Peripheral blood mononuclear cell (PBMC) phenotypes as biomarkers in patients with Alzheimer's disease and / or mild cognitive impairment

Abstract

The present invention relates to a method for identifying whether a mammal suffers or may suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) using a phenotypic signature generated using a multivariate classification algorithm. Furthermore, the present invention relates to a method for stratifying a mammal into a healthy group or a group suffering or may suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) for treatment against said disease. Furthermore, the present invention relates to a method for monitoring the progression and / or treatment status of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal undergoing treatment for Alzheimer's disease (AD) and / or mild cognitive impairment (MCI). The present invention also relates to a marker panel comprising the identified PBMC markers for identifying whether a mammal suffers or may suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI). Furthermore, the present invention relates to the use of said marker panel and / or phenotypic signature for diagnostic and / or drug risk avoidance applications.

Description

[Technical field]

[0001] The present invention relates to a method for identifying whether a mammal suffers or may suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) using a phenotypic signature generated using a multivariate classification algorithm. Furthermore, the present invention relates to a method for stratifying a mammal into a healthy group or a group suffering or may suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) for treatment against said disease. Furthermore, the present invention relates to a method for monitoring the progression and / or treatment status of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal undergoing treatment for Alzheimer's disease (AD) and / or mild cognitive impairment (MCI). The present invention also relates to a marker panel comprising the identified PBMC markers for identifying whether a mammal suffers or may suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI). Furthermore, the present invention relates to the use of said marker panel and / or phenotypic signature for diagnostic and / or drug risk avoidance applications. [Background technology]

[0002] Alzheimer's disease and other dementias are serious pathological conditions that afflict many people worldwide. These diseases represent a societal challenge and currently have no cure. Diagnostic and clinical trial research for neurodegenerative diseases is hampered in part by a limited number of biomarker tests that can accurately diagnose these diseases, appropriately stratify and enrich cohorts of patients with target pathology, and reliably monitor disease progression, as well as a lack of affordable, non-invasive tests.

[0003] An increasing number of scientific studies have shown that inflammatory processes in the brain are involved in the development of chronic neurodegenerative diseases and that systemic / peripheral immune responses contribute to disease progression. In fact, the brain does not retain absolute immune privilege, and in addition to infiltrating peripheral immune cells, peripheral inflammatory mediators, such as cytokines, danger signals, or other signaling molecules, can reach the brain and central nervous system (CNS) (Non-Patent Document 1, Non-Patent Document 2, Non-Patent Document 3). For example, peripheral inflammatory events and infections, such as airway infections, bacterial, viral, and fungal infections, have been found to be associated with cognitive decline and progression of dementia and contribute significantly to dementia progression (Non-Patent Document 3, Non-Patent Document 4, Non-Patent Document 5). By adopting a biocomputing system-level approach based on constructed molecular networks and integrating preclinical, genetic, and bioinformatics data, it has been revealed that immune regulatory pathways strongly contribute to late-onset Alzheimer's disease (Non-Patent Document 6).

[0004] For inflammation-related diseases, especially Alzheimer's disease, CSF biomarkers have been used in clinical diagnosis with high accuracy. However, there is a strong need for less invasive, cost-effective, and reliable methods to diagnose early-stage Alzheimer's disease (Non-Patent Document 7, Non-Patent Document 8).

[0005] Recently, ultrasensitive immunoassays based on immunoprecipitation mass spectrometry have been developed and successfully used to measure plasma Aβ biomarkers and to distinguish between healthy controls and patients with AD or mild cognitive impairment (MCI) (Non-Patent Document 9). The authors were able to demonstrate that the brain Aβ burden can be accurately predicted by assessing the plasma APP / Aβ ratio and the Aβ40 / 42 ratio (Non-Patent Document 9, Non-Patent Document 10, Non-Patent Document 11).

[0006] Moreover, it has been recently shown that a highly accurate plasma Aβ42 / 40 ratio can accurately diagnose brain amyloidosis and can therefore be used to screen cognitively normal individuals to identify those at risk for AD dementia (Non-Patent Document 12, Non-Patent Document 13). Furthermore, plasma neurofilament light chain (NfL) has recently emerged as a reliable peripheral biomarker for neurodegenerative diseases (Non-Patent Document 10, Non-Patent Document 11), showing excellent diagnostic accuracy in distinguishing neurodegenerative patients from controls and starting from early stages of disease progression (Non-Patent Document 14). Indeed, Preische et al. were able to demonstrate that CSF and serum NfL levels correlated and were found to be elevated in the presymptomatic stage of familial Alzheimer's disease (Non-Patent Document 15). This suggests that the assessment of serum NfL may be able to predict disease progression and brain neurodegeneration at an early presymptomatic stage of familial AD. Taken together, these facts indicate that blood-based biomarker testing may be possible.

[0007] However, there is a very high need for methods that can be used to detect changes with confidence, a necessary step that involves the combination and analysis of multiple parameters, and this is supported by a recent expert review by Penner et al., who viewed biomarkers in blood as “shadows of events happening in the brain” and described them as “fingerprints in blood” that are worth following up to overcome limitations in biomarker development (Non-Patent Document 16).

[0008] In recent years, based on the hypothesis that the use of a panel of markers may be superior to a single candidate marker, a number of blood-based multivariate biomarker panels have been examined and proposed for the purposes of diagnosing, classifying, and prognosing AD (Non-Patent Document 8).

[0009] Overall, however, there is a lack of harmonization and reproducibility of findings (Non-Patent Document 8).Recently, a hypothesis-based plasma biomarker search was conducted analyzing 53 inflammatory proteins in healthy controls, individuals with MCI and AD, and a collection of analytes and markers of inflammation and complement dysregulation, showing significant group differences that could be reproduced in independent cohorts (Non-Patent Document 17).

[0010] Increasingly, drugs are used in a personalized way, to be able to more specifically exploit the individual mechanism of action of the drug in the patient and to avoid inflammatory responses. There is a need for suitable biomarkers that allow reliable statements about whether the drug under consideration is likely to be effective in this patient; whether a particular patient is likely to tolerate the drug under consideration; and how the drug can be optimally administered for this patient.

[0011] Overall, there is a lack of methods to diagnose Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a robust, reliable, fast, cheap and convenient manner, especially compared to healthy controls, while at the same time allowing for patient stratification, monitoring of subsequent treatment, and proactive exclusion of potentially inflammatory drugs through drug risk avoidance.

[0012] A major challenge of personalized medicine is the analysis of the diagnostic data obtained. For example, genetic data obtained from procedures such as next-generation sequencing require computationally complex data processing steps before the actual analysis of the data can take place. The subsequent analysis requires multidisciplinary collaboration between experts in medical sciences, clinical oncologists, biologists, and software engineers. Thus, the provision of personalized treatments entails a high degree of diagnostic effort.

[0013] Currently, the automated workflows disclosed in the prior art are optimized for homogenous or image-based readout in human cell lines or mouse cell models. These approaches cannot link intracellular processes at the single-cell level with extracellular signals of the innate immune response, which is essential for understanding and evaluating the mechanism of action of perturbation effects in predictive drug discovery and innate immune risk mitigation.

[0014] The prior art discloses many approaches for obtaining and evaluating quantitative data from microscopic imaging.

[0015] Loo et al. (Non-Patent Document 18) disclose a multivariate method for classifying untreated and treated human cancer cells based solely on phenotypic single cell measurements.

[0016] Bray et al. (Non-Patent Document 19) describe a phenotypic profiling method based on microscopic images of cellular components or organelles, identifies biologically relevant similarities and differences between samples based on these profiles, and assesses the effect of chemical, genetic, or disease insults on the phenotypic signatures.

[0017] Collinet et al. (Non-Patent Document 20) have demonstrated phenotypic profiling based on multiparametric image analysis using an automated confocal microscope to analyze the human genome in terms of endocytosis of transferrin and epidermal growth factor.

[0018] US Patent No. 5,399,633 discloses a method and system for displaying measurements of complete blood count ("CBC") parameters, the method including a step of displaying measurements of CBC parameters obtained from a plurality of samples of a first lot of a quality control composition, the displaying including displaying a marker corresponding to each measurement from the first lot on a plot including a two-dimensional coordinate system, the two-dimensional coordinate system including a first dimension corresponding to the time the measurement of the CBC parameter was obtained and a second dimension corresponding to the numerical value of the CBC parameter.

[0019] There are previous applications using PBMCs for specific diseases, for example, Patent Document 2 uses miRNAs present in PBMCs to diagnose lung cancer.

[0020] US Patent No. 5,999,633 utilizes the reduced levels of Hsp70 in PBMC samples to serve as a biomarker for diseases that show reduced levels of Hsp70, such as lysosomal storage diseases, neurodegenerative diseases, and muscular diseases. However, all of these methods use single feature marker analysis and not multivariate analysis.

[0021] There have also been some attempts to use blood to develop tests for the diagnosis of Alzheimer's disease and other dementias. In most cases, these attempts are based on a single marker (e.g., Aβ-42) or a combination of multiple proteins (Patent Document 4 or Patent Document 5 or Patent Document 6). These methods have limited sensitivity and specificity and lack the ability to be implemented in a scalable and economical manner. To date, bioassays combining lymphocytes and specific peptides have also been reported (Patent Document 7). [Prior art documents] [Patent documents]

[0022] [Patent Document 1] US 2013-0024130 [Patent Document 2] US 2018-0251853 (Biomarkers inPeripheral Blood Mononuclear Cells for Diagnosing or Detecting Lung Cancers)

Patent document 3

Patent document 4

Patent document 5

Patent document 6

Patent document 7

Non-licensed literature

[0023]

Non-licensed literature 1

Non-licensed Document 2

Non-licensed Document 4

Non-licensed Document 5

Non-licensed Document 6

Non-licensed Document 7

Non-licensed literature 9

Non-licensed literature 10

Non-licensed Document 11

Non-licensed Document 12

Non-licensed Document 13

Non-licensed Document 14

Non-licensed Document 15

Non-licensed Document 16

Non-licensed Document 17

Non-licensed Document 18

Non-licensed Document 19

[0024] However, methods based only on single-point readouts (e.g., blood-based biomarkers), such as microscopic phenotypic assessment, are prone to failure. Therefore, the inventors have developed a phenotypic signature / marker panel that allows to improve all the above-mentioned shortcomings in the diagnosis of Ad and MCI, patient stratification, disease progression monitoring, and compound risk avoidance in a reliable, convenient, reproducible, and cost-effective manner. The prior art does not disclose or suggest comparable assays and panels. [Means for solving the problem]

[0025] In a first aspect thereof, the present invention provides a phenotypic signature or marker panel for identifying whether a mammal suffers from or is likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or is healthy, comprising the phenotypic signature or marker panel disclosed herein: TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell;activCell_EstAbsBcell; TNFaRel_Monoc; activCell_EstAbsMonoc; TNFaRel_Tcell;IL1bRel_actCellsObj; Activationrate_percent; IL1bRel_Monoc; IL1b_pgml; TNFaRel_Bcell;TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell; IL1bRel_IL1bPosCell;TNFa_pos_percent; The above problem is solved by providing a phenotypic signature or marker panel comprising PBMC markers selected from the group consisting of IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte;IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef;TNFaPosCells_Monocyte; and TNFa_pgml.

[0026] Preferably, the present invention solves the above problems by providing a phenotypic signature or marker panel for identifying whether a mammal suffers from or is likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) or is healthy, comprising at least the PBMC markers consisting of IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte;IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef;TNFaPosCells_Monocyte; TNFa_pgml; optionally including TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell and activCell_EstAbsBcell as disclosed herein.

[0027] Markers of the phenotypic signature were determined based on analysis of a selected set of features derived from microscopy, and biochemical and protein markers of PBMCs, as described below. Phenotypic signatures were derived as described herein by identifying patients and / or patient groups suffering from Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) and healthy controls, i.e., AD, AD / MCI (MCI with AD, also referred to herein as MCI-AD or MCI and AD (MCI-to-AD), see, e.g., Varatharajah, Y., Ramanan, VK, Iyer, R. et al. Predicting Short-term MCI-to-AD Progression Using Imaging, CSF, Genetic Factors, Cognitive Resilience, and Demographics. Sci Rep 9, 2235 (2019). https: / / doi.org / 10.1038 / s41598-019-38793-3), MCI (i.e., MCI but not AD), and healthy.

[0028] In its second aspect, the present invention solves the above problem by providing a kit comprising a marker panel and / or a phenotypic signature according to the present invention together with auxiliary materials, preferably suitable buffers, dyes and / or labels, for carrying out the methods according to the present invention as described herein.

[0029] In a third aspect thereof, there is provided an in vitro method for identifying whether a mammal is suffering from or likely to suffer from Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI), or is healthy, said method comprising: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from a) no stimulation (i.e., untreated, untreated), b) Toll-like receptor activation (e.g., lipopolysaccharide, LPS, etc.), c) NOD-like receptor activation (i.e., ATP as an NLRP3 activator), and d) a combination of Toll-like receptor activation (e.g., LPS and NOD-like receptor activation (e.g., ATP)); iii) detecting in the PBMCs at least one phenotypic signature of said PBMCs against said agent, said ... Analyzing phenotypic markers selected from the group consisting of the markers IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; and iv) identifying said mammal as suffering or likely to suffer from Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI), or as healthy, based on the analysis of phenotypic markers, using a multivariate classification algorithm; The above problem is solved by providing the method.

[0030] In one preferred embodiment of the method according to the invention, the method is for identifying whether a mammal suffers from or may suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), wherein the stimulation in step ii) is LPS stimulation, and the phenotypic markers analyzed in step iii) include at least IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; optionally TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell, and activCell_EstAbsBcell, and based on analysis of the phenotypic markers and using a multivariate classification algorithm, identify whether the mammal is suffering from or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI).

[0031] Markers of the phenotypic signature are determined based on analysis of intracellular phenotypic markers, including appropriate staining of microscopically assessable intracellular and / or extracellular structures of PBMCs, followed by microscopic evaluation of said stained intracellular and / or extracellular structures, and analysis of phenotypic markers in the culture medium, including appropriate identification of soluble marker proteins and / or their biological activity.

[0032] In its fourth aspect, the present invention solves the above problem by providing an in vitro method for stratifying a mammal into a healthy group or a group suffering or likely to suffer from Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) or a healthy group, said method comprising carrying out a method according to the invention and further comprising v) stratifying the mammal into a healthy group or a group suffering or likely to suffer from Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) or a healthy group based on the analysis of step iv).

[0033] In a fifth aspect thereof, the present invention solves the above-mentioned problems by providing an in vitro method for monitoring the progression of Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) and / or the status of a treatment thereof in a mammal treated for Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI), said method comprising the steps of: i) providing a first biological sample obtained at a first time point from said mammal to be treated, said first biological sample comprising peripheral blood mononuclear cells (PBMCs); ii) providing a second biological sample obtained at a second time point from said mammal to be treated, said second biological sample comprising peripheral blood mononuclear cells (PBMCs); iii) performing the method according to the invention on both the first and second samples; and iv) monitoring the progression of Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI) and / or the status of a treatment thereof in said mammal based on the analyzed changes in the first and second phenotypic signatures.

[0034] In its sixth aspect, the present invention provides an in vitro method for identifying a compound for the prevention and / or treatment of Alzheimer's disease (AD) or Mild Cognitive Impairment (MCI) in a mammal, said method comprising the steps of: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from: a) no stimulation (i.e. untreated, untreated); b) Toll-like receptor activation (lipopolysaccharide, LPS, etc.); c) NOD-like receptor activation (i.e. ATP as an NLRP3 activator); and d) a combination of Toll-like receptor activation (e.g. LPS and NOD-like receptor activation (e.g. ATP)); iii) suitably contacting said sample with at least one candidate compound; iv) detecting in said PBMCs a candidate compound, as disclosed herein, TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell; activCell_EstAbsBcell;TNFaRel_Monoc; activCell_EstAbsMonoc; TNFaRel_Tcell; IL1bRel_actCellsObj;Activationrate_percent; IL1bRel_Monoc; IL1b_pgml; TNFaRel_Bcell;TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell; IL1bRel_IL1bPosCell;TNFa_pos_percent; IL1bRel_Tcell; activCell_EstAbsTcell; TNFaCells_IL1bCells; activCell_EstAbsUndef;TNFaPosCells_Monocyte;and TNFa_pgml, and v) using a multivariate classification algorithm to identify the compound based on changes in markers identified after comparing the analysis of step iv) with an analysis of the phenotypic markers in a sample in the absence of the candidate compound and / or with an analysis of the phenotypic markers in a control sample;

[0035] In one preferred embodiment of the method according to the invention, said method is for identifying compounds for the prevention and / or (preferably or) treatment of Alzheimer's disease (AD) and / or Mild Cognitive Impairment (MCI), wherein the stimulation in step ii) is an LPS stimulation and the phenotypic markers analyzed in step iii) comprise at least IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent;IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef;TNFaPosCells_Monocyte; and TNFa_pgml; optionally TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell, and activCell_EstAbsBcell, and based on analysis of the phenotypic markers and using a multivariate classification algorithm, the mammal is identified as suffering from or likely to suffer from Alzheimer's disease (AD) and / or (preferably, or) Mild Cognitive Impairment (MCI).

[0036] In a seventh aspect thereof, the above-mentioned object is solved by providing a use of a marker panel according to the invention for stratifying a mammal according to the invention into a healthy group or a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), for monitoring the progression and / or treatment status of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal according to the invention, or for identifying compounds for the prevention and / or treatment of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal according to the invention, for identifying whether a mammal suffers or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) or is healthy.

[0037] The technical features of the present invention are further described below. These features are listed with specific embodiments, but it should be understood that they can be combined in any manner and in any number to create additional embodiments. The various described examples and preferred embodiments should not be interpreted as limiting the present invention to only the explicitly described embodiments. The present description should be understood to support and encompass embodiments that combine two or more of the explicitly described embodiments, or combine one or more of the explicitly described embodiments with any number of the disclosed elements and / or preferred elements. Furthermore, any permutation and combination of all elements described in this application should be considered to be disclosed by the description of this application, unless the context indicates otherwise.

[0038] The present invention provides phenotypic signatures and / or marker panels and in vitro methods for diagnostic applications for identifying patients suffering from Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI), for stratifying patients, for monitoring the progression of Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) during patient treatment, and for screening for effective therapies and / or drugs. As referred to herein, "Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI)" includes the phenotypic diseases AD, AD / MCI (MCI with AD, also referred to herein as MCI-AD or MCI and AD (MCI-to-AD), see, e.g., Varatharajah, Y.,Ramanan, VK, Iyer, R. et al. Predicting Short-term MCI-to-AD Progression Using Imaging, CSF, Genetic Factors, Cognitive Resilience, and Demographics.Sci Rep 9, 2235 (2019). https: / / doi.org / 10.1038 / s41598-019-38793-3), and MCI (i.e., MCI that is not AD).

[0039] PBMCs are components of the immune system and consist of lymphocytes (T cells, B cells, NK cells) and monocytes. PBMCs are available in large quantities from venous blood collection and are routinely produced in many analytical laboratories. Therefore, we consider PBMCs to be a valuable and reliable source of cell-based biomarkers. In our invention, we use PBMC cell-based functional assays in combination with predictive analytical methods using machine learning and artificial intelligence (AI) algorithms to deep profile PBMCs obtained by individuals towards a unique database containing statistically relevant samples of a population.

[0040] The phenotypic markers obtained according to the present invention represent numerical vectors reflecting the unique phenotypic signature of PBMCs and provide a more robust representation of the mammal's status with respect to Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI), especially versus the healthy state, since the measurement error is distributed over multiple readouts, compared to the above-mentioned conventional assays based on a single-point readout.

[0041] Phenotypic markers were obtained by stimulating PBMCs in biological samples obtained from patients suffering from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) or from healthy individuals. The PBMCs were contacted with an inducer or inhibitor of innate immune response, which was stimulated and isolated from at least a portion of the supernatant of the biological sample, and at least a portion of the supernatant was collected and analyzed for factors that induce or inhibit innate immunity released from the cells. Furthermore, the PBMCs were appropriately stained for microscopically assessable intracellular and / or extracellular structures, and then analyzed by microscopic evaluation of the stained intracellular and / or extracellular structures.

[0042] Preferably, in the method according to the invention, the agent is an agent selected from the group of i) no stimulation (untreated or untreated), ii) Toll-like receptor activation (such as lipopolysaccharide, LPS), iii) NOD-like receptor activation (e.g. ATP, LeuLeuOMe, or nigericin, which are all NLRP3 inflammasome activators) and iv) a combination of Toll-like receptor activation (e.g. LPS and NOD-like receptor activation (e.g. ATP, LeuLeuOMe or nigericin)).

[0043] After analyzing the factors released in the supernatant and said microscopic evaluation using a multivariate classification algorithm, as described herein, a phenotypic signature was obtained that identifies patients and / or patient groups with Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) and healthy controls, i.e. AD, AD / MCI (MCI with AD, also referred to herein as MCI-AD or MCI and AD (MCI-to-AD)), MCI (i.e. MCI without AD), and healthy controls. The present invention is based on a set of cell-based markers and / or biochemical and / or genetic markers combined with functional assays investigating the immune system as described herein. The resulting 25, preferably 6 features selected from the assay readouts constitute the "phenotypic signature" or "marker panel" of an individual's PBMC according to the present invention.

[0044] As used herein, the term "phenotypic signature" covers a multivariate data set of PBMCs of interest in the form of a numerical vector obtained by the method according to the invention. The phenotypic signature is composed of at least data from the analysis of factors released into the supernatant (e.g., TNF-α, IL-1-β) and data from microscopic evaluation (e.g., chromatin, PYCARD, IL1β, TNF-α, and CD marker staining).

[0045] A machine learning multivariate classification algorithm processes the dataset obtained by the method according to the invention and generates therefrom a numerical vector representing a phenotypic signature indicative of the innate immune response of cells from said PBMCs or a control sample from said mammal. A baseline signature of such a control sample is constituted by a data collection of control cells from a sample obtained from a healthy donor. Once all data is collected, the data points of the vector undergo a statistical quality check and the most robust data points are used for the final analysis. All numerical vectors of analyzed cells are classified using a machine learning multivariate classification algorithm to finally assign the samples to innate immune response groups (i.e. healthy or diseased). Preferably, the identification of phenotypic markers in cells comprises generating a cluster map providing a clear p-value signature of pairwise comparisons and evaluating said comparisons for statistical differences, preferably using a p-value <0.05 or <0.01 using Kruskal-Wallis and / or Games-Howell post-hoc tests.

[0046] As used herein, the term "multivariate classification algorithm" refers to a process of identifying similarities between a multivariate dataset representing the phenotypic signature of said cells obtained by the method according to the invention and a multivariate dataset representing a phenotypic signature from a control cell, a phenotypic signature from untreated cells with a known innate immune response, or a phenotypic signature from a reference cell with a known innate immune response included in a database. Conventional multivariate classification algorithms are available to the skilled artisan and can be adapted for the purposes of the present invention. For example, Nicholson et al. have revealed a machine learning multivariate pattern analysis for classifying posttraumatic stress disorder. Apart from that, further documents are disclosed in the prior art (Nicholson, Densmore, Kinnon, Cambridge University Press; Volume 49, Issue 12; September 2019, pp. 2049-2059: Machine learning multivariate pattern analysis predicts classification of posttraumatic stress disorder and its dissociative subtype: a multimodal neuroimaging approach).

[0047] As mentioned above, the present invention further provides an in vitro method for identifying whether a mammal suffers from or is likely to suffer from Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI), or is healthy, said method comprising: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from a) no stimulation, b) Toll-like receptor activation, c) NOD-like receptor activation, and d) a combination of Toll-like receptor activation and NOD-like receptor activation; iii) detecting in the PBMCs at least one phenotypic signature of said PBMCs against said agent, said ... Analysing a phenotypic marker selected from the group consisting of the markers IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; and iv) identifying said mammal as suffering or likely to suffer from Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI), or as healthy, based on the analysis of phenotypic markers, using a multivariate classification algorithm; The above problem is solved by providing the method.

[0048] Still further preferably, the present invention relates to a method for identifying whether a mammal suffers from or is likely to suffer from Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI), said method comprising: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from: a) no stimulation, b) LPS; iii) analyzing in the PBMCs phenotypic markers selected from the group comprising the markers IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte;IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef;TNFaPosCells_Monocyte; and TNFa_pgml; optionally including TNFaRel_undefC; IL1bRel_TNFaPosCell;IL1bRel_Bcell, and activCell_EstAbsBcellas as disclosed herein to generate at least one phenotypic signature of the PBMCs for the agent; and iv) identifying said mammal as suffering or likely to suffer from Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI), or as healthy, based on the analysis of phenotypic markers, using a multivariate classification algorithm; The above problem is solved by providing the method.

[0049] Markers of the phenotypic signature were determined based on analysis of a selected set of features derived from microscopy and analysis of biochemical and protein markers of PBMCs, as described below. Identification of soluble marker proteins for analysis may include obtaining cytokine release data such as TNFα and IL1β, and microscopic evaluation may include chromatin, PYCARD, IL1β, TNF-α, and CD marker staining.

[0050] Thus, the present invention provides a method to detect whether an individual PBMC phenotype deviates from a healthy population, whether an individual PBMC phenotype is associated with a given disease phenotype, or whether an individual PBMC phenotype of a given disease is responsive to a certain treatment. The present invention is based on a cell-based marker set and / or a biochemical marker and / or a genetic marker. The resulting features selected from the assay readout constitute the phenotypic signature of the individual PBMC. The obtained associated individual PBMC phenotypic signature is analyzed, i.e. compared against a database, and associated with a particular risk / disease / treatment group (AD and / or MCI) versus health, as disclosed herein.

[0051] The identifying step preferably comprises assigning said phenotypic signature to a risk / disease / treatment group (AD and / or MCI) as disclosed herein, and comprises the use of a multivariate classification algorithm, wherein a change in said phenotypic signature of said mammal when compared to said phenotypic signature in a healthy mammal identifies a mammal suffering from or at risk of suffering from an AD and / or MCI related disease.

[0052] As used herein, the term "biological sample" preferably refers to a liquid biological sample, i.e. any mammalian body fluid that contains PBMCs to be analyzed from said mammal, which allows the method according to the invention to be carried out. For example, such a liquid biological sample may be selected from, but is not limited to, blood, organ or cell-type blood samples. Liquid biological samples also include PBMCs in an appropriate (nutrient) medium.

[0053] The term "in vitro" as used herein refers to methods carried out outside a mammal and in a controlled artificial environment. For example, cells (PBMCs) for "in vitro" methods may be taken from said mammal or may be obtained from (immortalized) mammalian cell lines. These methods may be carried out, for example, but are not limited to, in a test tube or a petri dish.

[0054] The term "supernatant" as used herein refers to the nutrient medium necessary for the culture and preservation of said PBMCs of interest. The culture supernatant is in direct contact with said cells and therefore contains factors released by said cells that are essential for obtaining the phenotypic signature of said cells.

[0055] In the present invention, phenotypic markers identified in PBMCs as being associated with a phenotypic signature in the context of AD and / or MCI versus health are selected from the group comprising the following analyses:

[0056] [Table 1] JPEG2024529522000002.jpg128137

[0057] Measurements and determinations were performed using assays described in the art and in, for example, PCT / EP2021 / 052306, which is incorporated herein by reference.

[0058] Preferably, the subsets identified in PBMCs in the present invention as associated with a phenotypic signature in relation to AD and / or MCI versus health, selected from the group comprising the following analyses:

[0059] [Table 2]

[0060] As used herein, the term "Toll-like receptors (TLRs)" refers to structures of innate immunity that belong to the group of pattern recognition receptors (PRRs). TLRs are transmembrane proteins that are used to recognize PAMPs (pathogen-associated molecular patterns). PAMPs are highly conserved pathogen-specific patterns that are not found in mammals. Such PAMPs can be, for example, proteins, lipids, and carbohydrates, such as lipopolysaccharides or flagellins, present on the surface of microorganisms. In contrast, so-called "damage-associated molecular patterns (DAMPs)" are host molecules, such as heat shock proteins (HSPs) or interleukin-1 (IL-1). These are endogenous, usually intracellular molecules that enter the extracellular space when cells are damaged. Binding to PRRs triggers an inflammatory response.

[0061] The term "NOD-like receptors (NLRs)" as used herein is related to the structure of innate immunity and also belongs to a further group of PRRs. In contrast to TLRs, NLRs exist as soluble proteins in the cytosol and are responsible for the recognition of intracellular pathogen-associated molecular patterns (PAMPs) introduced into cells via bacterial protein secretion or PAMPs taken up by immune cells via phagocytosis. In addition, danger-associated molecular patterns (DAMPs) were also used (Yang X, Lin G, Han Z, Chai J.Structural Biology of NOD-Like Receptors. Adv Exp Med Biol. 2019;1172:119-141.doi: 10.1007 / 978-981-13-9367-9_6. PMID: 31628654.).

[0062] In another preferred embodiment, the present invention solves the above problem by providing an in vitro method for stratifying a mammal into a healthy group, or into a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or into a healthy group, said method comprising carrying out the method according to the invention and further comprising a step v) of stratifying the mammal into a healthy group, or into a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or into a healthy group based on the analysis of step iv). Thus, if the phenotypic signature of said mammal deviates from the same phenotypic signature in a healthy mammal, said mammal is stratified into at least one group for treatment against said Alzheimer's disease (AD) and / or mild cognitive impairment (MCI).

[0063] In another preferred aspect thereof, the present invention provides an in vitro method for monitoring the progression of Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) and / or the status of a treatment therefor in a mammal undergoing treatment therefor, said method comprising the steps of: i) providing a first biological sample obtained at a first time point from said mammal to be treated, said first biological sample comprising peripheral blood mononuclear cells (PBMCs); ii) providing a second biological sample obtained at a second time point from said mammal to be treated, said second biological sample comprising peripheral blood mononuclear cells (PBMCs); The above problem is solved by providing a method comprising the steps of: iii) carrying out the method according to the invention on both the first sample and the second sample; and iv) monitoring the progression of Alzheimer's disease (AD) or mild cognitive impairment (MCI) and / or its treatment status in said mammal based on the analyzed changes in the first and second phenotypic signatures.

[0064] "Monitoring" as used in the context of the present invention means periodic provision of said biological sample from said mammal and implementation of the method according to the present invention to provide a reliable and complete indication of the progression of Alzheimer's disease (AD) or mild cognitive impairment (MCI) during treatment based on phenotypic signature analysis, as described herein. Suitable intervals can be determined by the skilled artisan.

[0065] During said treatment of the mammal, the effect of said treatment on the progression of Alzheimer's disease (AD) or mild cognitive impairment (MCI) (i.e., AD, MCI or AD / MCI) is monitored by changes in the phenotypic signature of the mammal.

[0066] In a preferred embodiment of the method according to the invention, the drug used in the treatment induces or inhibits the activity of a factor selected from the group consisting of Toll-like receptors (TLRs) and / or NOD-like receptors (NLRs).

[0067] In another preferred aspect thereof, the present invention relates to an in vitro method for identifying a compound for the prevention and / or treatment of Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) in a mammal, said method comprising the steps of: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from: i) no stimulation; ii) Toll-like receptor activation; iii) NOD-like receptor activation; and iv) a combination of Toll-like receptor activation and NOD-like receptor activation; iii) suitably contacting said sample with at least one candidate compound; iv) detecting in the medium of said sample and / or in said PBMCs at least one phenotypic signature of said PBMCs to said agent; activCell_EstAbsBcell; TNFaRel_Monoc; activCell_EstAbsMonoc;TNFaRel_Tcell; IL1bRel_actCellsObj; Activationrate_percent; TNFaRel_Bcell; TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell;IL1bRel_IL1bPosCell; TNFa_pos_percent; IL1bRel_Tcell; The above problem is solved by providing a method comprising the steps of: analyzing phenotypic markers selected from the group consisting of TNFaCells_IL1bCells;activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; and v) using a multivariate classification algorithm to identify the compound based on changes in markers identified after comparing the analysis of step iv) with an analysis of the phenotypic markers in a sample in the absence of the candidate compound and / or with an analysis of the phenotypic markers in a control sample.

[0068] Preferably, in the method according to the invention, the phenotypic markers analysed in step iv) comprise at least IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; optionally including TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell, and activCell_EstAbsBcell.

[0069] Preferably, the analysis of said phenotypic markers comprises appropriate staining of microscopically assessable intracellular and / or extracellular structures of the PBMCs followed by microscopic evaluation of said stained intracellular and / or extracellular structures, and preferably analysis of phenotypic markers in the culture medium including appropriate identification of soluble marker proteins, including obtaining cytokine release data such as TNFα and IL1b, and microscopic evaluation including chromatin, PYCARD, IL1β, TNF-α, and CD marker staining.

[0070] More preferably, identifying phenotypic markers in cells comprises generating a cluster map providing distinct p-value signatures of pairwise comparisons and assessing said comparisons for statistical differences, preferably using a p-value <0.05 or <0.01 using Kruskal-Wallis and / or Games-Howell post-hoc tests.

[0071] Further preferred, in an embodiment of the method according to the invention, said compound is selected from a drug, a small molecule, a polypeptide, an antibody or a fragment thereof, a peptide, a polynucleotide, an oligonucleotide, a nanoparticle, a carbohydrate, or a lipid.

[0072] The term "drug" as used in the context of the present invention encompasses any substance or pharmaceutical composition intended to treat or prevent a disease in a mammal. A drug may be a known substance or pharmaceutical composition, but also includes substances or pharmaceutical compositions that are not yet known. The term "pharmaceutical composition" encompasses all medical preparations, where a pure substance is formulated in combination with an excipient to form a pharmaceutical product, and thus the substance is not in a pure form, but is present as an administrable pharmaceutical composition. For example, to prevent inactivation of the compound, it may be necessary to coat the compound with a substance or to co-administer the compound, or the compound may be administered to the mammal in a suitable carrier, such as liposomes or a diluent. Pharmaceutically acceptable diluents include, for example, saline and aqueous buffer solutions to put the substance into the form of a tablet, capsule, injection, infusion, solution, ointment or syrup. In addition to a pure substance, an excipient or whole drug may also induce or suppress the innate immune response, whereby the effect of this induction or suppression can be identified in the method according to the present invention based on the phenotypic signature of the cells from the mammal.

[0073] As used herein, the term "small molecule" includes low molecular weight compounds having low molecular weight. It refers to a class of active substances having a molecular weight not exceeding about 800 g-mol-1.

[0074] In the context of the present invention, the term "protein" is used to refer to a polymer composed of amino acid monomers linked by peptide bonds. It refers to a molecular chain of amino acids, not a specific length of the product, which can be modified in vivo or in vitro, as appropriate, for example, by glycosylation, amidation, carboxylation, or phosphorylation. Amino acid chains less than about 100 amino acids in length are called "peptides." The terms "peptide" and "protein" are included in the definition of "polypeptide." A "peptide bond" is a covalent bond between two amino acids in which the α-amino group of one amino acid is attached to the α-carboxyl group of the other amino acid. All amino acid or polypeptide sequences are written from the amino terminus (N-terminus) to the carboxy terminus (C-terminus) unless otherwise specified.

[0075] The term "antibody" as used herein includes full-length antibodies and any fragment or derivative thereof in which the "antigen-binding portion" or "antigen-binding region" or a single chain thereof is retained, such as the binding domain of an antibody specific for the immunogenic active portion of the conjugate. Naturally occurring "antibodies" (immunoglobulins) are glycoproteins that contain at least two heavy (H) chains and two light (L) chains interconnected by disulfide bonds. Each heavy chain is composed of a heavy chain variable region (abbreviated herein as VH) and a heavy chain constant region. Heavy chains are classified as mu, delta, gamma, alpha, or epsilon, which define the antibody isotype as IgM, IgD, IgG, IgA, and IgE, respectively. The heavy chain constant region is composed of three domains: CH1, CH2, and CH3. Each light chain is composed of a light chain variable region (abbreviated herein as VL) and a light chain constant region. The light chain constant region is composed of one domain, CL. The VH and VL regions can be further subdivided into hypervariable regions, called complementarity determining regions (CDRs), interspersed with more conserved regions, called framework regions (FRs). Each VH and VL is composed of three CDRs and four FRs, arranged from the amino terminus to the carboxy terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The heavy and light chains form two regions: the Fab (fragment, antigen binding) region, also called the variable (Fv) region, and the Fc (fragment, crystallizable) region. The variable regions (Fv) of the heavy and light chains contain the binding domains that interact with antigens. The constant (Fc) region of an antibody may mediate binding to host tissues or factors, including various cells of the immune system (e.g., effector cells) and the first component (Clq) of the classical complement system. As used herein, the term "Fc" includes native and mutein forms of polypeptides derived from the Fc region of an antibody. Also included are truncated forms of such polypeptides that contain the hinge region that promotes dimerization. Fusion proteins that contain an Fc portion (and oligomers formed therefrom) have the advantage that they are easier to purify by affinity chromatography compared to protein A or protein G columns. One suitable Fc polypeptide is derived from a human IgG1 antibody.

[0076] As used herein, the term "polynucleotide" includes linear biopolymers composed of 20 or more nucleotides with alternating nucleic acid bases. Nucleic acid polymers with a small number of nucleotides are called "oligonucleotides". Such polynucleotides can be, for example, DNA, cDNA, RNA, or synthetically produced DNA or RNA, or recombinantly produced chimeric nucleic acid molecules that contain one of these nucleic acids alone or in combination.

[0077] As used herein, the term "nanoparticle" defines a particle having a particle size of less than 0.1 μm (<100 nm), which can be naturally occurring or synthetically produced. Such nanoparticles can be, for example, but are not limited to, carbon-containing nanoparticles, metal and semi-metal oxides, semiconductors, metal sulfides, or polymers.

[0078] As used herein, the term "carbohydrate" encompasses all carbon compounds composed of carbon, hydrogen, and oxygen and may exist as monosaccharides or polysaccharides.

[0079] As used herein, the term "lipid" refers to the whole range of fats and fat-like substances. Lipids are chemically heterogeneous substances that are poorly soluble in water but highly soluble in non-polar solvents.

[0080] In another preferred embodiment thereof, the present invention solves the above problem by providing the use of a marker panel according to the invention for stratifying mammals according to the invention into a healthy group or a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) and a healthy group, for monitoring the progression and / or treatment status of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal according to the invention or for identifying compounds for the prevention and / or treatment of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal according to the invention, for identifying whether a mammal suffers or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) according to the invention.

[0081] More preferably, in one embodiment of the method according to the present invention, the proteome, transcriptome, lipidome and / or genome of said cells are evaluated in a step of analyzing both factors released into the supernatant and microscopic evaluation to obtain a combined phenotypic signature, and clinical data are included to complement said combined phenotypic signature.

[0082] As used herein, the term "proteome" refers to the totality of proteins expressed by a single additional cell obtained from said mammal, which is evaluated in parallel with the microscopic evaluation of said cells in said liquid biological sample from said mammal according to the method of the present invention. The expressed proteins may occur as intracellular, extracellular, or transmembrane proteins. Many methods for analyzing these proteins are known to those skilled in the art. For example, proteomes can be analyzed by Western blot or mass spectrometry, but are not limited to these methods. The amount of cellular protein may vary depending on the measurement time.

[0083] The term "transcriptome" as used herein refers to the totality of all genes transcribed in an additional single cell obtained from said mammal (i.e., the totality of all RNA molecules present in said additional single cell) that is evaluated in parallel with the microscopic evaluation of said cells in said liquid biological sample obtained from said mammal according to the method of the present invention. The transcribed RNA molecules are present in a cell. In the prior art, there are many different methods for the analysis of cellular RNA, such as, but not limited to, reverse transcriptase polymerase chain reaction (RT-PCR). The amount of RNA present in a cell may vary depending on the measurement time.

[0084] The term "lipidome" as used herein refers to the totality of all lipids in a single additional cell obtained from the mammal, which is evaluated in parallel with the microscopic evaluation of the cells in the liquid biological sample obtained from the mammal according to the present invention. There are many methods known to those skilled in the art for analyzing lipidome, including, but not limited to, mass spectrometry.

[0085] The term "genome" as used herein refers to the totality of genetic information of a further single cell obtained from said mammal, which is evaluated in parallel with the microscopic evaluation of said cells in said liquid biological sample obtained from said mammal according to the method of the present invention. The term "genetic information" includes all genetic material contained in the cell nucleus, mitochondria, and plastids. The prior art provides many methods, such as, but not limited to, fluorescent in situ hybridization (FISH) or polymerase chain reaction (PCR).

[0086] The term "peripheral blood mononuclear cells (PBMCs)" as used herein encompasses mononuclear cells of blood with round nuclei. After providing the liquid biological sample from the mammal, PBMCs from blood, for example, can be purified by centrifugation through hydrophilic colloid. Both dead and live cells can be used in the method according to the invention.

[0087] Further provided in one embodiment of the method according to the invention is that said mammal is selected from mouse, rat, monkey, camel, llama, rabbit, pig, donkey, cow, horse, sheep and preferably human.

[0088] In another preferred aspect thereof, the present invention solves the above problem by providing a kit comprising a marker panel and / or a phenotypic signature according to the present invention, preferably together with suitable carriers, such as solid carriers, buffers, dyes and / or labels, together with auxiliary agents for carrying out the methods according to the present invention as described herein.

[0089] Further preferred, in an embodiment of the method according to the invention, said separation comprises immobilizing said cells on a transparent, preferably chemically inert solid support material, preferably selected from a glass or plastic material.

[0090] The term "solid support" refers to any support that does not affect the cellular reaction and is insoluble and chemically inert under the reaction conditions. In addition, the support must allow the inflow and outflow of materials. Many supports suitable for cell immobilization are known.

[0091] The term "fixation" as used herein refers to the spatial fixation of cells in a liquid biological sample from a mammal to a solid support, for example, but not limited to, by the action of a chemical such as methanol, ethanol, or acetone.

[0092] In the method according to the invention, the microscopically assessable intracellular and / or extracellular structures are stained and the microscopic assessment is carried out optically. Those skilled in the art are familiar with the various methods that can be used for optical detection of cellular structures. For example, but not limited to, neutral dyes can be used to bind lipid structures and alkaline dyes can be used to stain acidic nucleic acids. For example, conventional immunofluorescence methods can be applied to evaluate intracellular and / or extracellular proteins such as PYCARD, TNF-α, CD14, CD19, CD3, IL-1-β, CASP1, NFkappaB, and NLRP3 using commercially available antibodies that specifically bind to the respective proteins. By adding the respective second fluorescent dye-labeled detection antibodies, the corresponding structures on the proteins can be detected by fluorescence microscopy according to the wavelength range of the respective fluorescent dyes used.

[0093] More preferably, in one embodiment of the method according to the invention, the released factors in the analyzed supernatant comprise cytokines, such as TNF-α, and IL-1-β, and optionally IL18, IL1α, IL1-RA, IL6, IL8, IL10, IL18, CXCL1 / GROalpha, CXCL6, MCP1 / CCL2, MCP4 / CCL13, MIP1α / CCL3, MIP1β / CCL4, ENRAGE, and / or VEGFA. The term "released factors" as used herein encompasses all signaling molecules secreted or released by living or dead cells into the surrounding supernatant, such as, but not limited to, hormones, neurotransmitters, nitric oxide, and especially cytokines. The term "cytokine" as used in the context of the present invention encompasses all proteins secreted by cells in the course of an immunological or inflammatory response. The skilled artisan knows many methods for analyzing released factors, in particular cytokines, such as, but not limited to, enzyme-linked immunosorbent spot (ELISpot) or enzyme-linked immunosorbent assay (ELISA).

[0094] In another preferred embodiment of the method according to the invention, the microscopic evaluation comprises at least one of morphology, subcellular location of stained subcellular structures, and marker intensity of said stained subcellular structures of said cells.

[0095] The term "morphology" as used in the context of the present invention refers to the shape and structure of the cell analysed beyond the molecular level.

[0096] The term "subcellular location" as used herein refers to the subcellular location of a stained intracellular structure within a cell or an extracellular structure on the surface of a cell.

[0097] As used herein, the term "marker intensity" refers to the intensity at which a detection signal is emitted. The intensity of the detection signal may be proportional to the amount of a detectable intracellular structure present in the cell.

[0098] In one preferred embodiment of the method according to the invention, the assessment of the phenotypic signature is performed automatically.

[0099] The term "automatically" as used herein refers to autonomous microscopic evaluation by using an automated microscope, as well as autonomous collection, storage, processing and evaluation of further data by said self-learning multivariate algorithm obtained by the method according to the invention.

[0100] In one preferred embodiment of the method according to the invention, the method comprises comparing with a control from a database.

[0101] The term "comparison" as used herein means that a numerical vector representing a phenotypic signature obtained by the method according to the invention is searched for overlap with a numerical vector representing a phenotypic signature of a control stored in said database.

[0102] The term "control" as used herein encompasses a numerical vector of phenotypic signatures of untreated control cells obtained from a healthy mammal and representative of the innate immune response of a healthy mammal stored in a database for comparison. A control allows matching of the numerical vector of phenotypic signatures of the cells of interest with the numerical vector of phenotypic signatures of controls known to be healthy in immune status or immune response group. In a further preferred embodiment of the method according to the invention, said innate immune response is determined based on the change in the phenotypic signature when compared to the control provided by said database.

[0103] The term "alteration" as used herein refers to a change or deviation of a phenotypic signature from a control cell, from an untreated cell with a known innate immune response, or from a reference cell with a known innate immune response provided by a database, for determining the innate immune response of a cell, in particular the inflammasome activation state. Each tested biological sample from the mammal is compared to a baseline control and a multivariate classification algorithm determines the alteration and assigns a numerical vector, i.e. the corresponding sample, to a designated innate immune response group.

[0104] In summary, the present invention establishes an in vitro diagnostic method that will be an integral part of the physician's decision-making in the diagnosis of AD pathology and / or MCI versus health. The present invention supports diagnosis at different levels: prevention and screening of healthy subjects, diagnosis, prognosis and stratification of asymptomatic and symptomatic disease subjects, and treatment and monitoring of chronic and active disease. Thus, the present invention can be applied in healthy people as a method to thwart disease for prevention and screening. The present invention can also be applied to the diagnosis, prognosis and stratification of symptomatic and asymptomatic patients. Finally, the present invention can be applied in the later stages of disease development for the selection of optimal treatment and monitoring of disease progression.

[0105] As used herein, the terms "of the present invention," "according to the present invention," "in accordance with the present invention," and the like are intended to refer to all aspects and embodiments of the invention described and / or claimed herein.

[0106] In the context of the present invention, the terms "about" and "approximately" indicate a range of accuracy that a person skilled in the art can understand to ensure the technical effect of the feature in question. The terms usually indicate a deviation of ±20%, ±15%, ±10%, e.g. ±5% from the indicated numerical value, preferably ±5% or ±10%. As will be understood by a person skilled in the art, the specific such deviation from the numerical value of a given technical effect will depend on the nature of the technical effect. For example, a natural or biological technical effect will generally be more likely to have such deviation than an artificial or engineered technical effect. As will be understood by a person skilled in the art, the specific such deviation from the numerical value of a given technical effect will depend on the nature of the technical effect. For example, a natural or biological technical effect will generally be more likely to have such deviation than an artificial or engineered technical effect. When an indefinite or definite article is used to refer to a singular noun, such as "a", "an", or "the", this also includes the plural of that noun, unless otherwise specified.

[0107] It will be understood that adapting the teachings of the present invention to a particular problem or environment, and including modifications of the present invention or additional features thereto (such as further aspects and embodiments), will be within the capabilities of one of ordinary skill in the art in light of the teachings contained herein.

[0108] All references, patents, and publications cited herein are hereby incorporated by reference in their entirety.

[0109] The present invention relates to the following items: Item 1. An in vitro method for identifying whether a mammal is suffering from or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or is healthy, comprising: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from: i) no stimulation, ii) Toll-like receptor activation, iii) NOD-like receptor activation, and iv) a combination of Toll-like receptor activation and NOD-like receptor activation; iii) detecting in the medium of the sample and / or in the PBMCs of the sample: TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell;activCell_EstAbsBcell; TNFaRel_Monoc; activCell_EstAbsMonoc; TNFaRel_Tcell;IL1bRel_actCellsObj; Activationrate_percent; IL1bRel_Monoc; IL1b_pgml; TNFaRel_Bcell;TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell; IL1bRel_IL1bPosCell;TNFa_pos_percent; IL1bRel_Tcell; activCell_EstAbsTcell; Analysing a phenotypic marker selected from the group consisting of IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; and iv) identifying said mammal as suffering or likely to suffer from Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI), or as healthy, based on the analysis of phenotypic markers, using a multivariate classification algorithm; A method comprising:

[0110] Item 2. An in vitro method for stratifying a mammal into a healthy group or a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or a healthy group, the method comprising carrying out the method according to item 1, and further comprising: v) stratifying the mammal into a healthy group or a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or a healthy group, based on the analysis of step iv); A method comprising:

[0111] Item 3. An in vitro method for monitoring the progression and / or treatment status of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal receiving treatment for Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), the method comprising: i) providing a first biological sample obtained at a first time point from the mammal to be treated, the first biological sample comprising peripheral blood mononuclear cells (PBMCs); ii) providing a second biological sample obtained at a second time point from the mammal to be treated comprising peripheral blood mononuclear cells (PBMCs); iii) carrying out the method according to item 1 on both the first sample and the second sample; and iv) monitoring the progression of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) and / or the treatment status thereof in said mammal based on the analyzed changes in the first and second phenotypic signatures; A method comprising:

[0112] Item 4. The method of item 3, wherein the first biological sample is obtained prior to initiation of treatment.

[0113] Item 5. The method according to item 3 or 4, wherein the first biological sample is a control sample provided from a healthy patient or patient group, or from a patient or patient group that has undergone another treatment.

[0114] Item 6. The method according to any one of items 1 to 5, wherein the agent is selected from the group consisting of lipopolysaccharide (LPS), LPS and ATP, Leu Leu, and nigericin.

[0115] Item 7. The method according to item 1 for identifying whether a mammal is suffering from or likely to suffer from Alzheimer's disease (AD) or mild cognitive impairment (MCI), comprising: The stimulation in step ii) is an LPS stimulation, wherein the phenotypic markers analyzed in step iii) comprise at least IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; optionally including TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell, and activCell_EstAbsBcell, and The method according to claim 1, wherein a multivariate classification algorithm is used to identify whether the mammal suffers from or is likely to suffer from Alzheimer's disease (AD) or mild cognitive impairment (MCI) based on the analysis of phenotypic markers.

[0116] Item 8. The method according to Item 7, further comprising the stratification step of Item 2 or the monitoring step of any one of Items 3 to 5.

[0117] Item 9. The analysis of phenotypic markers in cells comprises appropriate staining of microscopically assessable intracellular and / or extracellular structures of PBMCs, followed by microscopic evaluation of said stained intracellular and / or extracellular structures, and 9. The method according to any one of items 1 to 8, comprising analysis of phenotypic markers in the culture medium, including suitable identification of soluble marker proteins.

[0118] Item 10. The method of item 9, wherein identifying soluble marker proteins includes obtaining cytokine release data such as TNF-α and IL1β, and microscopic evaluation includes chromatin, PYCARD, IL1β, TNF-α, and CD marker staining.

[0119] Item 11. Identifying phenotypic markers in cells includes generating a cluster map that provides distinct p-value signatures of pairwise comparisons and evaluating the comparisons for statistical differences; Preferably, the method according to any one of items 1 to 10, using a p-value <0.05 or <0.01 using Kruskal-Wallis and / or Games-Howell post-hoc tests.

[0120] Item 12. An in vitro method for identifying a compound for the prevention and / or treatment of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal, the method comprising: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from: i) no stimulation, ii) Toll-like receptor activation, iii) NOD-like receptor activation, and iv) a combination of Toll-like receptor activation and NOD-like receptor activation; iii) suitably contacting said sample with at least one candidate compound; iv) detecting in the medium of the sample and / or in the PBMCs TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell;activCell_EstAbsBcell; TNFaRel_Monoc; activCell_EstAbsMonoc; TNFaRel_Tcell;IL1bRel_actCellsObj; Activationrate_percent; IL1bRel_Monoc; IL1b_pgml;TNFaRel_Bcell; TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell; IL1bRel_IL1bPosCell;TNFa_pos_percent; IL1bRel_Tcell; activCell_EstAbsTcell; Analysing a phenotypic marker selected from the group consisting of IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; and v) identifying the compound based on changes in markers identified after comparing the analysis of step iv) with an analysis of the phenotypic markers in a sample in the absence of the candidate compound and / or with an analysis of the phenotypic markers in a control sample using a multivariate classification algorithm.

[0121] Item 13. The method according to item 12, wherein the phenotypic markers analyzed in step iv) include at least IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; optionally including TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell, and activCell_EstAbsBcell.

[0122] Item 14. The method according to Item 13, wherein the step of analyzing a phenotypic marker is any one of Items 9 to 11.

[0123] Item 15. The method according to any one of items 12 to 14, wherein the compound is selected from a drug, a small molecule, a protein, a peptide, an antibody or a fragment thereof, a polynucleotide, an oligonucleotide, a nanoparticle, a carbohydrate, and a lipid.

[0124] Item 16. The method according to any one of items 1 to 15, wherein the analysis of the phenotypic signature is carried out automatically and preferably comprises a comparison with respective controls in a database.

[0125] Item 17. A marker panel for identifying whether a mammal has or is likely to have Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or is healthy, comprising: TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell;activCell_EstAbsBcell; TNFaRel_Monoc; activCell_EstAbsMonoc; TNFaRel_Tcell;IL1bRel_actCellsObj; Activationrate_percent; IL1bRel_Monoc; IL1b_pgml;TNFaRel_Bcell; TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell; IL1bRel_IL1bPosCell;TNFa_pos_percent; IL1bRel_Tcell; activCell_EstAbsTcell; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef;TNFaPosCells_Monocyte; and TNFa_pgml; and Preferably, the marker panel comprises PBMC markers selected from the group consisting of: IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC;TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; TNFa_pgml; optionally including TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell, and activCell_EstAbsBcell.

[0126] Item 18. A kit comprising the marker panel according to item 17 together with auxiliary agents, preferably suitable buffers, dyes and / or labels, for carrying out the method according to any one of items 1 to 16.

[0127] Item 19. A method for identifying whether the mammal according to Item 1 is suffering from or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or is healthy; For stratifying the mammals according to item 2 into a healthy group or a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or into a healthy group; For monitoring the progression of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) and / or the treatment status thereof as described in item 3, or For identifying a compound for the prevention and / or treatment of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal according to any one of items 12 to 15, 18. Use of a marker panel according to item 17. [Brief description of the drawings]

[0128] [Figure 1] FIG. 1 shows a preferred example of feature fingerprinting (phenotypic signatures) identified according to the present invention for Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) versus healthy donors, showing feature fingerprinting extracted from stimulation with LPS (Toll-like receptor activation, also called "primed"). For markers, see the Examples section. [Diagram 2] FIG. 2 shows the confusion matrix for prediction and classification of AD and / or MCI patients, i.e., AD, MCI, AD / MCI and healthy, according to the present invention. [Figure 3-1] (A) Shown is the final cell-based assay format including the timeline and experimental setup. [Figure 3-2](B) The investigated experimental conditions were: i) "untreated", i.e., no stimulation, i.e., untreated; ii) "primed", i.e., activation of Toll-like receptors, i.e., an activator of TLR2 / 4 receptors (e.g., LPS) was applied; iii) "activated", i.e., activation of NOD-like receptors (e.g., ATP as an NLRP3 activator), and iv) "primed and activated", i.e., a combination of Toll-like receptor activation (e.g., LPS) and NOD-like receptor activation (i.e., ATP); assessed homogeneous readouts and analyzed markers of immunostaining are displayed. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0129] Certain aspects and embodiments of the invention are described herein, by way of example, with reference to the description, figures, and tables set forth herein. Such examples of methods, uses, and other aspects of the invention are merely representative and should not be construed as limiting the scope of the invention to only such representative examples. EXAMPLES

[0130] The present invention provides methods to detect whether an individual's PBMC phenotype deviates from a healthy population, or whether an individual's PBMC phenotype is associated with a given disease phenotype (i.e., AD, MCI, AD / MCI), or whether an individual's PBMC phenotype of a given disease is responsive to a particular treatment for a given disease phenotype (i.e., AD, MCI, AD / MCI). The phenotypic signatures were obtained as described herein by identifying patients and / or patient groups suffering from Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) and healthy individuals, i.e., AD, AD / MCI (MCI with AD, also referred to herein as MCI-AD or MCI and AD (MCI-to-AD)), and healthy.

[0131] The present invention is based on a set of cell-based and / or biochemical markers combined with functional assays investigating the immune system. The resulting features selected from the assay readouts constitute the "phenotypic signature" or "marker panel" of an individual's PBMCs according to the present invention.

[0132] Common methods - overview

[0133] The method of the present invention uses human peripheral blood mononuclear cells (hPBMCs), their phenotypic signature obtained by an automated cell-based assay, and finally focuses on the assessment of TLR2 / 4 signaling and inflammasome responses.

[0134] hPBMCs were left i) unstimulated (i.e., untreated), ii) primed (i.e., Toll-like receptor activation) using an activator of the TLR2 / 4 receptor (e.g., LPS), iii) “activated”, i.e., with NOD-like receptor activation (e.g., ATP as an NLRP3 activator), and iv) “primed and activated”, i.e., with a combination of Toll-like receptor activation (e.g., LPS) and NOD-like receptor activation (e.g., ATP). Priming of hPBMCs induces the activation of specific cellular pathways, resulting in the expression of numerous genes involved in immune responses (e.g., TNF-α and IL-1β). Primed hPBMCs can initiate inflammasome responses upon appropriate stimulation with pathogen-associated molecular patterns (PAMPs) or danger-associated molecular patterns (DAMPs). In this method, the DAMP ATP was used to induce innate immune responses.

[0135] Using a fully automated robotic platform driven cell-based assay, cells were seeded in microtiter plates suitable for automated image acquisition by confocal microscopy. PBMCs are investigated under the following experimental conditions: i) unstimulated (i.e. untreated), ii) Toll-like receptor activation (e.g. lipopolysaccharide, LPS), iii) "activated", i.e. NOD-like receptor activation (i.e. ATP as NLRP3 activator), and iv) combined Toll-like receptor activation (e.g. LPS) and NOD-like receptor activation (e.g. ATP). Thus, hPBMCs are investigated under the following experimental conditions: 1) untreated (resting or untreated), 2) primed (mediated by LPS, a Toll-like receptor (TLR4 / TLR2) activator), 3) primed and activated (via LPS as Toll-like receptor (TLR4 / TLR2) activator in combination with the NLRP3 inflammasome inducer ATP), and 4) activated (via the NLRP3 inflammasome inducer ATP). After a given incubation time, at the end of the assay, cell culture supernatants for each experimental condition were collected, stored frozen, and processed for homogenous biochemical read-out assessment, such as measurement of cytokine release (e.g., TNF-α, IL-1β) by homogenous time-resolved fluorescence (HTRF) assay techniques.

[0136] The hPBMCs on the assay plates are then fixed with paraformaldehyde and processed for immunocytochemistry (ICC) analysis.

[0137] The following markers were investigated: (1) chromatin (associated with features related to cell number data) to assess cell nuclei number and nuclear morphology, (2) IL-1β (associated with IL1b-pos- / IL1bPos-related features) to assess priming and induction of IL-1β expression required for a functional inflammasome response, (3) TNF-α (associated with TNFa-pos- / TNFaPos-related features) to assess the response to priming, and finally (4) PYCARD (Apoptosis-associated speck-like protein containing a CARD) (associated with Activationrate / activCell-related features) to assess inflammasome assembly and inflammasome speck formation.

[0138] The samples are then imaged using an automated microscope and the resulting images are analyzed using an automated image analysis workflow. Figure 3 shows a schematic diagram of the assay workflow, from cell seeding through the various steps of priming, activation, and detection of assay markers, until cell fixation.

[0139] The assay readout core was built based on the integration of different readout values, including but not limited to cellular phenotype descriptors obtained by high content analysis (HCA) of the described immunocytochemical markers, biochemical readouts including but not limited to cytokine analysis (HTRF assays), and clinical data. At this stage, different experimental quantitative information was collected for different markers and readouts. The results of the analysis were converted into a quantitative alphanumeric dataset represented by a set of individual features (see table below).

[0140] Feature collection was performed for each experimental condition (stimuli) and the combined data was constructed into a single numerical vector that represents a phenotypic signature. The phenotypic signature can be combined with other information such as clinical data, medical history data, etc. Each sample and its phenotypic signature are stored in a database.

[0141] Each of the described experimental conditions may induce the expected pheno-molecular phenotype or deviations from it. Deviations from the expected phenotype represent the pathological phenotype captured by the assay technology. The vector obtained for each patient represents the pheno-molecular signature of the patient or control. A baseline signature is constituted by the collection of healthy donors, this data set represents the baseline to which all the samples examined are referred. After data construction, statistical analysis was performed on all features of the vectors to assess the significance of the differences between group-wise comparisons for all the experimental conditions investigated.

[0142] All pheno-molecular vectors were then analyzed using machine learning / artificial intelligence algorithms to finally classify the samples as healthy or diseased. The results of each test sample were then compared to a baseline control, and computational methods were used to assign the samples to either a healthy class or a disease class, as described herein.

[0143] In addition to patient samples, the database also stores information on sample controls, i.e. disease-free individuals and / or individuals suffering from different diseases. These samples constitute the reference set or baseline. Using a dedicated machine learning algorithm, each sample was analyzed and classified as a) control, b) Alzheimer's disease (or other dementia modality under investigation), or c) other disease.

[0144] Analysis example

[0145] We used microscopic images as input data together with features obtained from the supernatant analysis of the analyzed PBMC preparations.

[0146] Features extracted from microscopic images were obtained as follows: Digital images were analyzed by specific image analysis software. Objects of interest were segmented in the images. Individual regions of interest (ROIs) were quantified by extracting quantitative parameters (i.e., intensity-based, morphology-based, object number-based, texture-based, and area-based) and primary feature pools were generated. Quantitative secondary features were obtained by combining two or more primary features when necessary.

[0147] A subset of extracted features was selected to constitute a vector on which the phenotypic signature was based. Within each treatment group, features were assessed for statistical differences (p-value <0.05) using Kruskal-Wallis and Games-Howell post-hoc tests. These tests are non-parametric statistical tests and are generally used to provide evidence of differences between medians of three or more independent groups. Selected features showed significant statistical differences between groups, indicating that patient and control PBMCs respond differently to each treatment.

[0148] The phenotypic markers / phenotypic signatures identified and analyzed herein are:

[0149] [Table 3] JPEG2024529522000005.jpg126134

[0150] Measurement and identification were performed using assays as described in the art, for example as disclosed in PCT / EP2021 / 052306, specifically incorporated by reference.

[0151] As shown in Figure 1, a subset of the features described herein were observed to be statistically different for AD patients (Box A), while another set of features is statistically significant for healthy samples (Box B). This allows identifying and monitoring the response of PBMCs from each treatment group based on features that discriminate between healthy and AD / MCI and / or between AD vs. MCI. Thus, particularly preferred markers are: [Table 4]

[0152] To validate our approach, we analyzed PBMCs with a specific disease phenotype, namely Alzheimer's disease (AD). Thirty AD patients, 20 mild cognitive impairment (MCI) patients (MCI group with AD symptoms), and 35 healthy donors were included. Our method successfully assigned samples from unidentified patients to the correct class with high accuracy, even with limited sample size.

[0153] The resulting cluster maps provided clear p-value signatures (e.g., Games_Howell test-based) for pairwise comparisons of groups (e.g., healthy vs. AD, healthy vs. MCI). The resulting classification of different groups revealed differences between various features within pathological and healthy samples, generating a fingerprint (phenotypic signature) unique to each sample and condition, allowing for improved diagnosis of the associated disease in question. This information can be used to establish a disease-or-health diagnosis.

[0154] In the AD / MCI cohort considered in the context of the present invention, three distinct patient classes were established: 1) healthy, 2) MCI_AD, 3) AD, and 4) MCI without AD. The healthy class corresponds to donors with no predisposition to cognitive impairment. The MCI_AD class corresponds to donors with mild cognitive impairment with onset of Alzheimer's disease. The AD class corresponds to donors diagnosed with Alzheimer's disease.

[0155] For machine learning based classification with a support vector machine algorithm, 4-12 replicate samples were tested per patient. Therefore, to eliminate sample bias, we randomly sampled each patient's data for the replicates so that each patient had the same number of replicates in the dataset. To evaluate the predictive ability of new patients, it is important that the training and testing algorithms do not refer to these patient samples at all during the learning phase, but instead use them only as indicators for predictability. Therefore, we extracted 3 patients from each class that constituted the validation dataset. The remaining data was randomly split in a 70:30 ratio as training and testing data.

[0156] To avoid the algorithm "getting lucky" by random chance, we prepared a 20k-fold validation dataset. This means that the dataset was randomly split 20 times into training, testing and validation sets in the manner described above and evaluated by a classifier to check robustness. The k-fold validation allowed us to evaluate the uncertainty of whether the collected patient samples constituted representative samples of a given disease class or whether more samples were needed to improve the ability to accurately predict the disease class. The resulting feature set was the one according to the invention that gave the highest predictive potential for predicting which disease category a given donor belongs to, with high confidence in the classifier.

[0157] Furthermore, we ranked the robustness of features based on how frequently they are selected as robust features in different k-fold datasets. The more frequently occurring features are ranked as robust. As a result, as shown in Figure 2, the above-mentioned method was used to classify Alzheimer's disease, mild cognitive impairment (MCI) patients, and healthy individuals. Within this approach, the present invention suggests the probability that a given PBMC sample is associated with a given disease. Thus, Figure 2 shows a confusion matrix that predicts that a given sample belongs to either a healthy population or the disease phenotype under observation (AD and / or MCI), as described herein.

[0158] Drawing Terminology p-value P value Healthy True labels Predicted labels PBMC seeding in 384-well assay plate PBMC priming via activation of TLR2 / TLR4 pathway Inflammasome activation via defined and validated activators Supernatant collection for homogenous readout analysis Readouts soluble factors single-plex PBMC fixation and staining for imaging and high content analysis image-based multi-parametric PBMC seeding PBMC seeding LPS priming Inflammasome activation Supernatant collection Fixation and staining Untreated Primed Primed & activated Cytokine assay TNFa release IL1b release IL1b release Activated Automated image analysis Cell type classification

Claims

1. 1. An in vitro test method for aiding in identifying whether a mammal suffers from or is likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or is healthy, said method comprising: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from: i) no stimulation, ii) Toll-like receptor activation, iii) NOD-like receptor activation, and iv) a combination of Toll-like receptor activation and NOD-like receptor activation; iii) detecting in the medium of the sample and / or in the PBMCs of the sample TNFaRel_undefC; IL1bRel_TNFaPosCell;IL1bRel_Bcell;activCell_EstAbsBcell; TNFaRel_Monoc; activCell_EstAbsMonoc;TNFaRel_Tcell; IL1bRel_actCellsObj; Activationrate_percent; IL1bRel_Monoc;IL1b_pgml; TNFaRel_Bcell; TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell;IL1bRel_IL1bPosCell; TNFa_pos_percent; IL1bRel_Tcell; Analysing a phenotypic marker selected from the group consisting of: activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; and iv) based on the analysis of phenotypic markers, using a multivariate classification algorithm to identify said mammal as suffering or likely to suffer from Alzheimer's Disease (AD) or Mild Cognitive Impairment (MCI), or as healthy; A method comprising:

2. 1. An in vitro test method for reference to stratify a mammal into a healthy group or into a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) or into a healthy group, said method comprising the steps of carrying out the method according to claim 1 and further comprising the steps of v) stratifying the mammal into a healthy group or into a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) or into a healthy group based on the analysis of step iv), A method comprising:

3. 1. An in vitro test method for aiding in monitoring the progression and / or treatment status of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal undergoing treatment for Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), said method comprising: i) providing a first biological sample obtained at a first time point from the mammal to be treated, the first biological sample comprising peripheral blood mononuclear cells (PBMCs); ii) providing a second biological sample obtained at a second time point from the mammal to be treated comprising peripheral blood mononuclear cells (PBMCs); iii) performing the method of claim 1 on both the first sample and the second sample; and iv) monitoring the progression of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) and / or the treatment status thereof in said mammal based on the analyzed changes in the first and second phenotypic signatures; A method comprising:

4. The first biological sample comprises: are control samples provided from healthy patients or patient groups, or from patients or patient groups that have undergone a different treatment; or The method of claim 3, obtained prior to the initiation of treatment.

5. 2. The method of claim 1, wherein the agent is selected from the group consisting of lipopolysaccharide (LPS), LPS and ATP, Leu Leu, and nigericin.

6. 13. A method according to claim 1 for aiding in identifying whether a mammal suffers from or is likely to suffer from Alzheimer's disease (AD) or mild cognitive impairment (MCI), comprising: The stimulation in step ii) is an LPS stimulation, The phenotypic markers analyzed in step iii) comprise at least IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml, 2. The method of claim 1, wherein a multivariate classification algorithm is used to identify whether the mammal has or is likely to have Alzheimer's disease (AD) or mild cognitive impairment (MCI) based on the analysis of phenotypic markers.

7. 2. The method of claim 1, wherein the analysis of phenotypic markers in the cells comprises suitable staining of microscopically assessable intracellular and / or extracellular structures of the PBMCs and subsequent microscopic evaluation of said stained intracellular and / or extracellular structures, and the analysis of phenotypic markers in the culture medium comprises suitable identification of soluble marker proteins, or The analysis of phenotypic markers in the cells comprises suitable staining of microscopically assessable intracellular and / or extracellular structures of the PBMCs and subsequent microscopic evaluation of said stained intracellular and / or extracellular structures, and the analysis of phenotypic markers in the culture medium comprises suitable identification of soluble marker proteins, The method of claim 1, wherein identifying soluble marker proteins includes obtaining cytokine release data for TNF-α and IL1β, and microscopic evaluation includes chromatin, PYCARD, IL1β, TNF-α, and CD marker staining.

8. 2. The method of claim 1, wherein identifying phenotypic markers in a cell comprises generating a cluster map providing distinct p-value signatures of pairwise comparisons and evaluating the comparisons for statistical differences; or Identifying phenotypic markers in cells includes generating a cluster map providing distinct p-value signatures of pairwise comparisons and evaluating the comparisons for statistical differences; 2. The method of claim 1, using a p-value <0.05 or <0.01 using Kruskal-Wallis and / or Games-Howell post-hoc tests.

9. 1. An in vitro testing method for aiding in identifying compounds for the prevention and / or treatment of Alzheimer's Disease (AD) and / or Mild Cognitive Impairment (MCI) in a mammal, said method comprising: i) providing at least one biological sample obtained from said mammal comprising peripheral blood mononuclear cells (PBMCs) in a suitable liquid medium; ii) stimulating said PBMCs in said at least one biological sample with an agent selected from: i) no stimulation, ii) Toll-like receptor activation, iii) NOD-like receptor activation, and iv) a combination of Toll-like receptor activation and NOD-like receptor activation; iii) suitably contacting said sample with at least one candidate compound; iv) detecting in the medium of the sample and / or in the PBMCs of the sample: TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell;activCell_EstAbsBcell; TNFaRel_Monoc; activCell_EstAbsMonoc; TNFaRel_Tcell;IL1bRel_actCellsObj; Activationrate_percent; IL1bRel_Monoc; IL1b_pgml;TNFaRel_Bcell; TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell; IL1bRel_IL1bPosCell;TNFa_pos_percent; IL1bRel_Tcell; activCell_EstAbsTcell; Analysing a phenotypic marker selected from the group consisting of IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml; and v) identifying the compound based on changes in markers identified after comparing the analysis of step iv) with an analysis of the phenotypic markers in a sample in the absence of the candidate compound and / or with an analysis of the phenotypic markers in a control sample using a multivariate classification algorithm.

10. The method of claim 9, wherein the phenotypic markers analyzed in step iv) include at least IL1bRel_Tcell; activCell_EstAbsTcell; IL1bPosCells_Monocyte; IL1b_pos_percent; IL1bRel_undefC; TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml.

11. 10. The method of claim 9, wherein the compound is selected from a drug, a small molecule, a protein, a peptide, an antibody or fragment thereof, a polynucleotide, an oligonucleotide, a nanoparticle, a carbohydrate, and a lipid.

12. 10. The method of claim 9, wherein the analysis of the phenotypic signature is performed automatically; or 10. The method of claim 9, wherein the analysis of the phenotypic signatures is performed automatically and includes comparison with respective controls in a database.

13. 1. A marker panel for identifying whether a mammal has or is likely to have Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or is healthy, comprising: TNFaRel_undefC; IL1bRel_TNFaPosCell; IL1bRel_Bcell; activCell_EstAbsBcell;TNFaRel_Monoc; activCell_EstAbsMonoc; TNFaRel_Tcell; IL1bRel_actCellsObj;Activationrate_percent; IL1bRel_Monoc; TNFaRel_Bcell;TNFaRel_IL1bPosCell; TNFaRel_TNFaPosCell; IL1bRel_IL1bPosCell;TNFa_pos_percent; IL1bRel_Tcell; A marker panel comprising PBMC markers selected from the group consisting of: TNFaCells_IL1bCells; activCell_EstAbsUndef; TNFaPosCells_Monocyte; and TNFa_pgml.

14. A kit comprising the marker panel according to claim 13 together with auxiliary agents, suitable buffers, dyes and / or labels for carrying out the method according to any one of claims 1 to 12.

15. For aiding in identifying whether a mammal according to claim 1 suffers from or is likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or is healthy, 3. To aid in stratifying the mammals according to claim 2 into a healthy group or a group suffering or likely to suffer from Alzheimer's disease (AD) and / or mild cognitive impairment (MCI), or into a healthy group; For monitoring the progression and / or treatment status of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) according to claim 3, or For reference in identifying compounds for the prevention and / or treatment of Alzheimer's disease (AD) and / or mild cognitive impairment (MCI) in a mammal according to any one of claims 9 to 12, Use of the marker panel according to claim 13.

Images