ASC) I Ap (Fig. 3D), which is increased 6 fold in early AD and 9.7 fold at late disease stage (AD dementia) and could distinguish early AD from control serum with an accuracy of 92% (AUG = 92%) and AD dementia from control with an accuracy of 92% (AUG = 95%) (Fig. 3E). Also, the use of (p-tau202 / 205 + ASC) I Ap ratio in serum allows to distinguish AD dementia from FTD patients with an accuracy of 90% (AUC = 90%) (Fig. 3E).

[0232] We also explored different combinations of the number of ASC specks, a-syn aggregates (a- syn) and Ap aggregates (see the data analysis and statistical testing section for details) in serum from the two early PD cohorts (Fig. 3) which were assayed independently (see Table 1 and the participants section for details) see the participants section for details). In both cohorts we found that the serum (a-syn + ASC) / Ap ratio (Fig. 3J, K) is increased 3.1 - 4.9-fold in disease (cohort 1 and 2) and accurately differentiated people with early PD from controls (AUC = 97% in cohort 1 and 93% in cohort 2, Fig. 3L). Interestingly, the ASC / Ap ratio (Fig. 3G, H) alone also gave an AUC of 96% and 3.4-fold increase in cohort 1 and AUC of 98% and 3-fold increase in cohort 2 (Fig. 3I).

[0233] Overall, these results show that by combining ASC specks with measurements of other aggregates in the serum to form a composite biomarker achieves excellent discrimination and a larger dynamic range between patients and controls. For PD (a-syn + ASC) / Ap seems most promising (in cohort 1 : AUC= 97% and 3.1 -fold increase, in cohort 2: AUC = 93% and 4.9-fold increase) and for early AD the serum (p-tau202 / 205 + ASC) / Ap ratio gives excellent discrimination (AUC = 92%) and a dynamic range of 6. As for differential biomarkers, (p- tau202 / 205 + ASC) I Ap and ASC I p-tau202 / 205 show a good promise to distinguish FTD from AD dementia giving AUC of 90% and 91%, respectively.

[0234] Example 4: Size and shape of ASC specks in biofluids of people with PD / AD as a biomarker

[0235] We super-resolved ASC specks in human PD and AD biofluids and controls to extract information about the size and shape of individual aggregates. We used this information as additional metrics which may allow us to detect larger differences between controls and people with early disease and increase the accuracy of aggregate-based biomarkers even further. In our previous work, we showed that by combining the ratio-metric measurements of a-syn to Ap with the size and shape of a-syn aggregates imaged with dSTORM we could improve the accuracy for distinguishing PD from control serum from 85% to 90% AUC (Zhang, Y. P. et al. (2023)). Individual ASC aggregates are characterised by their area, perimeter, and circularity. Cumulative histograms of the area, perimeter and circularity distributions and their relative differences can be generated to investigate the morphological differences of the aggregates. By combining single-aggregate ASC speck imaging (SiMPull) with dSTORM, we can distinguish between aggregates of different sizes bigger than 30 nm (our resolution limit) which are the majority of our detected aggregates in serum and brain samples (-99%).

[0236] To characterise ASC specks accumulated in the brain, we super-resolved soluble ASC specks extracted from post-mortem human brain of PD patients. We found that ASC specks ranged in size between 0.035 pm (area = 0.001 pm2) to 0.37 pm (area = 0.108 pm2) (Fig. 4E) and they were smaller and rounder in PD than non-demented control brain (Fig. 4A). To check the trend was consistent across all patients, we examined individual aggregate trends for each patient and observed a clear separation in size and shape distributions of ASC specks between each patient and the controls. There were 12% and 8% higher proportions of ASC specks with area smaller than 0.018 pm2(p = 0.0028) and with circularity larger than 0.8 (p = 7.4 x 10'5) detected in PD compared to control brains representing the maximum fractional difference. By combining both parameters together (the area < 0.018 pm2and circularity > 0.8) we observed 8% higher proportion of ASC specks smaller than 150 nm and rounder in PD soaked brain compared to control soaked brain samples (p = 7.1 x 10'5, Fig. 4A). To relate to the ASC specks present in the periphery, we characterised the size and shape of ASC specks in serum (Figs. 5, 6). We found that ASC specks in both early PD (Fig. 5A) and AD (Fig. 6A) serum ranged in size between 35 nm (area = 0.001 pm2) to 360 nm (area = 0.1 pm2) (Fig. 5, 6G) with a higher proportion of smaller and rounder ASC specks than controls, similar to those in the brain (Fig. 4A). By combining the proportion of ASC specks that were smaller and rounder than a defined area and circularity threshold (area < 0.05 pm2and circularity > 0.5 in PD serum cohort and area < 0.04 pm2and circularity > 0.7 in AD serum cohort) we observed 2% and 7% higher proportion of these ASC specks in PD (p = 0.00063, Fig. 5A) and AD (p = 0.0039, Fig. 6A) than in control serum. When converting area into the size, Alzheimer’s and Parkinson’s patients have a significantly higher fraction of ASC specks smaller than 226 and 252 nm, respectively, in serum, than people without disease. We then made a combined threshold based on the proportion of aggregates with an area smaller and circularity higher than the thresholds (morphologically distinctive ASC specks) established above. This allowed us to distinguish controls from people with disease with an excellent accuracy (AUC = 100 % in PD human brain (Fig. 4B), AUC = 94 % in early PD serum (Fig. 5B) and AUC = 90 % in AD serum (Fig. 6B)). We have also been able to distinguish small but significant differences in size and shape of ASC specks between PD and AD serum. There was about 2.5% higher proportion of smaller and rounder ASC aggregates with area < 0.0021 pm2and circularity > 0.5 or perimeter < 0.145 pm and circularity > 0.5 present in serum of PD patients compared to AD. This discriminative area (0.0021 pm2) or perimeter (0.145 pm) is equivalent to the size of ASC aggregates of ~50 nm with more of them present in PD than AD serum. When setting these parameters of area and circularity as our combined threshold we can distinguish PD from AD diagnosis with an accuracy of 82% (AUC = 82%). We have been also able to improve the discrimination accuracy to 85% (AUC = 85%) using a combination of the defined perimeter and circularity thresholds for ASC specks (perimeter < 0.145 pm and circularity > 0.5).

[0237] Finally, we explored whether a combination of morphology of ASC specks together with the quantity of ASC specks and disease-specific protein aggregates measured in our SiMPull assay could further improve diagnostic biomarker performance (Fig. 5, 6D). We took the proportion of morphologically distinctive serum ASC specks established above and added them to a panel of the number of ASC specks, a-syn aggregates and Ap aggregates in PD serum and ASC specks, p-tau202 / 205 aggregates and Ap aggregates in AD serum to explore their different combinations using an unbiased approach (see the data analysis and statistical testing section for details). We found that the total number of ASC specks divided by the morphologically distinct fraction of ASC specks (total ASC / morphologically distinct fraction ASC ratio) increased 4.7-fold in disease and worked best at differentiating AD from controls (AUC = 100%, Fig. 6E). In PD (cohort 1), the ratio of the sum of the morphologically distinct fraction of ASC specks and total number of ASC speck to the number of Ap aggregates ((morphologically distinct fraction ASC + total ASC) / AP) was the most promising at discriminating early PD from control serum giving us AUC of 97% (Fig. 5E) and increased 1.7- fold in PD (Super resolution imaging of ASC specks was not performed for PD cohort 2).

[0238] Example 5: Single-molecular array analysis

[0239] The aim of this work was to develop a sensitive assay for the detection of Ap, a-syn, and pTau aggregates including common post-translational modifications, based on the SIMOA platform. Methods:

[0240] Protein aggregates / tau lysate preparation:

[0241] Recombinant a-syn aggregates.

[0242] Wild type a-syn samples were expressed, purified in E. coli and stored at -80°C as previously described (Cremades et al. Cell 149, 1048-1059 (2012)) and kindly provided by the Centre for Misfolding Diseases (CMD) at the University of Cambridge. To remove pre-aggregation seeds, the solution was centrifuged at 91000 g at 4°C for 1 h by an ultracentrifuge (Optima TLX Ultracentrifuge, Beckman). The concentration of the supernatant was then determined by A280 (E 280 = 5960 M -1 cm -1). The supernatant was then diluted to 70 pM in 1x PBS supplemented with 0.01% NaN3 (Merck, Cat. No. 71290) and incubated at 37°C with shaking at 200 rpm for 12 h or 48 h. The aggregates were aliguoted (20 pL) to avoid multiple freezethaw cycles. The aliguots were then snap-frozen and stored at -80°C.

[0243] In vitro A(342 one-week sonicated aggregates.

[0244] Lyophilised monomeric recombinant Ap42 peptide (Stratech, Cat. No. A-1170-2-RPE-1.0mg) was dissolved in PBS (pH = 7.4) at 200 pM on ice. The solution was guickly aliguoted and snap-frozen. To prepare recombinant Ap42 fibrils, an aliguot was thawed and diluted to 4 pM in 1x PBS supplemented with 0.01 % NaN3 (Merck, Cat. No. 71290) and incubated at 37 °C under guiescent conditions for one week. The Ap42 fibril was then sonicated as described previously (Corbett et al. Acta Neuropathol 139, 503-526 (2020)) with modification. The one- week aggregated Ap42 aliguot was immersion sonicated in an ice water bath with a 3-mm- titanium probe (Sonicator microprobe 4422, Qsonica) mounted on a tip sonicator (Ultrasonic processor Q125, QSonica) at 20 kHz with 40% of power for 24*5-s bursts with 15-s rests between bursts. Thereafter, the sonicated aggregate was centrifuged, aliguoted (50 pL) and snap-frozen. The aliguots were stored at -80 °C until use.

[0245] Preparation of cell-derived tau.

[0246] HEK293 cells expressing tau P301S-Venus under the CMV promoter (pcDNA3) were maintained in DMEM supplemented with 10% FCS, 100 U / ml penicillin, 100 pg / ml streptomycin and grown at 37 °C and 5% CO2. The cells were seeded with 50 nM heparin- assembled recombinant 6xHis-tau P301S assemblies in the presence of 1 % Lipofectamine2000 and a single clone was isolated (R1 E5) that stably propagates tau P301S- Venus aggregates. Cells were lysed in lysis buffer (1x PBS, 1% w / v Triton X-100, 1x complete™, EDTA-free Protease Inhibitor Cocktail mix, 1x PhosSTOP™ phosphatase inhibitor mix) on ice for 30 min. The lysate was then centrifuged at 14,000 x g for 15 min at 4 °C and the clarified lysate was aliguoted and stored at -20 °C. The total tau concentration was determined through ELISA, which is an upper limit for the amount of tau aggregates and used this sample as calibration standard. By measuring the HEK cell lysate using the tau silica bead calibrant using HT7 antibodies, which bind pan-tau, for capture and detection, we found that 0.1 ng / mL total tau monomer corresponds to 5.8 nM tau aggregates so that the conversion factor from 1 ng / mL to nM is 58. We ensured that the Venus tag does not interfere with the 488nm dye-labelled beads by testing 750nm dye-labelled beads and determined the total tau aggregate concentration in the cell lysate using tau silica bead calibrant (data not shown).

[0247] Calibration sample preparation

[0248] It is important to have calibration samples that can be included in each run in order to reduce run-run to variation. We therefore made silica nanoparticles coupled with a-synuclein, betaamyloid or tau as calibration samples as described below. g-synuclein SiNaP a-synuclein SiNaPs were prepared following the protocol described by Herrmann and colleagues (Herrmann et al. Clin Chim Acta 466, 152-159 (2017)). To 500 pL of carboxylated silica nanoparticles (cSiNaP) (particle size with 15-nm diameter), 11.47 pM in dimethylformamide (DMF) and 500 pL of 18.2-MO cm water was added. The mixture was then centrifuged (10,000 g, room temperature, 1 h) and the pellet was resuspended with 500 pL of MES buffer (2-(N-morpholino)ethanesulfonic acid) buffer (10 mM, pH 5.7). Meanwhile, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and sulfo-N-hydroxysuccinimide (sulfo-NHS,) were freshly dissolved in cold MES buffer (10 mM, pH 5.7) at 52.16 mM (10 mg / mL) and 92.11 mM (20 mg / mL), respectively. The cSiNaP was then diluted to 100 nM in MES buffer. EDC and sulfo-NHS were then introduced into the diluted cSiNaP suspension such that the mixture contained 400 pM EDC and 100 pM sulfo-NHS. The reaction mixture was sonicated for 30 min and then centrifuged (10,000 g, room temperature, 1 h). The pellet, i.e. activated SiNaP, was resuspended in fresh 10 mM MES buffer to give 200 nM suspension. Meanwhile, the monomeric a-synuclein solution (kindly provided by the Centre for Misfolding Diseases (CMD) at the University of Cambridge) was then diluted in 10 mM MES buffer to give 20 pM solution. To 1 mL of the activated cSiNaP suspension, 1 mL of diluted a-synuclein solution was introduced. The reaction mixture was placed on a revolver rotator for overnight incubation at room temperature. Finally, it was centrifuged (10,000 g, 4 °C, 1 h) and the pellet was redispersed in 1 mL of 1 :1 H2O:DMSO (v / v). The suspension was sonicated for 10 min and then centrifuged (5,000 g, 4 °C, 1 h). The a-synuclein-conjugated SiNaP was then redispersed in 400 pL of 1 :1 H2O:DMSO (v / v) to give a 500 nM (based on the SiNaP) suspension, and stored at -20 °C until use. AB42 SiNaP

[0249] Ap42 SiNaPs were prepared following the protocol described by Hulsemann and colleagues (Hulsemann et al. J Alzheimers Dis 54, 79-88 (2016)). The carboxylated silica nanoparticles (cSiNaP) were buffer-exchanged and activated by EDC and sulfo-NHS as described above. The reaction mixture was then centrifuged (10,000g, room temperature, 1 h) and the resulting pellet, i.e. activated SiNaP, was resuspended in fresh 10 mM MES buffer to give 200 nM suspension. Meanwhile, lyophilised monomeric recombinant Ap42 peptide (Stratech, Cat. No. A-1170-2-RPE-1 .Omg) was dissolved in PBS (pH = 7.4) at 200 pM on ice. An aliquot of Ap42 peptide solution was then diluted in 10 mM MES buffer to give 20 pM solution. To 1 mL of the activated SiNaP suspension, 1 mL of diluted Ap42 solution was introduced. The reaction mixture was placed on a revolver rotator for overnight incubation at room temperature. Finally, it was centrifuged (10,000 g, 4 °C, 1 hour) and the pellet was redispersed in 1 mL of 1 :1 H2O:HFIP (1 ,1 ,1 ,3,3,3-Hexafluoro-2-propanol) (v / v). The suspension was sonicated for 10 min and then centrifuged (5,000 g, 4 °C, 1 hour). The Ap42-conjugated SiNaP was then redispersed in 200 pL of 1 :1 H2O:DMSO (v / v), and stored at -20 °C until use.

[0250] Tau SiNaP

[0251] To the 30-nm triethoxylpropylaminosilane silica nanoparticles (nSiNaP, Merck, Cat. No. 791334, 16.59 pM in water, 500 pL), 500 pL of DMF was introduced. The mixture was firstly centrifuged (10,000 g, room temperature, 1 hour) and resuspended with 500 pL of DMF. Meanwhile, the NHS-activated carbonylacrylic reagent was dissolved in DMF to give a 1 M stock solution. N,N-diisopropylethylamine (DIPEA) was diluted in DMF to give a 1 M stock solution. To 5.6 pL of 1 M DI PEA solution, 96.4 pL of the bead suspension was introduced, followed by the addition of 5.6 pL of 1 M carbonylacrylic linker solution. The reaction was incubated overnight at 37 °C with shaking at 200 rpm in dark. The reaction mixture was centrifuged (15,000 g, room temperature, 15 min) and the pellet was resuspended in 500 pL of DMF. The suspension was then centrifuged at (15,000 g, room temperature, 15 min) and the pellet was resuspended with 96.4 pL of water. Meanwhile, a 10x Tris buffer (500 mM Tris, pH 8.5) was prepared. Tris(2-carboxyethyl)phosphine hydrochloride (TCEP) was dissolved in water to give a concentration of 100 mM and the pH was adjusted to 7-8. To the pellet suspension, 60 pL of 10x Tris buffer and 3.2 pL of TCEP solution were added, followed by the addition of 0.5 mL of RP hTAU solution (32 pM in 100 mM Tris, 150 mM NaCI and 1 mM EGTA). The reaction was incubated overnight at 37 °C with shaking at 200 rpm. Next, the reaction mixture was centrifuged (15,000 g, room temperature, 15 min) and the pellet was resuspended in 500 pL of 1 :1 H2O:DMSO solution (v / v). The mixture was then sonicated for 10 min and centrifuged (15,000 g, room temperature, 15 min). The pellet was resuspended in 160 pL of 1 :1 H2O:DMSO solution (v / v) to give 10 pM, and stored at -20 °C until use.

[0252] Synthesis of NHS-activated carbonylacrylic reagent

[0253] All reagents and solvents were purchased from commercial suppliers and used as received. THF was purified as reported by Pangborn et al. (Organometallics 15, 1518-1520 (1996)), solvent was pre-dried over sodium wire and then distilled from calcium hydride and lithium aluminium hydride. Merck Silica gel 60 was used for the flash column chromatography. Monitoring of reactions was performed using TLC Silica gel 60 F254 plates. Compounds were detected using shortwave (254 nm) UV lamp or by staining with an indicated solution prepared by known procedures. NMR spectra were recorded on Bruker 400-Avance III HD, Avance DPX-400, 400-QNP Cryoprobe (400.1 MHz for 1 H) in DMSO-cftS (referenced to the residual solvent signal). Chemical shifts are given in ppm (b-scale), coupling constants ( ) in Hz.

[0254] 2,5-Dioxopyrrolidin-1-yl (E)-4-oxo-4-phenylbut-2-enoate (2)

[0255] A solution of 3-benzoylacrylic acid (1) (1 g, 5.68 mmol) and N-hydroxysuccinimide (690 mg, 6.0 mmol) in anhydrous THF (25 ml) was cooled to 0°C and N,N' dicyclohexylcarbodiimide (1.24 g, 6 mmol) was added with stirring. The reaction mixture was stirred at 0°C for 1 h and then it was kept in a freezer (-20°C) overnight. The formed N,N-dicyclohexylurea was removed by filtration, washed with ether and solvents were removed under reduced pressure. The crude product 2 (1.084 g, 4.0 mmol, 70%) was isolated as a yellow solid after recrystallization form isopropanol (40 ml).

[0256] Rf 0.93 (10% MeOH / DCM, UV).

[0257] 1 H NMR (400.1 MHz, DMSO-d6) 5 8.29 (d, J = 15.7 Hz, 1 H, H3), 8.10-8.07 (m, 2H, H2”), 7.76-7.72 (m, 1 H, H4”), 7.62-7.53 (m, 2H, H3”), 7.03 (d, J = 15.7 Hz, 1 H, H2), 2.88 (s, 4H, H2’). 1 H NMR data are in accordance with the previous reports (Jakubec & Berkes, Tetrahedron: Asymmetry 21 , 2807-2815 (2010)). Single-molecule Pull-down (SIMPull)

[0258] Single-molecule Pull-down (SiMPull) experiments were performed as previously described (Bdken et al. Angew. Chem. Int. Ed. Engl., e202317756 (2024)). Briefly, functionalised coverslips were incubated with NeutrAvidin, washed, incubated with biotinylated capture antibody (a-syn: SC211 , Ap: 6E10, tau: HT7). After a further wash, the diluted silica- nanoparticle calibrators were incubated for 1 h at RT or overnight at 4 °C. Subsequently, the wells were washed, blocked with BSA and labelled detection antibody (a-syn: SC211 , A : 6E10, tau: HT7) was added, followed by a further wash. Imaging was performed on a homebuilt total internal reflection fluorescence (TIRF) microscope. Super-resolution imaging was performed either using STORM (Ap and tau calibrators) or DNA-PAINT (a-syn calibrators).

[0259] Single-molecular array magnetic bead coupling

[0260] We selected two tau antibodies: the phospho-tau specific antibody AT8 (p-Ser202, p-Thr205) and the total tau antibody HT7 for detection of tau aggregates. 4B12 was used for detection of a-synuclein aggregates and 6E10 was used for detection of Amyloid beta (Ap) aggregates. The use of the same capture antibody for detection and capture ensures that monomers give no signal and the signal comes from dimers or larger protein complexes. This was confirmed by performing denaturation experiments on synthetic aggregates.

[0261] Antibody-bead conjugation was performed as per manufacturer’s instructions (Quanterix). Briefly, paramagnetic carboxylated beads were vortexed for 30 sec and left on a rotator to mix gently for at least 10 min. The beads were then washed three times with wash buffer followed two times with conjugation buffer. EDC (0.3 mg / mL) was used to activate the beads by placing the mixture in a rotator at 2-8°C for 30 min. The beads were once again washed with conjugation buffer. Buffer exchange of 100 pg of antibody was performed with a 50 kDa Amicon Ultra Spin Column device to change the storage buffer of the antibody to conjugation buffer. The concentrated antibody was recovered to ~ 100 pL volume and the antibody concentration was measured by A280. The antibody was diluted to 0.2 mg / mL in 300 pL of ice cold bead conjugation buffer and used to resuspend the washed beads. The mixture was placed on a rotator at 2-8°C for 120 min. The antibody-coated beads were then washed twice with wash buffer and antibody coating efficiency was determined by measuring the residue antibody concentration in the solution and supernatant of the first wash. Bead blocking buffer was used to block the beads with 45 min incubation on a rotator. Finally, the beads were washed once with wash buffer followed by two washes with diluent buffer, and stored as a pellet at 4°C until use. Antibodies

[0262] The following antibodies were used in this study: SC211 (Santa Cruz, Cat. No. SC767), 4B12 (BioLegend, Cat. No. 807-808), 6E10 (BioLegend, Cat. No. 803007), HT7 (ThermoFisher Scientific, Cat. No. MN1000), AT8 (ThermoFisher Scientific, MN1020).

[0263] Single-molecular array plate preparation

[0264] A total volume of 100 pL of each sample, diluted with Detector / Sample Diluent to desired concentration, were added to a SIMOA 96-Well Plate (Quanterix). These samples included bead calibration samples over a range of different concentrations. Approximately 500k antibody-conjugated beads were added to each well and incubated on a shaker at 30°C and 800 rpm for 30 min (or 60, 120 min). All washing steps were performed using the automated 3-step-assay program on the SIMOA Microplate Washer. The assay wells were washed three times with wash buffer. To each well, 100 pL of biotinylated detector antibody (0.3 / 0.5 / 1 pg / mL, in Detector / Sample Diluent) was added and incubated on a shaker at 30°C and 800 rpm for 10 min (or 30 min) followed by three washes with wash buffer. SpG (streptavidin p - galactosidase) was diluted to a concentration of 150 pM (or 50 pM, 300 pM) with SpG Dilution Buffer and 100 pL of the solution was added to the assay wells. The microplate was once again incubated on a shaker at 30°C and 800 rpm for 10 min followed by three washed with wash buffer. Two final washes were performed with wash buffer and 1 min incubation on a shaker. Finally, the buffer in the assay wells was completely removed and the beads were left to dry for 10 min.

[0265] Single-molecular array processing

[0266] The following steps were automated by using a Quanterix SR-X™ Instrument. The beads were resuspended using pre-warmed RGP (resorufin p-D-galactopyranoside) substrate. Then these beads suspensions were transferred to a sample well (the “flow cell”) on the SIMOA Disc. A low-power vacuum pulls the bead suspension into the disc channel, allowing the bead suspension to flow over an array of more than 200 thousand microwells, each large enough to hold a single bead. Oil was used to remove the excess beads on the surface of the array and seal the loaded wells. Then the wells were imaged. The wells containing beads with at least one immunocomplex attached gave out fluorescence signals, while the plain beads did not. The fraction of beaded wells with enzyme activity (LN) was then reported by the machine.

[0267] Single-molecular array data analysis

[0268] The readout of SIMOA assays is presented in terms of the fraction of beaded wells with enzyme activity (foN). This is subsequently used to calculate the average enzyme per bead (AEB). When the fraction of beaded wells with enzyme activity (LN) is lower than 0.7 (i.e. with less than 70% active beads), AEB is calculated in digital mode, assuming Poisson statistics for the number of captured molecules per “on” bead. In this case, AEB is determined by:

[0269] As for the wells with foN greater than 0.7, uncertainty rapidly grows to unacceptable levels for the Poisson distribution, the AEB is then switched to analog mode based on the intensity of a single enzyme, and calculated by: n p-'p analog

[0270] WHERE: / ON X fbead

[0271] — In [1 — ON] in arrays where O < 0-1

[0272] We note however that for these assays, aggregates are being detected which may be bound by more than one enzyme. This means that the analog mode for calculating AEB cannot be used, and we revert to using digital mode. We avoid any sample or calibrant concentrations where foN > 0.9. Furthermore, while we use the term AEB (average enzyme per bead) for the purpose of consistency, in our case more than one enzyme can be bound per aggregate.

[0273] After obtaining the AEB values at each calibration level, a four parameter logistic (4PL) curve was fitted (concentration of calibrators as x and AEB as y, and blanks were not included when establishing the calibration curve). The data points are weighted by 1 over ? / 2during the fitting process. (4PL equation)

[0274] The preliminary limit of detection (LoD) for each assay was determined by the parameter A with a set multiplier (normally 1.3) and fitted back to the curve to get the LoD in units of concentration. The lower limit of quantification (LLoQ) was the lowest valid calibration level with coefficient of variation (%CV) less than 25%.

[0275] Denaturation protocol

[0276] A342 early oligomers.

[0277] Ap42 early oligomers were diluted in PBS to 500 nM with different concentrations of guanidine hydrochloride (GdnHCI, 6 M, 4 M, 2 M, 1 M, 0.5 M, 0.25 M and 0 M) from an 8 M guanidine hydrochloride solution (pH 8.5, Merck, Cat. No. G7294). The mixture was then heated to 80°C for 10 min. It was then immediately diluted to 0.5 nM with Detector / Sample Diluent (Quanterix, Cat. No. 101359). The diluted mixture was then immediately either loaded into a SIMOA 96- Well Plate (Quanterix, Cat. No. 101457) or stored on ice until use. g-syn early oligomers.

[0278] Recombinant 12 h a-syn aggregates were diluted in PBS to 17.5 nM with different concentrations of guanidine hydrochloride (GdnHCI, 6 M, 4 M, 2 M, 1 M, 0.5 M, 0.25 M and 0 M) from an 8 M guanidine hydrochloride solution (pH 8.5, Merck, Cat. No. G7294). The mixture was then heated to 80°C for 10 min. It was then immediately diluted to 0.2 nM with Detector / Sample Diluent (Quanterix, Cat. No. 101359) and kept on ice.

[0279] Assay validation

[0280] For the validation of each assay, four plates were run across two days using calibrator samples as well as independently prepared quality control samples across the calibration range. The quality control samples were analysed against the calibration curve, and the obtained concentrations were compared with the nominal value to obtain the accuracy as a percentage of the nominal value. The accuracy was determined within a single plate (within-run accuracy) as well as across different plates (between-run accuracy).

[0281] Information on serum samples

[0282] Patients with early Alzheimer’s disease includes the patients who sought medical advice at a memory clinic for the first time and were diagnosed for AD based on the positive AD CSF biomarker profile. AD CSF biomarker profile was defined by a CSF Ap42 / p-tau181 ratio < 10.25 (as measured by Lumipulse G600II, Fujirebio). The study protocol was approved by the regional ethics committee at the University of Gothenburg.

[0283] Results:

[0284] Establishing SIMOA assays for the selective detection of protein aggregates

[0285] We set out to develop SIMOA aggregate assays to detect protein aggregates relevant to neurodegenerative diseases, including native and C-terminal truncated a-synuclein (a-syn), beta-amyloid (Ap), tau, and phosphorylated tau (p-tau). For the specific detection of these aggregates - and not monomers - we used the same monoclonal antibody for both capture and detection. This configuration required the presence of two or more identical epitopes within a single aggregate; capturing the aggregate would occupy one binding site and require a second site for binding of the detection antibody, ensuring that the detected species are, at a minimum, dimeric. The assay development process consisted of antibody pair selection, assay condition optimisation (detector concentration and SBG concentration), optimisation to achieve lower coefficient of variation at all calibration levels including adjusting the dynamic range, assay validation, and diluent optimisation for specific sample matrices.

[0286] To achieve accurate quantification and ensure normalisation between runs, we first developed suitable calibrators for these assays. This is a challenging task since the calibrator needs to have a known concentration and be reproducible. Since the aggregation process of monomers is stochastic and the produced aggregates have an unstable, unknown concentration, in vitro aggregates are not suitable calibrators. To have a reliable aggregate mimic, we coupled a-syn, Ap42, or tau monomers to silica nanoparticles. We used single-molecule and super-resolution microscopy to check if the calibrators aggregate (Figures 7A-C). These experiments showed no significant aggregates of the coated silica nanoparticles making them suitable calibrators for our purposes. Since it is not possible to get a sample of tau in which every single monomer is phosphorylated at S202 and T205 (AT8), the beads could not be used for (p-)tau. To solve this issue, we used cell lysate from HEK cells which stably propagate tau aggregates that are highly phosphorylated (Bdken et al. Angew. Chem. Int. Ed. Engl., e202317756 (2024)). This allowed us to use the same sample for measurement of total tau aggregates and tau aggregates phosphorylated at S202 and T205 (AT8-positive).

[0287] We then proceeded to select the antibody pairs for each assay. For the development of the a-syn assays, we selected the 4B12-4B12 and sc12767-sc12767 (SC211) antibody pairs based on their target specificity (truncated and total, respectively) and limits of detection. It is known that significant truncation of a-syn occurs during the pathogenesis of PD, increasing aggregation. Importantly, this truncation can remove the epitope detected by the SC211 antibody, which is closer to the C-terminus. Use of the 4B12 antibody, which has an epitope closer to the centre of the a-syn molecule, allows us to detect truncated a-syn along with full length a-syn. By detecting both we can estimate the concentration of truncated a-syn aggregates. For the tau assay, we selected two antibodies that are commonly used in the field: the Ser202 and Thr205 phosphorylated tau-specific antibody AT8 and the total tau antibody HT7. For the Ap assay, we used the 6E10 antibody, which has a binding region within residues 3-8 of A 36. All five antibodies selected (4B12, SC211 , 6E10, HT7, AT8) are monoclonal and have a single epitope, and as such only bind to a single epitope on each monomer. We ensured that using these antibody pairs we were able to detect the respective silica- nanoparticle and lysate calibrators explained above, generating a concentration-dependent signal which can be used as calibration curves for each assay (Figure 8A-D). To ensure the accuracy of our protein aggregate quantification assays, we evaluated specificity from two key perspectives. Firstly, the assay must be specific to the analyte of interest. While this specificity largely depends on the antibodies used, it can also be influenced by other reagents involved in the assay. To test this, we assessed the cross-reactivity between Ap, a-syn, and tau aggregates by measuring the recombinant aggregate samples using mismatched SIMOA assays. The results showed negligible cross-reactivity, confirming high analyte-specificity (Figure 9A). Secondly, the assay must specifically detect aggregates rather than monomer. To evaluate this, we performed denaturation experiments of the aggregates with increasing concentrations of guanidinium chloride (GdnHCI), and a short heat treatment (85 °C for 10 min). Following this treatment, 99% of the signal disappeared showing that the assays are aggregate specific (Figure 9B).

[0288] To further enhance assay performance, we optimised the sample diluent, detector, and SBG concentration for each assay individually, evaluating performance based on the signal-to- background ratio at various calibration levels (data not shown). Initially, we identified the optimal sample diluent by testing a range of commercially available diluents (Quanterix Sample Diluents A-E), alongside the standard sample diluent and, specifically for the tau aggregate assays, the tau 2.0 sample diluent. The standard diluent is phosphate buffer with saline. The Quanterix sample diluents are: A, phosphate buffer with bovine serum components, a heterophilic blocker, and a surfactant; B, phosphate buffer with protein stabilisers (bovine), a heterophilic blocker, and a high surfactant concentration; C, phosphate buffer with low concentration of protein stabilisers (bovine), a heterophilic blocker, and a surfactant; D, phosphate buffer with newborn calf serum, a heterophilic blocker, and a surfactant; E, Tris buffer with high pH, bovine serum components, a heterophilic blocker, and a surfactant. The Tau 2.0 diluent contains BSA and Calf serum and an antimicrobial.

[0289] For the detector and SBG enzyme concentrations, we explored various combinations of their concentrations together to account for potential combined effects, ensuring the optimal interplay between these components. Using these optimised assay conditions, we achieved the following limits of detection (LoDs): 4.2 pM for 4B12 assay, 0.63 pM for SC211 , 0.92 pM for 6E10, 17 pM for HT7, 37 pM for AT8. The optimal assay conditions and respective LoDs are shown in Table 4. Table 4: Optimised conditions and limits of detection for aggregate assays. * indicates assay conditions for serum sample specific assays.

[0290] To confirm the reproducibility of the assays, we validated their accuracy and precision across using four technical replicates across tyvo days using calibrator samples as well as independently prepared quality control samples across the calibration range. All assays had a coefficient of variation (CV) below the accepted 20% threshold throughout the working range. The accuracy of all assays laid between 80%-120% for all calibration samples (Figure 9C) as well as for independently prepared quality control samples (n = 5 technical replicates on each plate) at multiple concentrations selected in the dynamic range of the respective assay.

[0291] SIMOA aggregate assays detect soluble aggregates in the human brain

[0292] Once the assays were optimised, we applied them to post-mortem human brain homogenate samples to demonstrate their capabilities to soluble protein aggregates in relevant biologically complex samples. Given the critical role of these aggregates in AD, post-mortem brain homogenate from AD and age-matched control samples was tested. We first ensured that these assays detect aggregates in a concentration-dependent manner in a linear range. For this purpose, brain homogenate from 5 AD and 5 control samples were combined and tested across a wide range of dilutions (a-syn 1 :4 to 1 :256, Ap 1 :6.25 to 1 :400, tau: 1 :2000 to 1 :64000, data not shown). We confirmed the linear working range of the assays for brain homogenate by interpolating the protein aggregate concentration from the standard curves and calculating the final concentration of the aggregates depending on the dilution factor, observing only little variation across the dilution range. This positive correlation of the brain homogenate concentration and the readout confirms the detection of aggregates in this sample type.

[0293] We then proceeded to test brain homogenate from 5 AD (frontal cortex, BA6 / 8, Braak Stage VI) and 5 age-matched controls (frontal cortex, BA6 / 8, Braak Stage 0; Figure 10A). Using the calibrator samples to calculate the aggregate concentration, all samples were above zero, showing the detection of aggregates in these samples. Notably, we saw a significant increase in tau and p-tau aggregates with AD (Figure 10B-G). The mean concentration of a-syn aggregates detected by SC211 was 109 pM and 248 pM (SC211 assay, Welch’s t-test, p = 0.42) and 107 pM and 88 pM for aggregates detected by 4B12 (Welch’s t-test, p = 0.69), respectively in control and AD brains (Figure 10B and C), with no significant differences between control and disease. The Ap concentration was 3060 pM in control brain and 1272 pM in AD (Welch’s t-test, p = 0.14, Figure 3D). This difference may reflect that there is a higher number of small A aggregates in control brain and there is a small number of large aggregates in AD brain. The total tau (HT7) and AT8-positive aggregate concentration were and 6446 pM and 6372 pM respectively in AD brains, compared to 328 pM and 218 pM in control brain (Welch’s t-test, HT7: p = 0.023, AT8: p = 0.018, Figure 10E and F). Thus, there were significantly increased levels of aggregated tau in AD brain. Overall, these results show that the dominant protein aggregate present in AD brain is tau. The ratio of AT8 / HT7 aggregates was ~1 in AD and < 0.5 in controls, showing that the majority of aggregates detected in AD were phosphorylated at S202 and T205 (Welch’s t-test, p < 0.0001 , Figure 10G).

[0294] SIMOA aggregate assays detect aggregates in human serum

[0295] We confirmed the assays are compatible with readily available and clinically relevant samples, such as human serum. The high concentration of albumin, as well as other proteins and lipids, creates a complex matrix that may affect the diffusion and detection of the aggregates in serum. To address this, the detector and SBG concentration was optimised for serum samples specifically for Ap, pTau, and a-syn aggregates. Similar to the brain homogenate, we verified the detection of aggregates in a linear range by testing a range of concentrations, observing a positive correlation between serum concentration and signal (data not shown). After further optimising the assays for human serum, we analysed serum samples taken from patients from a memory clinic in Gothenburg, Sweden, who diagnosed with early-stage AD based on a positive CSF biomarker profile (n = 20). Patients who were negative for this biomarker profile were used as the control group (n = 20). The results are shown in Table 5. Table 5. Results of aggregate measurements on AD and control serum. Values shown in unit of pM in format of mean (S.D.).

[0296] While a-syn aggregate levels did not differ between AD and control (Figure 11 A,B), there was a trend for higher Ap (Figure 11C) and total tau (Figure 11 D) in the CSF biomarker positive cases accompanied by significantly higher levels of AT8-positive tau aggregates (Figure 11C). Moreover, we observed tenfold higher levels of total (C terminally truncated + full length a-syn) aggregates than full length only aggregates, suggesting that most of the a-syn aggregates in these samples are C-terminally truncated. The ratio of AT8-positive tau aggregates to total tau aggregates was 3.1 and 0.67 for AD and control serum respectively (Figure 11 F), suggesting higher phosphorylation of tau in early AD..

[0297] We then explored combining these measurements into a combined biomarker for AD to see if this could improve the diagnostic accuracy, as was done for PD. All the potential biomarkers evaluated, along with their performance are tabulated in Table 6. The ratios were calculated with fitted concentrations in unit of pM. If the numerator was measured at a concentration below the LoD, the ratio was considered as 0. While on the other hand, when the denominator was below the LoD, the ratio was considered as infinity. We found that the combined biomarker (SC211 / 6E10) + (HT7 / AT8) (i.e. (a-syn I A ) + (tau I p-tau)) gave good discrimination with an area under the curve in the ROC curve of 0.91 , where a higher value of biomarker represents lower likelihood of AD.

[0298] We also found that Ap aggregates / (Ap + a-synuclein + tau aggregates) gave a ROC curve with an AUC of 0.81. Other combined biomarkers tested included: (Ap + p-tau aggregates) I a-syn aggregates gave a ROC curve with an AUC of 0.84, and (Ap + p-tau aggregates) I (a- syn aggregates + (p-tau aggregates I tau aggregates)) gave a ROC curve with an AUC also of 0.84. Table 6: Table for all combined biomarkers tested. Example 6: Brain-derived tau aggregates on SiMPull

[0299] SiM Pull experiments were conducted as described in Example 1 , for detection of brain-derived tau aggregates in plasma from AD and non-AD patients (Figure 12). The anti-tau antibody, 8B9, was used to detect tau aggregates. In particular, this antibody detects tau aggregates that have formed in the brain (i.e. are brain derived). SiMPull methods were able to detect brain-derived tau aggregates in biofluids and distinguish between AD and non-AD patients.

[0300] Example 7: Further investigation of size and shape of ASC specks in biofluids of people with AD as a biomarker

[0301] The size of ASC specks was investigated as a sensitive biomarker in early-stage AD and the results are presented in Figure 13. Morphology analysis of ASC specks was detected in the control (n = 14) and early-stage AD (n = 14) cerebrospinal fluid (CSF) samples using dSTORM. Comparing the number (Figure 13A) and fraction (Figure 13B) of individual ASC aggregates indicated that are smaller (area < 0.03 pm2) and rounder (circularity > 0.5) than the defined threshold in the CSF of people with early AD vs age-matched controls. Examples of super-resolved ASC aggregates in AD and control CSF samples are shown in Figure 13C. The findings were confirmed by correlation in the number of ASC specks in paired serum versus CSF samples from 20 early AD patients and 20 HC participants (Figure 13D) and ROC curve analysis (Figure 13E and F).

[0302] Example 8: Size and shape of ASC specks as a differential biomarker

[0303] The size of ASC specks was investigated as a candidate differential biomarker between FTD syndromes and AD and the results are presented in Figure 14. Morphology analysis of ASC specks was detected in the FTD (n = 10) and AD (n = 10) serum using dSTORM. The FTD cohort included 3 patients with Behavioural Variant Frontotemporal dementia [bvFTD], 1 patient with Behavioural Variant Semantic Frontotemporal dementia [BvSdFTD], 3 patients with Semantic Variant Primary Progressive Aphasia [PPAsv] and 3 patients with Nonfluent / agrammatic Primary Progressive Aphasia [PPAnfv]. The fraction of individual ASC aggregates with an area > 0.0122 pm2(Figure 14A) or circularity > 0.88 (Figure 14D) was compared in the serum of people with FTD syndromes vs AD. It was found that ASC specks were smaller in FTD serum, with the smallest ASC specks in patients with Semantic Variant Primary Progressive Aphasia (PPAsv). ROC curve analysis of the identified phenotypes allowed FTD to be distinguished from AD with AUC = 86% using the area metric (Figure 14B) and AUC = 81% using the circularity metric (Figure 14E) of ASC specks. Example 9: ASC speck composite serum biomarkers as a differential biomarker

[0304] Composite serum biomarkers using ASC speck were investigated as differential biomarkers using SiMPull methods as described hereinbefore. The results are presented in Figure 15. It was found that the ASC I p-tau ratio improves differential diagnosis between FTD and AD serum. The (ASC + p-tau-AT8) / Ap (wherein “p-tau-AT8” represents p-tau202 / 205) composite biomarker was able to better distinguish different FTD syndromes (same cohort as described in Example 8).

[0305] Discussion:

[0306] Immune system dysregulation and neuroinflammation are major pathophysiological drivers of disease progression in AD (Heneka, M. T. et al. (2015)) and PD (Jewell, S. et al. (2022)). We introduce a new assay that has potential as an ultrasensitive biomarker of inflammation in the blood of people with AD and PD. The single-molecule pull-down detection assay was combined with dSTORM, enabling measurement of the size (to 30 nm resolution) and shape of individual extracellular inflammasome ASC specks in the serum, as well as in the CSF and brain samples from people with PD and AD.

[0307] We used the same ASC antibody for capturing and detecting ASC specks from non-processed biofluids which allowed us to specifically measure inflammasome-active ASC aggregates. The quantity of extracellular, secreted ASC specks measured in the blood and CSF has utility as a biomarker of inflammation and inflammasome formation for PD and AD. By combining the quantity of ASC specks with measures of Ap, a-syn and p-tau aggregates into different ratios, we achieved high diagnostic accuracies of 92% for AD ((p-tau202 / 205 + ASC) / Ap ratio) and of up to 98% for PD (both (a-syn + ASC) / Ap and ASC / AP). The use of ratios normalised to Ap levels instead of absolute aggregate quantities helped us to achieve greater biomarker performance and a larger dynamic range between patients and controls (from 1.5-1.7 fold increased single ASC speck biomarker to 3-5 fold increased (a-syn + ASC) / Ap ratio in PD serum compared to controls and from 1.5-fold increased single ASC speck biomarker to 6- fold increased (p-tau202 / 205 + ASC) / Ap ratio in AD serum compared to controls). We also showed that ASC / p-tau202 / 205 ratio might be useful as a sensitive differential biomarker between different dementia types that allowed us to distinguish AD dementia from FTD with an accuracy of 91 %. These combined biomarkers suggest that there is increased inflammation and production of a-syn aggregates in PD and increased inflammation and production of tau aggregates in AD and FTD. There is little change in the number of Ap aggregates in serum in either disease allowing the number of Ap aggregates to be used for normalisation. The increased accuracy of these new ASC speck composite biomarkers in discriminating PD cases which were close to diagnosis (average duration 6 months) from age-matched controls, together with validation in an independent group of samples, suggests that these markers may show promise as an early diagnostic biomarker for PD. Future studies using blood samples from patients with prodromal disease will be important to assess whether the marker can accurately predict conversion to manifest disease, whilst longitudinal studies are needed to assess whether inflammation changes during disease progression and in response to future disease modifying drugs can be assessed.

[0308] By super-resolving ASC specks, we detected aggregates ranging in size from 0.035 pm to about 0.36 pm in both serum and brain. The soluble ASC aggregates from the post-mortem brain were present in similar morphology to those detected in serum, but PD patients had a significantly higher proportion of smaller and rounder ASC aggregates than controls. In comparison, in the THP-1 cell experiment with nigericin stimulation for 3.75, 7.5, 15, 30 and 60 mins there were about 30% reduction in the number of ASC specks detected in the cell lysates at 15 min nigericin accompanied by high levels of released ASC specks in the matched media. This suggest that pyroptosis of significant proportion of cells occurs by this early time point. Morphologically ASC specks at 15 min post nigericin were the smallest and roundest compared to other time points. This correlates with the morphology of ASC specks detected in disease samples. This may suggest that the increased number of smaller, rounder ASC specks we see in the disease state are from cells dying due to rapid inflammasome activation in disease. For diagnostic potential, measuring the proportion of ASC specks that are smaller and rounder than threshold (morphologically distinctive ASC specks) can distinguish people with early disease from controls with an excellent accuracy (AUC = 94 % in PD serum and AUC = 92 % in AD serum) and between different types of dementia (AUC = 91%, AD dementia from FTD). ASC speck morphological information has an even better diagnostic value when combined with the quantity of ASC specks: the total number of ASC specks divided by the morphologically distinct fraction of ASC specks (total ASC / morphologically distinct fraction ASC) in AD serum had an accuracy (AUC) of 100% and the (morphologically distinct fraction ASC + total ASC) / Ap in PD serum had an accuracy (AUC) of 97%. It will be important to perform similar measurements on a larger number of samples in future work to confirm these findings.

[0309] Collectively, our data support the hypothesis that the NLRP3 inflammasome is activated in early PD and AD, and lend further support to the idea that specific targeting of inflammasomes and / or their downstream constituents such as ASC, caspase 1 or gasdermin D (Mangan, M. S. J. et al. (2018)) warrants testing as a disease-modifying strategy for PD / AD (Heneka, M. T. et al. (2013)). Previous studies in multiple rodent PD and AD models have shown that administration of the NLRP3 inflammasome inhibitor, MCC950 (Coll, R. C. et al. (2015)), helped to reduce neuroinflammation and neurodegeneration, promote a-syn and Ap aggregate clearance and improve cognitive and motor symptoms in these rodents (Dempsey, C. et al. (2017); Gordon, R. et al. (2018)). In addition, pharmacological blockade of the NLRP3 inflammasome with OLT1177 reduces neuroinflammation, dopaminergic degeneration and motor deficits in the MPTP mouse model of PD and increases clearance of alpha-synuclein in microglia in vitro (Amo-Aparicio, J. et al. (2023)). Blocking the inflammasome ASC protein from aggregation with an antibody (Venegas, C. et al. (2017)) or nanobody (Bertheloot, D. et al. (2022)) appears as another attractive pharmacological target. In this regard, our method of detecting the decrease in ASC aggregates might be useful for screening the effectiveness of different ASC aggregation inhibitors in vitro, and as a biomarker to detect target engagement of inflammasome inhibitors in clinical trials.

[0310] We have also developed quantitative, highly sensitive, and selective SIMOA-based assays for detecting aggregates of a-synuclein (a-syn), beta-amyloid (Ap) and tau with detection limits in the low picomolar range. We confirmed the specificity for detecting aggregates only (and not monomers) and showed that the assays are not cross-reacting with other protein aggregates. Additionally, our assays can detect key post-translational modifications, including C-terminally truncated a-syn and AT8-phosphorylated tau aggregates. We performed extensive bioanalytical validation to show the reproducibility and stability of all our assays.

[0311] These assays have been optimised for use in a wide range of complex biological samples, including human serum, post-mortem brain tissue, as well as model systems such as cerebral organoid conditioned media and mouse models. The high sensitivity of our assays is crucial, as small, soluble aggregate levels in these samples are typically very low and often fall below the detection limits of other techniques. This sensitivity allows for the reliable detection of aggregates even in challenging biological matrices. This versatility makes our assays a valuable tool for studying disease mechanisms across various biological contexts. By applying these assays to different sample types, the assays can be used to gain insights into the aggregation processes and post-translational modifications occurring in vivo and in vitro, enhancing our understanding of neurodegenerative disease progression.

[0312] We demonstrated the capabilities of our assays to detect soluble aggregates in biological samples by applying them to human brain tissue. Post-mortem brain homogenate from AD samples was first tested at different dilutions providing concentration-dependent readouts. This was a proof of principle for detection of all the aggregates in this sample type and helped determining the linear range. Then we compared the AD brain homogenates with age- and sex-matched control samples. The decision of measuring all the aggregates in the same sample was made to allow their direct comparison. These assays can detect a-syn, Ap and tau aggregates in human post-mortem brain samples which will enable them to be used in future studies providing novel insights into the molecular mechanisms underlying neurodegenerative diseases. This approach of detecting small soluble aggregates with high sensitivity and specificity also holds potential for investigating the efficacy of therapeutic interventions aimed at reducing or modifying protein aggregates in various neurodegenerative conditions.

[0313] We then explored the application of our assays to human serum which is an easily accessible (non-invasive) biofluid that can be used for the diagnosis of dementia and track its progression. Serum samples from individuals who individuals who visited a memory clinic for the first time and showed a positive CSF biomarker profile were compared to samples from patients who also visited the memory clinic but showed a negative CSF biomarker profile. The assays' high sensitivity allowed us to use less than 150 pL of sample in total to test for all the aggregates, making them suitable for samples with limited availability and enabling the application of multiple assays to each patient. Similar to human brain homogenate, we showed a concentration-dependent readout for all aggregates showing the assays’ capability of detecting a-syn, Ap and tau aggregates in human serum samples. Overall, cases classified as AD by the CSF biomarker profile, had significantly higher levels of AT8-positive tau aggregates and a strong trend for higher total tau (CI95 = -2.62, 54.65) and A (CI95 = -2.83, 71.54) aggregate levels. Once we added Ap, HT7 and AT8 for each patient the difference was significant (CI95 = 70.61 , 378.97).

[0314] In conclusion, we have developed a robust and sensitive method for detecting a-syn, Ap and tau aggregates, including post-translationally modified forms, in human brain homogenate and serum. It should be noted that the techniques and calibrators described here are not limited to the antibodies used in this study. We selected these antibodies for their common usage in the field of dementia research, but any monoclonal antibody can be applied. Our results demonstrate the potential of this method for understanding how aggregates change with AD progression and for early disease diagnosis. This approach is also broadly applicable to other protein aggregates and other diseases. REFERENCES

[0315] 1. Hampel, H. etal. TheAmyloid-p Pathway in Alzheimer’s Disease. Molecular Psychiatry vol. 26 5481-5503 Preprint at https: / / doi.org / 10.1038 / s41380-021-01249-0 (2021).

[0316] 2. Aarsland, D. et al. Parkinson disease-associated cognitive impairment. Nat Rev Dis Primers 7, (2021).

[0317] 3. Porsteinsson, A. P., Isaacson, R. S., Knox, S., Sabbagh, M. N. & Rubino, I. Diagnosis of Early Alzheimer’s Disease: Clinical Practice in 2021. Journal of Prevention of Alzheimer’s Disease vol. 8 371-386 Preprint at https: / / doi.org / 10.14283 / jpad.2021.23 (2021).

[0318] 4. Angiulli, F. et al. Blood-based biomarkers of neuroinflammation in Alzheimer’s disease: A central role for periphery? Diagnostics vol. 11 Preprint at https: / / doi.org / 10.3390 / diagnostics11091525 (2021).

[0319] 5. Zimmermann, M. & Brockmann, K. Blood and Cerebrospinal Fluid Biomarkers of Inflammation in Parkinson’s Disease. Journal of Parkinson’s Disease vol. 12 S183-S200 Preprint at https: / / doi.org / 10.3233 / JPD-223277 (2022).

[0320] 6. King, E. et al. Inflammation in Mild Cognitive Impairment due to Parkinson’s disease, Lewy Body disease and Alzheimer’s disease.

[0321] 7. Chatterjee, P. et al. Plasma glial fibrillary acidic protein is elevated in cognitively normal older adults at risk of Alzheimer’s disease. Transl Psychiatry 11 , (2021).

[0322] 8. Williams-Gray, C. H. et al. Serum immune markers and disease progression in an incident Parkinson’s disease cohort (ICICLE-PD). Movement Disorders 31, 995-1003 (2016).

[0323] 9. Hakobyan, S. et al. Complement biomarkers as predictors of disease progression in Alzheimer’s disease. Journal of Alzheimer’s Disease 54, 707-716 (2016).

[0324] 10. Julian, A. et al. There is no correlation between peripheral inflammation and cognitive status at diagnosis in Alzheimer’s disease. Aging Clin Exp Res 27, 589-594 (2015).

[0325] 11. Janelidze, S. et al. Plasma p-amyloid in Alzheimer’s disease and vascular disease. Sci Rep 6, (2016).

[0326] 12. Chong, J. R. et al. Blood-based high sensitivity measurements of beta-amyloid and phosphorylated tau as biomarkers of Alzheimer’s disease: a focused review on recent advances. Journal of neurology, neurosurgery, and psychiatry vol. 92 1231-1241 Preprint at https: / / doi.org / 10.1136 / jnnp-2021-327370 (2021).

[0327] 13. Rust, M. J., Bates, M. & Zhuang, X. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat Methods 3, 793-795 (2006).

[0328] 14. Lobanova, E. et al. Imaging protein aggregates in the serum and cerebrospinal fluid in Parkinson’s disease. Brain 145, 632-643 (2022). 15. De, S. et al. Soluble aggregates present in cerebrospinal fluid change in size and mechanism of toxicity during Alzheimer’s disease progression. Acta Neuropathol Commun 7, 120 (2019).

[0329] 16. Kouli, A., Horne, C. B. & Williams-Gray, C. H. Toll-like receptors and their therapeutic potential in Parkinson’s disease and a-synucleinopathies. Brain, Behavior, and Immunity vol. 81 41-51 Preprint at https: / / doi.Org / 10.1016 / j.bbi.2019.06.042 (2019).

[0330] 17. Tahara, K. et al. Role of toll-like receptor signalling in Ap uptake and clearance. (2006).

[0331] 18. Meng, J. X. et al. Hyperphosphorylated tau self-assembles into amorphous aggregates eliciting TLR4-dependent responses. Nat Commun 13, (2022).

[0332] 19. Latz, E., Xiao, T. S. & Stutz, A. Activation and regulation of the inflammasomes. Nature Reviews Immunology vol. 13 397-411 Preprint at https: / / doi.org / 10.1038 / nri3452 (2013).

[0333] 20. Gordon, R. et al. Inflammasome inhibition prevents a-synuclein pathology and dopaminergic neurodegeneration in mice. Sci Transl Med 10, (2018).

[0334] 21. Halle, A. et al. The NALP3 inflammasome is involved in the innate immune response to amyloid- . Nat Immunol 9, 857-865 (2008).

[0335] 22. Venegas, C. et al. Microglia-derived ASC specks cross-seed amyloid-p in Alzheimer’s disease. Nature 552, 355-361 (2017).

[0336] 23. Dempsey, C. et al. Inhibiting the NLRP3 inflammasome with MCC950 promotes non- phlogistic clearance of amyloid-p and cognitive function in APP / PS1 mice. Brain Behav Immun 61 , 306-316 (2017).

[0337] 24. Holbrook, J. A. et al. Neurodegenerative Disease and the NLRP3 Inflammasome. Frontiers in Pharmacology vol. 12 Preprint at https: / / doi.org / 10.3389 / fphar.2021.643254 (2021).

[0338] 25. Jewell, S., Herath, A. M. & Gordon, R. Inflammasome Activation in Parkinson’s Disease. Journal of Parkinson’s Disease vol. 12 S113-S128 Preprint at https : / / do i . org / 10.3233 / J PD-223338 (2022) .

[0339] 26. Je, G. et al. Endogenous alpha-synuclein protein analysis from human brain tissues using single-molecule pull-down assay. Anal Chem 89, 13044-13048 (2017).

[0340] 27. Sideris, D. I. et al. Soluble amyloid beta-containing aggregates are present throughout the brain at early stages of Alzheimer’s disease. Brain Commun 3, (2021).

[0341] 28. Emin, D. et al. Small soluble a-synuclein aggregates are the toxic species in Parkinson’s disease. Nat Commun 13, (2022).

[0342] 29. Zhang, Y. P. et al. Imaging protein aggregates in Parkinson’s disease serum using aptamer-assisted single-molecule pull-down. Chem Sci (2023).

[0343] 30. Hong, W. et al. Diffusible, highly bioactive oligomers represent a critical minority of soluble Ap in Alzheimer’s disease brain. Acta Neuropathol 136, 19-40 (2018). 31 . Zhang, Y. P. et al. A simplified imaging surface for versatile and high-throughput singlemolecule studies. Anal Chem (2023).

[0344] 32. Tsukakoshi, K., Abe, K., Sode, K. & Ikebukuro, K. Selection of DNA aptamers that recognize a-synuclein oligomers using a competitive screening method. Anal Chem 84, 5542- 5547 (2012).

[0345] 33. Fazekas, F. J., Shaw, T. R., Kim, S., Bogucki, R. A. & Veatch, S. L. A mean shift algorithm for drift correction in localization microscopy. Biophysical Reports 1 , 100008 (2021).

[0346] 34. Ovesny, M., Kfizek, P., Borkovec, J., Svindrych, Z. & Hagen, G. M. ThunderSTORM: a comprehensive Imaged plug-in for PALM and STORM data analysis and super-resolution imaging. Bioinformatics 30, 2389-2390 (2014).

[0347] 35. Legland, D., Arganda-Carreras, I. & Andrey, P. MorphoLibJ: integrated library and plugins for mathematical morphology with Imaged. Bioinformatics 32, 3532-3534 (2016).

[0348] 36. Heneka, M. T. et al. Neuroinflammation in Alzheimer’s disease. The Lancet Neurology vol. 14 388-405 Preprint at https: / / doi.org / 10.1016 / S1474-4422(15)70016-5 (2015).

[0349] 37. dewell, S., Herath, A. M. & Gordon, R. Inflammasome Activation in Parkinson’s Disease, dournal of Parkinson’s Disease vol. 12 S113-S128 Preprint at https : / / do i . org / 10.3233 / d PD-223338 (2022) .

[0350] 38. Mangan, M. S. d. et al. Targeting the NLRP3 inflammasome in inflammatory diseases.

[0351] Nature Reviews Drug Discovery vol. 17 588-606 Preprint at https: / / doi.Org / 10.1038 / nrd.2018.97 (2018).

[0352] 39. Heneka, M. T. et al. NLRP3 is activated in Alzheimer’s disease and contributes to pathology in APP / PS1 mice. Nature 493, 674-678 (2013).

[0353] 40. Coll, R. C. et al. A small-molecule inhibitor of the NLRP3 inflammasome for the treatment of inflammatory diseases. Nat Med 21 , 248-257 (2015).

[0354] 41. Gordon, R. et al. Inflammasome inhibition prevents a-synuclein pathology and dopaminergic neurodegeneration in mice. Sci Transl Med 10, (2018).

[0355] 42. Amo-Aparicio, d., Daly, d., Hojen, d. F. & Dinarello, C. A. Pharmacologic inhibition of NLRP3 reduces the levels of a-synuclein and protects dopaminergic neurons in a model of Parkinson’s disease, d Neuroinflammation 20, (2023).

[0356] 43. Bertheloot, D. et al. Nanobodies dismantle post-pyroptotic ASC specks and counteract inflammation in vivo . EMBO Mol Med 14, (2022).

[0357] 44. Narayan, P. et al. Single Molecule Characterization of the Interactions between Amyloid-p Peptides and the Membranes of Hippocampal Cells, d Am Chem Soc, 4, 1491-1498 (2013).

[0358] 45. Fish, K.N. Total internal reflection fluorescence (TIRF) microscopy. Curr Protoc Cytom. Chapter 12: Unit 2 (2009). 46. Moerner, W. E. et al. Methods of single-molecule fluorescence spectroscopy and microscopy. Rev Sci Instrum. 74(8) (2003).

[0359] 47. Ross, C. A. & Poirier, M. A. Protein aggregation and neurodegenerative disease. Nat Med 10 Suppl, S10-17 (2004).

[0360] 48. Tsoi, P. S., Quan, M. D., Ferreon, J. C. & Ferreon, A. C. M. Aggregation of Disordered Proteins Associated with Neurodegeneration. Int J Mol Sci 24, 3380 (2023).

[0361] 49. Armstrong, R. A. Plaques and tangles and the pathogenesis of Alzheimer’s disease. Folia Neuropathol 44, 1-11 (2006).

[0362] 50. Spillantini, M. G. et al. Alpha-synuclein in Lewy bodies. Nature 388, 839-840 (1997).

[0363] 51. De, S. et al. Different soluble aggregates of Ap42 can give rise to cellular toxicity through different mechanisms. Nature Communications 2019 10:1 10, 1-11 (2019).

[0364] 52. Alquezar, C., Arya, S. & Kao, A. W. Tau Post-translational Modifications: Dynamic Transformers of Tau Function, Degradation, and Aggregation. Front Neurol 11 , 595532 (2021).

[0365] 53. Zhang, J., Li, X. & Li, J.-D. The Roles of Post-translational Modifications on a-Synuclein in the Pathogenesis of Parkinson’s Diseases. Front Neurosci 13, 381 (2019).

[0366] 54. Tarutani, A. et al. Distinct tau folds initiate templated seeding and alter the post- translational modification profile. Brain 146, 4988-4999 (2023).

[0367] 55. Manzanza, N. de O., Sedlackova, L. & Kalaria, R. N. Alpha-Synuclein Post- translational Modifications: Implications for Pathogenesis of Lewy Body Disorders. Front Aging Neurosci 13, 690293 (2021).

[0368] 56. Kummer, M. P. & Heneka, M. T. Truncated and modified amyloid-beta species. Alzheimers Res Ther 6, 28 (2014).

[0369] 57. Standke, H. G. & Kraus, A. Seed amplification and RT-QuIC assays to investigate protein seed structures and strains. Cell Tissue Res 392, 323-335 (2023).

[0370] 58. Deng, H.-X., Bigio, E. H. & Siddique, T. Detection of protein aggregation in neurodegenerative diseases. Methods Mol Biol 793, 259-272 (2011).

[0371] 59. Pan, T. et al. An Aggregation-Specific Enzyme-Linked Immunosorbent Assay: Detection of Conformational Differences between Recombinant PrP Protein Dimers and PrPSc Aggregates. J Virol 79, 12355-12364 (2005).

[0372] 60. Lincon, A., Das, S. & DasGupta, S. Capturing protein denaturation using electrical impedance technique. Journal of Molecular Liquids 360, 119301 (2022).

[0373] 61. Novo, M., Freire, S. & Al-Soufi, W. Critical aggregation concentration for the formation of early Amyloid-p (1-42) oligomers. Sci Rep 8, 1783 (2018). 62. Pedersen, M. E., Ostergaard, J. & Jensen, H. Flow-Induced Dispersion Analysis (FIDA) for Protein Quantification and Characterization. Methods Mol Biol 1972, 109-123 (2019).

[0374] 63. Xia, W. et al. A Specific ELISA for Measuring Amyloid p-Protein Oligomers in Human Plasma and the Brains of Alzheimer Patients. Arch Neurol 66, 190-199 (2009).

[0375] 64. Bruggink, K. A. et al. Amyloid-p oligomer detection by ELISA in cerebrospinal fluid and brain tissue. Anal Biochem 433, 112-120 (2013).

[0376] 65. Herrmann, Y. et al. sFIDA automation yields sub-femtomolar limit of detection for A aggregates in body fluids. Clin Biochem 50, 244-247 (2017).

[0377] 66. Bldmeke, L. et al. Quantitative detection of a-Synuclein and Tau oligomers and other aggregates by digital single particle counting. NPJ Parkinsons Dis 8, 68 (2022).

[0378] 67. Herrmann, Y. et al. Nanoparticle standards for immuno-based quantitation of a- synuclein oligomers in diagnostics of Parkinson’s disease and other synucleinopathies. Clin Chim Acta 466, 152-159 (2017).

[0379] 68. Hulsemann, M. et al. Biofunctionalized Silica Nanoparticles: Standards in Amyloid-p Oligomer-Based Diagnosis of Alzheimer’s Disease. J Alzheimers Dis 54, 79-88 (2016).

[0380] 69. Kuhbach, K. et al. Application of an Amyloid Beta Oligomer Standard in the sFIDA Assay. Front Neurosci 10, 8 (2016).

[0381] 70. Willbold, D. et al. Blood-based quantification of Ap oligomers indicates impaired clearance from brain in ApoE E4 positive subjects. Alzheimer’s & Dementia 19, e082922 (2023).

[0382] 71. Wilson, D. H. et al. The Simoa HD-1 Analyzer: A Novel Fully Automated Digital Immunoassay Analyzer with Single-Molecule Sensitivity and Multiplexing. Journal of Laboratory Automation 21 , 533-547 (2016).

[0383] 72. Bbken, D. et al. Single-Molecule Characterization and Super-Resolution Imaging of Alzheimer’s Disease-Relevant Tau Aggregates in Human Samples. Angew Chem Int Ed Engl e202317756 (2024) doi:10.1002 / anie.202317756.

[0384] 73. Islam, T. et al. Novel ultrasensitive immunoassay for the selective quantification of tau oligomers and related soluble aggregates. Alzheimer’s & Dementia 20, 2894-2905 (2024).

[0385] 74. Li, W. et al. Aggregation promoting C-terminal truncation of a-synuclein is a normal cellular process and is enhanced by the familial Parkinson’s disease-linked mutations. Proceedings of the National Academy of Sciences 102, 2162-2167 (2005).

[0386] 75. Kulenkampff, K. et al. An antibody scanning method for the detection of a-synuclein oligomers in the serum of Parkinson’s disease patients. Chem Sci 13, 13815-13828. 76. Goedert, M., Jakes, R. & Vanmechelen, E. Monoclonal antibody AT8 recognises tau protein phosphorylated at both serine 202 and threonine 205. Neurosci Lett 189, 167-169 (1995).

[0387] 77. Nobuhara, C. K. et al. Tau Antibody Targeting Pathological Species Blocks Neuronal Uptake and Interneuron Propagation of Tau in Vitro. Am J Pathol 187, 1399-1412 (2017).

[0388] 78. Baghallab, I., Reyes-Ruiz, J. M., Abulnaja, K., Huwait, E. & Glabe, C. Epitomic Characterization of the Specificity of the Anti-Amyloid Ap Monoclonal Antibodies 6E10 and 4G8. Journal of Alzheimer’s Disease 66, 1235 (2018).

[0389] 79. Reed, G. F., Lynn, F. & Meade, B. D. Use of Coefficient of Variation in Assessing Variability of Quantitative Assays. Clin Diagn Lab Immunol 9, 1235-1239 (2002).

[0390] 80. Shim, K. H., Kang, M. J., Youn, Y. C., An, S. S. A. & Kim, S. Alpha-synuclein: a pathological factor with Ap and tau and biomarker in Alzheimer’s disease. Alzheimers Res Ther 14, 201 (2022).

[0391] 81 . Fertan, E. et al. Cerebral organoids with chromosome 21 trisomy secrete Alzheimer’s disease-related soluble aggregates detectable by single-molecule-fluorescence and superresolution microscopy. Mol Psychiatry (2023) doi:10.1038 / s41380-023-02333-3.

[0392] 82. Dinh, T. L., Ngan, K. C., Shoemaker, C. B. & Walt, D. R. Using Antigen-antibody Binding Kinetic Parameters to Understand Single-Molecule Array Immunoassay Performance. Anal Chem 88, 11335-11339 (2016).

[0393] 83. Cremades, N. et al. Direct Observation of the Interconversion of Normal and Toxic Forms of a-Synuclein. Cell 149, 1048-1059 (2012).

[0394] 84. Corbett, G. T. et al. PrP is a central player in toxicity mediated by soluble aggregates of neurodegeneration-causing proteins. Acta Neuropathol 139, 503-526 (2020).

[0395] 85. McEwan, W. A. et al. Cytosolic Fc receptor TRIM21 inhibits seeded tau aggregation. Proc Natl Acad Sci U S A 114, 574-579 (2017).

[0396] 86. Bernardim, B. et al. Stoichiometric and irreversible cysteine-selective protein modification using carbonylacrylic reagents. Nat Commun 7, 13128 (2016).

[0397] 87. Pangborn, A. B., Giardello, M. A., Grubbs, R. H., Rosen, R. K. & Timmers, F. J. Safe and Convenient Procedure for Solvent Purification. Organometallics 15, 1518-1520 (1996).

[0398] 88. Jakubec, P. & Berkes, D. Crystallisation-induced asymmetric transformation (CIAT) for the synthesis of dipeptides containing homophenylalanine. Tetrahedron: Asymmetry 21 , 2807-2815 (2010).

Claims

CLAIMS1. A method of analysing one or more protein complexes in a blood, serum or plasma sample comprising:(a) performing an assay to quantify the level of the one or more protein complexes present in the sample, wherein the assay comprises a capture binding agent and a detection binding agent which specifically bind to the protein complex, wherein the same binding agent is used as the capture binding agent and the detection binding agent; and(b) performing a method for assessing one or more morphological features of the protein complexes present in the sample, wherein the combination of information obtained from performing steps (a) and (b) is used to analyse the protein complex.

2. The method of claim 1 , wherein the assay of step (a) is a single molecule pulldown (SiM Pull) assay comprising:(i) contacting the blood, serum or plasma sample with a solid phase comprising the capture binding agent which specifically binds to the protein complex;(ii) contacting the protein complex bound by the capture binding agent with a detection binding agent; and(iii) detecting the presence of the detection binding agent using single-molecule microscopy.

3. The method of claim 2, wherein the single-molecule microscopy is performed using a total internal fluorescence (TIRF) microscope.

4. The method of claim 2 or claim 3, wherein the solid phase is a bead or glass substrate.

5. The method of any one of claims 2 to 4, wherein the solid phase is passivated.

6. The method of any one of claims 2 to 5, wherein the capture binding agent is indirectly attached to the solid phase, such as through biotin and streptavidin, neutravidin or avidin interactions.

7. The method of any one of claims 1 to 6, wherein the one or more morphological features are selected from: size of the protein complex and shape of the protein complex.

8. The method of any one of claims 1 to 7, wherein step (b) is performed using an imaging method.

9. The method of claim 8, wherein the imaging method is Single-Molecule Localization Microscopy (SMLM).

10. The method of claim 9, wherein the SMLM is selected from: direct stochastic optical reconstruction microscopy (dSTORM) and DNA-Point Accumulation for Imaging in Nanoscale Topography (DNA-PAINT).11 . The method of any one of claims 1 to 7, wherein step (b) is performed by measuring brightness distribution of the protein complexes present in the sample.

12. The method of any one of claims 1 to 11 , wherein two or more types of protein complexes are analysed.

13. The method of any one of claims 1 to 12, wherein the protein complex comprises more than one type of protein and a second detection binding agent is used to detect another protein present in the protein complex.

14. The method of any one of claims 1 to 13, wherein the detection binding agent is labelled with a fluorescent marker, such as a fluorescent dye, in particular a fluorescent dye with photophysical properties.

15. The method of any one of claims 1 to 14, wherein the capture binding agent and detection binding agent are antibodies.

16. The method of any one of claims 1 to 15, wherein the sample is serum or plasma.

17. The method of any one of claims 1 to 16, wherein the protein complexes analysed are protein aggregates that form during the development of neurodegenerative disease and / or a protein complex associated with inflammation.

18. The method of claim 17, wherein the protein aggregates that form during the development of neurodegenerative disease are brain-derived.

19. The method of any one of claims 1 to 18, wherein the protein complexes analysed are selected from: ASC (adapter protein apoptosis associated speck-like protein containing a CARD) specks, Amyloid-p aggregates, tau aggregates, a-synuclein aggregates, TAR DNA- binding protein 43 (TDP-43) aggregates or co-aggregates thereof.

20. The method of any one of claims 1 to 19, wherein the protein complex comprises a post-translationally modified protein or a truncated protein, such as phosphorylated a-syn, truncated a-syn, ubiquitinated a-syn, ubiquitinated tau, phosphorylated tau, pyroglutamate- modified Amyloid-p, truncated tau or truncated Amyloid-p.

21. The method of any one of claims 1 to 20, wherein the assay is performed to quantify a total level of the protein complex and a level of the protein complex which has been post- translationally modified or truncated, to produce a combined biomarker.

22. The method of claim 21 , wherein the combined biomarker is the ratio of the total level of the protein complex to the level of protein complex which has been post-translationally modified or truncated.

23. A method for determining whether a subject has, or is at risk of developing, a neurodegenerative disorder, the method comprising performing the method of any one of claims 1 to 22 to analyse one or more protein complexes in a blood, serum or plasma sample obtained from the subject.

24. The method of claim 23, wherein the neurodegenerative disorder is selected from: Parkinson’s disease (PD) and other parkinsonian disorders, Alzheimer’s disease (AD), frontotemporal dementia (FTD), tauopathies, traumatic brain injury.

25. The method of claim 24, wherein the PD is prodromal PD or early PD, or wherein the AD is mild cognitive impairment (MCI) or early AD.

26. The method of any one of claims 23 to 25, wherein the method comprises comparing the information obtained for the protein complexes to one or more controls.

27. A combined biomarker for use in the detection or prognosis of a neurodegenerative disorder, wherein the combined biomarker comprises two types of protein complexes present in a sample, wherein one of the protein complexes is ASC speck.

28. The combined biomarker of claim 27, wherein the other type of protein complex is a protein aggregate that forms during the development of neurodegenerative disease, such as protein complexes selected from Amyloid-p aggregates, tau aggregates, a-synuclein aggregates and TDP-43 aggregates.

29. The combined biomarker of claim 27 or claim 28, wherein the combined biomarker comprises the level of ASC speck and / or the level of one or more protein aggregates that form during the development of neurodegenerative disease.

30. The combined biomarker of any one of claims 27 to 29, wherein the combined biomarker comprises the sum of the level of two or more protein aggregates that form during the development of neurodegenerative disease.

31. The combined biomarker of any one of claims 27 to 30, wherein the combined biomarker is the sum of the level of ASC speck and the level of one or more protein aggregates that form during the development of neurodegenerative disease.

32. The combined biomarker of any one of claims 27 to 31 , wherein the combined biomarker comprises a score for one or more morphological features for the ASC speck and / or a score for one or more morphological features for the protein aggregates that form during the development of neurodegenerative disease.

33. The combined biomarker of claim 32, wherein the one or more morphological features is selected from: the size or shape of the protein complex.

34. The combined biomarker of any one of claims 27 to 33, wherein the combined biomarker is determined as a ratio.

35. The combined biomarker of any one of claims 27 to 34, wherein the combined biomarker comprises:(i) the ratio of the level of ASC speck to the level of Amyloid-p aggregates;(ii) the ratio of the level of ASC speck to the level of p-tau aggregates;(iii) the ratio of the sum of the level of ASC speck and the level of p-tau aggregates to the level of Amyloid-p aggregates;(iv) the ratio of the sum of the level of ASC speck and the level of a-synuclein aggregates to the level of Amyloid-p aggregates;(v) the ratio of the sum of the level of ASC speck and the score for one or more morphological features for the ASC speck to the level of Amyloid-p aggregates.

36. A combined biomarker for use in the detection or prognosis of a neurodegenerative disorder, wherein the combined biomarker comprises the sum of the proportion of two or more protein aggregates that form during the development of neurodegenerative disease.

37. The combined biomarker of claim 36, which comprises the sum of the ratio of the level of a-synuclein aggregates to the level of Amyloid-p aggregates, and the level of tau aggregates to the level of p-tau aggregates (i.e. (a-syn I Ap) + (tau I p-tau)).

38. The combined biomarker of claim 36 or claim 37, wherein the combined biomarker comprises:(i) the sum of the ratio of the level of a-synuclein aggregates to the level of Amyloid-p aggregates, and the level of tau aggregates to the level of p-tau aggregates (i.e. (a-syn I Ap) + (tau I p-tau));(ii) the ratio of the level of p-tau aggregates to the level of tau aggregates (i.e. p tau I tau);(iii) the ratio of the level of Amyloid-p aggregates to the sum of the level of Amyloid- P, a-synuclein and tau aggregates (i.e. Ap aggregates / (Ap + a-synuclein + tau aggregates));(iv) the ratio of the sum of the level of Amyloid-p and p-tau aggregates to the level of a-synuclein aggregates (i.e. (Ap + p-tau aggregates) I a-syn aggregates); or(v) the ratio of the sum of the level of Amyloid-p and p-tau aggregates to the sum of the level of a-synuclein aggregates and the ratio of the level of p-tau aggregates to the level of tau aggregates (i.e. (Ap + p-tau aggregates) I (a-syn aggregates + (p-tau aggregates / tau aggregates))).

39. A method of diagnosing a neurodegenerative disorder in a subject, comprising:(a) measuring the combined biomarker of any one of claims 27 to 38 in a sample obtained from the subject; and(b) using the measurement detected for the combined biomarker to determine if the subject has, or is at risk of developing, a neurodegenerative disorder.

40. The method of claim 39, wherein the combined biomarker is measured using the method of any one of claims 1 to 26.

41. The method of claim 39 or claim 40, wherein the sample is a body fluid sample, such as blood, serum, plasma, cerebrospinal fluid or saliva.

42. The method of any one of claims 39 to 41 , wherein the method comprises comparing the measurement to one or more controls.

Images