Methods, systems and computer program for detecting effects of substances, particularly nanoparticles and toxins, on cell compartments

WO2026215301A3PCT designated stage Publication Date: 2026-11-19ARTIDIS AG
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/EP2026/060087
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-04-07
Filing Date
2026-04-07
Publication Date
2026-11-19

AI Technical Summary

Technical Problem

Conventional in vitro toxicity assays fail to detect the intracellular uptake of substances like nanoparticles and toxins, particularly at sublethal doses or early time points, which is crucial for understanding neurodegenerative diseases and drug efficacy.

Method used

A method using an indenter to detect force interactions between cells and a cantilever, generating force data to assess cellular uptake of substances like nanoparticles or toxins, enhanced by processing algorithms to determine mechanical properties and compartment localization.

Benefits of technology

Provides a direct, objective, and reproducible assessment of cellular uptake, improving sensitivity and consistency, suitable for toxicology and neurodegeneration research, and reducing dependence on expert interpretation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure EP2026060087_19112026_PF_FP_ABST
    Figure EP2026060087_19112026_PF_FP_ABST
Patent Text Reader

Abstract

The present invention relates to a method for detecting the uptake of a target substance by cells in vitro or ex vivo, the method comprising: contacting a population of cells in vitro or ex vivo with the target substance; indenting the sample with an indenter, detecting force signals indicative of a force interaction between the cells and the indenter, generating force data from the force signals, processing the force data, and detecting the intracellular uptake of the target substance based on said processed force data.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] Methods, systems and computer program for detecting effects of substances, particularly nanoparticles and toxins, on cell compartments

[0002] Description:

[0003] The present invention relates to methods, systems and computer programs for Understanding the interaction between exogenous substances and living cells is essential in fields ranging from nanomedicine and toxicology to neurodegenerative disease research. In particular, the uptake of materials such as nanoparticles and toxic compounds into cells plays a critical role in determining biological responses, transport mechanisms, and intracellular effects.

[0004] Conventional in vitro toxicity assays, such as the MTT, MTS, and luciferase-based assays, primarily assess cellular viability, metabolic activity, or reporter gene expression. While such assays are useful for detecting overt cytotoxicity, they often fail to detect whether a substance has actually entered the cells, particularly in cases where the intracellular presence of the substance does not result in immediate functional impairment or cell death. This is especially problematic in the study of substances that exert sublethal or compartment-specific effects, or that require uptake for mechanistic understanding.

[0005] In this context, the term "uptake" refers to the internalization of a target substance into the intracellular space, including cytosolic, vesicular, and organelle-associated compartments. Uptake may occur via active transporters, endocytosis, pinocytosis, or passive diffusion, depending on the nature of the substance and the cell type.

[0006] Among the materials of interest are nanoparticles, including metallic nanoparticles, metal oxide nanoparticles, and plastic nanoparticles. These are widely used in biomedical research, diagnostics, and environmental exposure studies. However, their interaction with cells is complex, and standard viability assays often yield false negatives when evaluating their uptake.

[0007] In parallel, the study of toxic compounds, particularly neurotoxins, is of central importance in the field of neurodegeneration. For example, 1-methyl-4-phenylpyridinium (MPP+) is a well-known neurotoxin and the active metabolite of MPTP, capable of selectively targeting dopaminergic neurons via the dopamine transporter (DAT). MPP+induces mitochondrial dysfunction and oxidative stress. The human neuroblastoma cell line SH-SY5Y can be differentiated to adopt a dopaminergic phenotype and is frequently used as an in vitro model for studying DA neuron function and toxin uptake.A / 0

[0008] Particularly, the study of dopaminergic (DA) neurons is essential for unravelling the mechanisms of Parkinson’s disease (PD) and developing effective therapeutic strategies. The versatility of the SH-SY5Y cell line in modelling DA neuronal properties and associated pathologies has established this neuroblastoma derived model as a valuable in vitro tool [1], Differentiation of SH-SY5Y cells using retinoic acid (RA) and the phorbol ester 12-0-tetradecanoyl-phorbol-13-acetate (TPA) induces a pronounced dopaminergic phenotype [2], These differentiated cells exhibit characteristics like DA neurons in vivo, including susceptibility to 1-methyl-4-phenylpyridinium (MPP+)-induced toxicity, a hallmark of PD related neurodegeneration.

[0009] Understanding the structural adaptations of differentiated SH-SY5Y cells is crucial for accurately mimicking the behaviour of DA neurons in vivo. While the biochemical and functional properties of these cells have been extensively studied, the mechanical properties remain significantly underexplored. Cellular mechanics influence a wide range of biological processes, including cytoskeletal organization, cellular signaling, and resilience to stress - all of which are integral to the pathology of neurodegenerative diseases like PD. Despite their importance, there is a critical lack of comprehensive studies linking nanomechanical changes to neuronal functionality and vulnerability in contexts of disease. Addressing this gap is essential to fully realize the potential of differentiated SH-SY5Y cells as a reliable model for DA neurons. However, standard cytotoxicity assays may fail to reveal whether MPP+or similar toxins have been taken up by these cells, especially at sublethal doses or early time points. Thus, based on the above, the problem to be solved by the present invention is to provide an improved method, system and computer program for detecting intracellular uptake.

[0010] This problem is solved by the method according to claim 1 , a system according to claim 22 and a computer program according to claim 23. Preferred embodiments thereof are stated in the corresponding dependent claims and are described below.

[0011] Further aspects of the present invention and embodiments thereof are also disclosed and described herein.

[0012] According to claim 1 a method for detecting the uptake of a target substance by cells in vitro (or ex vivo) is disclosed, the method comprising:

[0013] i) contacting (e.g. incubating) a population of cells in vitro (or ex vivo) with the target substance;

[0014] ii) indenting the sample with an indenter,

[0015] iii) detecting force signals indicative of a force interaction between the cells and the indenter,iv) generating force data from the force signals,

[0016] v) optionally processing the force data (to generate processed force data), and vi) detecting the cellular uptake of the target substance based on said force data and / or said processed force data.

[0017] In particular, the method according to the present invention enables a more direct, objective, and reproducible assessment of cellular uptake of target substances, such as nanoparticles or toxins (e.g. neurotoxins), thereby overcoming the limitations of conventional viability-based assays that rely on indirect or downstream cellular responses. Furthermore, the analysis is significantly enhanced in terms of sensitivity, consistency, and scalability, particularly when applied to heterogeneous cell populations or specialized cell types such as dopaminergic neurons. The method also reduces dependence on expert interpretation, increases experimental throughput, and supports integration into research and preclinical workflows, rendering it particularly suitable for applications in toxicology, neurodegeneration research, and drug screening.

[0018] Particularly, in the framework of the present invention, the target substance is any material to be detected within the cells after exposure, including but not limited to nanoparticles and toxins such as neurotoxic compounds.

[0019] Furthermore, said cellular uptake is understood as the internalization of the target substance into the intracellular space, including entry via transporters, endocytosis, membrane diffusion, vesicle-mediated transport, or any other mechanism.

[0020] The method and system are carried out on cells or tissues that have been removed from the human or animal body, or otherwise provided prior to execution of the method, such that the method is performed in vitro or ex vivo, and is not practiced on the human or animal body. Furthermore, in a preferred embodiment, the indenter is realized as a cantilever of an AFM or as a microindenter or a nanoindenter.

[0021] The force signals can thus correspond to a deflection of a cantilever forming at least a portion of the indenter, wherein the deflection is detected during the force interaction between the sample and the indenter. The deflection can be determined optically, for instance by detecting a light or laser beam reflected from the top side of the cantilever while it interacts with the sample with its bottom side. Specifically, case of an AFM, the force signals can comprise a vertical and a lateral deflection signal of the cantilever, reflecting the topology of the sample. A sum signal may be determined from the summation of the vertical and the lateral deflection signal. Moreover, the force signal can comprise displacements associated with the deflection. For example, each force signal can be associated with a displacement of the cantilever atwhich the force signal is detected. Additionally, the force signals can comprise temporal information and / or external vibrations, for example of a platform on which the AFM is used. For example, for the detection of force signals and the generation of the force data from the force signals, the indenter and the sample may be moved with respect to each other as follows while force signals indicative of a force interaction between the sample and the indenter are detected and a displacement of the indenter is recorded:

[0022] in an approach phase, approaching the sample and the indenter,

[0023] in a contact event, establishing contact between the surface of the sample and the indenter,

[0024] in a loading indentation phase, indenting the sample with the indenter until a predefined indentation depth of the indenter on the sample is reached, causing a deformation of the sample, particularly a surface of the sample in contact with the indenter, wherein the deformation may be particularly an elastic deformation or a plastic deformation, particularly for elastic samples and elastic deformations of the sample, in an unloading indentation phase, moving the sample indented by the indenter and the indenter away from each other, wherein the sample and the indenter remain in contact with each other and an elastic deformation of the sample caused by the indenter may be at least partially reversed and

[0025] in a separation phase, separating the sample and the indenter from each other.

[0026] The displacement of the indenter can be determined for example as a distance between a starting position of the indenter controlled by an actuator and a relative position of the indenter with respect to the starting position while moving the indenter toward, onto, into, out of and away from the sample. Alternatively, the sample may be moved toward the indenter and the relative position is determined with respect to a starting position of the sample.

[0027] The indentation depth of the indenter on the sample can be determined from a difference between the displacement of the indenter at its position at the contact event and its position at the predefined indentation depth.

[0028] Particularly, one or more force maps are determined by detecting force signals at a plurality of locations on a surface of the sample. Respective force data, particularly force-displacement data or force-displacement curves can be determined from the force signals detected at each location. Particularly, to this end, the plurality of locations can be arranged in a grid, particularly a regular grid on the surface of the sample. For example, each force map comprises a grid of 32x32 locations.

[0029] From the force signals and particularly the associated displacement of the indenter, said force data can be generated. For example, in case of an AFM, the force data can be generated from converting the cantilever deflection to a force acting between the cantilever and the sample,which force can be derived from the known geometry and material properties of the cantilever. As such, the force data can particularly reflect force-displacement data representing forcedisplacement curves, which relate forces between the sample and the indenter with distances between the indenter and the sample and / or indentation depths of the indenter on the sample. According to an embodiment of the invention, the processing of the force data includes a preprocessing, wherein the pre-processing comprises at least one of the following:

[0030] a smoothening algorithm for smoothening the detected force signals and / or the generated force data,

[0031] a baseline correction algorithm for correcting a baseline tilt of the detected force signals and / or the generated force data,

[0032] an indentation depth correction algorithm for determining a corrected indentation depth of the indenter on the sample, particularly accounting for a deflection of a cantilever forming at least a portion of the indenter,

[0033] a contact location determination algorithm for determining a contact location of the indenter on the sample.

[0034] The smoothing algorithm can be used to reduce noise in the force-distance data. This helps in better visualization and more accurate analysis of the curve. In particular, the smoothening algorithm can contribute to the removal of irrelevant features such as noise in the force data, which renders the employed machine learning algorithm more efficient. Common smoothing techniques include moving average, Savitzky-Golay filters, and Gaussian smoothing, which help in highlighting the underlying trends and features of the data by reducing random fluctuations.

[0035] By means of the baseline correction algorithm, a slope or tilt in a baseline of the force data, particularly force-distance data, can be at least partially, particularly completely removed. A baseline tilt can arise due to instrumental drift or uneven surfaces. Its removal further enhances the accuracy of the force measurements. For example, a linear fit is applied to a region of the force-distance curve in which the indenter and the sample are not in contact, particularly to adjust the baseline to a horizontal level, i.e. a level of zero slope.

[0036] The indentation depth correction algorithm is particularly advantageous when using an AFM with a cantilever as the indenter, as it allows to determine the actual indentation depth of the a of the cantilever and the sample surface. The displacement of the cantilever can correspond to a driveway of the actuator for moving the cantilever. However, once the cantilever starts bending due to the force interaction between the cantilever and the sample, the driveway does no longer reflect the actual indentation of the cantilever on the sample, due to the finite deflection of the cantilever. For the indentation depth correction algorithm, the deflection of thecantilever is subtracted from the driveway of the actuator, for example a piezo movement, to obtain the true indentation depth.

[0037] The contact location determination algorithm allows to determine a location at which the indenter, upon an approach between the indenter and the sample, first touches the sample surface. Identifying this point is advantageous as it can serve as a reference for further analysis of the force data. The contact location can be determined from detecting positional information of the indenter on the sample surface during a contact event. The contact event can for example be detected by an abrupt change of the detected force signal or the force data, or its gradients.

[0038] According to a preferred embodiment of the invention, processing of the force data comprises determining one or more mechanical properties of the cells from the force data and particularly comprises generating mechanical property data representing, particularly comprising, the one or more mechanical properties.

[0039] According to a preferred embodiment of the invention, the mechanical properties are selected from the list comprised of: an elastic modulus of the cells, a dissipation between the cells and the indenter, a stiffness (also denoted as modulus) of the cells, a topology of the cells, an adhesion between the cells and the indenter. The mechanical properties can be combined in as desired, i.e., all of them can be considered, a selection of them, or individual properties (such as stiffness) etc.

[0040] It can further be distinguished between mechanical force data determined during the loading phase of the indentation and mechanical force data determined during the unloading phase of the indentation.

[0041] In particular, the elastic modulus of the sample can be determined from the force data, particularly from the force-displacement data for example using the Hertz-Sneddon model or the Oliver-Pharr model. The Hertz-Sneddon model can be used to analyze the contact mechanics between a spherical or conical indenter tip and the sample. The Hertz-Sneddon model relates the indentation depth to the applied force, considering the elastic properties of the sample and the geometry of the tip. It allows for the extraction of the elastic modulus from the force data. For example, for a spherical tip, the force F can be related to the indentation 5 by means of the equation F = ^E*yjR63 / 2, where E* the reduced elastic modulus and R is the

[0042]

[0043] radius of the tip. For other tip geometries, for example a conical tip, different relationships can be used based on the tip geometry. The Oliver-Pharr model is a widely used method for analyzing indentation data to extract mechanical properties, such as stiffness and elastic modulus, from force data, particularly the force-displacement curves. The Oliver-Pharr model in particularly uses force data determined from the unloading indentation phase, using a stiffness determined during the loading phase of the indentation to determine the contact areabetween the indenter and the sample and subsequently the sample material's properties. To this end, the slope of the force-displacement curve during the unloading indentation phase is determined and this slope is used to calculate the reduced elastic modulus, taking into account the geometry of the indenter and the indentation depth.

[0044] The elastic modulus can be understood a mechanical property that indicates the resistance of the sample to being deformed elastically during interaction with the indenter. The elastic modulus E can be extracted from the measured force-displacement curves, for example using the Hertz-Sneddon model, using the relation between the reduced elastic modulus E* and the elastic modulus E, E* = E / (l - v2), wherein v denotes the sample-dependent Poisson’s ratio. For example, v can be between 0.4 and 0.5.

[0045] The stiffness can be understood as a mechanical property that indicates the resistance offered by an elastic body to deformation. For example, the stiffness can be determined from a gradient or slope of the force data, particularly the force-displacement curve, during the loading phase of the indentation and / or the unloading phase of the indentation.

[0046] The dissipation can be understood as a mechanical property that indicates the energy dissipated due to internal friction of the sample during the force interaction with the indenter. The dissipation can be determined for example in terms of a hysteresis between the loading phase and the unloading phase of the indentation.

[0047] The adhesion can be understood as a mechanical property that indicates a force interaction between the sample and the indenter during the separation phase. Specifically, the adhesion can be the force acting between the sample and the indenter during the separation phase. According to a preferred embodiment of the invention, detecting the intracellular uptake of the target substance comprises comparing the mechanical property data with a control, wherein particularly the control is a (corresponding) mechanical property reference data obtained for cells not exposed to the target substance.

[0048] According to yet another preferred embodiment of the invention, the target substance is or comprises nanoparticles.

[0049] According to a preferred embodiment of the invention, the nanoparticle is selected from the group comprised of: metallic nanoparticles (e.g. gold, silver, copper, and platinum), metal oxide nanoparticles (e.g. zinc oxide, titanium dioxide, and iron oxide), plastic nanoparticles (e.g. polystyrene, polyethylene, and polypropylene).

[0050] According to a preferred embodiment of the invention, the cells are selected from the group consisting of culture cells, human cells, mammalian cell lines, human cell lines, HeLa9903 cells.Furthermore, according to a preferred embodiment of the invention, the target substance comprises a toxic compound.

[0051] Further, in preferred embodiment of the invention, the toxic compound is or comprises 1-methyl-4-phenylpyridinium (MPP+).

[0052] According to a preferred embodiment of the invention, the cells are selected from the group consisting of: culture cells, human cells, mammalian cell lines, human cell lines, SH-SY5Y cells, SH-SY5Y cells differentiated by treatment with retinoic acid and 12-O-tetradecanoyl-phorbol-13-acetate (TPA). This treatment is mainly used for SH-SY5Y cell lines, but has also been applied to C4-2 and LNCaP prostate cancer cells (apoptosis), in V79 cell (inhibition of metabolic cooperation), HL-60 leukemia cells and SK-N-BE(2) neuroblastoma cell line.

[0053] Furthermore, in a preferred embodiment, cell compartments of the cells are identified from the force data, preferably by using a machine learning algorithm, yielding segmented spatial regions corresponding to cell nuclei, cell cytoplasms, and cell junctions.

[0054] In yet another embodiment of the invention, cell compartments of the sample are identified from the force data, particularly from said force maps, using the machine learning algorithm. Typically, different cell compartments, for example cell nucleus, cell cytoplasm and cell junction comprise different mechanical properties. For example, a cell nucleus of the sample is typically stiffer than cytoplasm. That is, the machine learning algorithm can identify and distinguish between different cell compartments by processing the force data, particularly force data determined at a plurality of locations on the surface of the sample. Different cell compartments can be visualized in real space using the force data of the force maps, for example parameterized by a mechanical property such as stiffness, or abstract features determined by the machine learning algorithm.

[0055] In particular, cell compartment information is comprised by the mechanical property data that can be input to the machine learning algorithm. The cell compartment information can be a classification of each location under investigation into different cell compartment types, such as cell nucleus, cell cytoplasm or cell junction.

[0056] According to yet another preferred embodiment, a spatial distribution of the uptaken target substance is determined with respect to the segmented spatial regions, to identify the localization of the target substance within one or more cell compartments.

[0057] A further embodiment of the method relates to detecting the uptake of a target substance in cells in vitro or ex vivo into a specific compartment of the cell.Another embodiment of the method relates to detecting the uptake of a target substance in cells in vitro or ex-vivo into the cell or its compartments and measuring effect on the function of the cell or compartments.

[0058] In another embodiment of the invention, the cell compartment is selected from the group consisting of: a cell nucleus, a cell cytoplasm, a cell junction.

[0059] According to an embodiment of the invention, an interaction state of the target substance with the cells is determined based on said force data and / or said processed force data.

[0060] In another embodiment of the invention, detecting the cellular uptake comprises evaluating at least one nuclei-associated backward modulus-family parameter together with at least one extra-nuclear parameter.

[0061] In an embodiment of the invention, detecting the cellular uptake comprises evaluating at least one nuclei-associated backward modulus-family parameter together with at least one extra-nuclear parameter.

[0062] In particular, the nuclei-associated backward modulus-family parameter is selected from Young's modulus (backward), reduced modulus (backward), precise modulus (backward), and precise reduced modulus (backward).

[0063] In particular, the extra-nuclear parameter comprises Backward convexity, dissipation (contact only), contact-area dissipation, an extra-nuclear backward modulus-family parameter, or any combination thereof.

[0064] A further aspect of the present invention relates to a system for detecting the uptake of a target substance by cells in vitro (or ex vivo), comprising:

[0065] an indenter for indenting the cells,

[0066] a detector for detecting force signals indicative of a force interaction between the cells and the indenter, and

[0067] a computer, adapted to execute the steps iv), v), and vi) of the method according to one of the present invention.

[0068] According to yet another aspect of the present invention, a computer program is disclosed comprising instructions to cause the computer of the system according to the invention (or the system according to the invention) to execute the steps of the method according to the invention.

[0069] According to a further aspect of the present invention a computer-readable medium having stored thereon the computer program according to the present invention is disclosed. Particularly the computer readable medium is a non-volatile computer readable storage medium.According to a further aspect of the present invention, a data carrier signal carrying the computer program according to the present invention is disclosed.

[0070] A further aspect of the invention pertains to a method for detecting cell compartments in cells in vivo. This process includes indenting the sample with an indenter and detecting force signals indicative of a force interaction between the cells and the indenter. Force data is generated from these force signals. Optionally, the force data is processed to generate processed force data, and cell compartments are identified based on the force data and / or the processed force data. It should be noted that the term 'cell compartment' refers to spatially and / or functionally distinct regions within a cell, including the nucleus, cytoplasm, and cell junctions.

[0071] One embodiment of this aspect elaborates that the step of identifying cell compartments utilizes a machine learning algorithm in the processing of the aforementioned force and / or processed force data.

[0072] In a further embodiment, the processing of the force data involves determining one or more mechanical properties of the cells from the force data. Further, mechanical property data encoding these properties might be generated. Specifically, the mechanical properties could be selected from a list including, but not limited to, an elastic modulus of the cells, a dissipation between the cells and the indenter, the cells' stiffness, their topology, or an adhesion between the cells and the indenter (see also above)

[0073] In yet another embodiment of this aspect of the invention, the identifying step provides segmented spatial regions corresponding to cell nuclei, cell cytoplasms, and cell junctions. Furthermore, an embodiment could also involve the method introducing an additional step prior to the indenting step, which includes contacting the cells in vitro (or ex vivo) with a specified target substance.

[0074] In a preferred embodiment, the specified target substance can be selected from diverse options, including nanoparticles, metallic nanoparticles, metal oxide nanoparticles, plastic nanoparticles, polystyrene nanoparticles, polyethylene nanoparticles, polypropylene nanoparticles, a neurotoxin, or 1-methyl-4-phenylpyridinium.

[0075] A further preferred embodiment can include specifying the cells as cancerous.

[0076] According to yet a further aspect of the present invention a system for detecting cell compartments in cells is disclosed. This system encompasses an indenter, a detector for picking up force signals indicative of a force interaction between the cells and the indenter, and a computer, structured to carry out steps of the method described in Aspect 1 and its aforementioned embodiments.

[0077] Yet a further aspect relates to a computer program consisting of instructions enabling a computer or the above-described system to perform the steps iii), iv), and v) of the method.According to a further aspect of the present invention a computer-readable medium having stored thereon the above-described computer program according to the present invention is disclosed. Particularly the computer readable medium is a non-volatile computer readable storage medium.

[0078] According to a further aspect of the present invention, a data carrier signal carrying this computer program according to the present invention is disclosed.

[0079] In the following, aspects of the present invention and embodiments thereof are stated as items. Particularly, these items may also be formulated as claims of the present application or a divisional application.

[0080] Item 1:

[0081] A computer-implemented method for determining an interaction state of a target substance with respect to a biological sample, the method comprising: (a) generating and / or receiving atomic force microscopy (AFM) force data from an exposed biological sample and from a reference biological sample or reference dataset; (b) deriving, from the AFM force data, values of one or more nanomechanical parameters at a plurality of measurement locations; (c) assigning the plurality of measurement locations to at least one cellular region; (d) generating one or more region-resolved features from the one or more nanomechanical parameters; (e) computing, from the one or more region-resolved features, one or more scores indicative of uptake, localization, burden, interaction, remodeling, stress, toxicity, viability-independent response, or a combination thereof for the target substance; and (f) outputting a classification, ranking, report, or decision based on the one or more scores.

[0082] Item 2. The method of item 1 , wherein the method does not require that the same party that performs the AFM analysis also performs cell exposure, nanoparticle addition, cell staining, confocal imaging, hyperspectral imaging, or flow-cytometric measurement.

[0083] Item 3. The method of item 1 or 2, wherein the at least one cellular region comprises a WHOLE_CELL region, a NUCLEI region, an EXTRA_NUCLEI region, or any combination thereof.

[0084] Item 4. The method of item 3, wherein EXTRA_NUCLEI is obtained by direct AFM measurement outside nuclei, by image-guided exclusion of nuclei, by a residual or difference function between WHOLE_CELL and NUCLEI representations, by a mask-based subtraction, or by a learned non-nuclear compartment model.

[0085] Item 5. The method of any one of items 1 to 4, wherein the one or more nanomechanical parameters comprise one or more of Young's modulus (forward), Young's modulus (backward), reduced modulus (forward), reduced modulus (backward), precise modulus (forward), precise modulus (backward), precise reduced modulus (forward), precise reducedmodulus (backward), adhesion force, retraction adhesion force, adhesion energy, dissipation, revised dissipation, dissipation (contact only), contact-area dissipation, forward convexity, forward concavity, backward convexity, backward concavity, topography, or transforms, ratios, z-scores, deltas, or embeddings thereof.

[0086] Item 6. The method of any one of items 1 to 5, wherein the reference biological sample or reference dataset comprises untreated cells, time-zero controls, intra-run controls, historical reference data, a reference atlas, particle-free controls, matched controls, or any combination thereof.

[0087] Item 7. The method of any one of item 1 to 6, wherein the one or more scores comprise a signed effect, an effect magnitude, a fold-change, a distance-from-reference value, an area-under-the-curve statistic, a classifier probability, a latent-variable score, a trajectory score, a burden-aligned score, or any combination thereof.

[0088] Item 8. The method of any one of items 1 to 7, wherein the target substance comprises one or more nanoparticles, toxins, biologies, drugs, impurities, formulations, nanoforms, process variants, or combinations thereof.

[0089] Item 9. The method of any one of items 1 to 8, wherein determining uptake or localization comprises evaluating at least one nuclei-associated backward modulus-family parameter together with at least one extra-nuclear parameter.

[0090] Item 10. The method of item 9, wherein the nuclei-associated backward modulus-family parameter is selected from Young's modulus (backward), reduced modulus (backward), precise modulus (backward), and precise reduced modulus (backward).

[0091] Item 11. The method of item 9 or 10, wherein the extra-nuclear parameter comprises Backward convexity, dissipation (contact only), contact-area dissipation, an extra-nuclear backward modulus-family parameter, or any combination thereof.

[0092] Item 12. The method of any one of items 1 to 11, wherein a response or stress score is computed using one or more adhesion-family and / or dissipation-family parameters from WHOLE_CELL and / or NUCLEI.

[0093] Item 13. The method of any one of items 1 to 12, wherein a multivariate model combines parameters from at least two cellular regions.

[0094] Item 14. The method of item 13, wherein the multivariate model combines WHOLE_CELL and NUCLEI measurements, WHOLE_CELL and EXTRA_NUCLEI measurements, NUCLEI and EXTRA_NUCLEI measurements, or WHOLE_CELL, NUCLEI, and EXTRA_NUCLEI measurements.Item 15. The method of item 13 or 14, wherein the multivariate model comprises a regularized linear model, logistic regression model, tree-based model, ensemble model, support-vector machine, neural network, latent-variable model, metric-learning model, or any combination thereof.

[0095] Item 16. The method of any one of items 1 to 15, wherein AFM measurements are acquired at a baseline time point and at one or more post-exposure time points and the one or more scores further comprise a time-phase feature.

[0096] Item 17. The method of item 16, wherein the time-phase feature comprises a Fourier-domain feature, a Laplace-domain feature, a cumulative burden integral, a mean response time, a modulation amplitude, a persistence ratio, or a compartment-specific trajectory descriptor. Item 18. The method of any one of items 1 to 17, wherein the output indicates that the biological sample exhibits a mechanically detectable interaction state notwithstanding absent or below-threshold signal in an orthogonal optical assay.

[0097] Item 19. A computer-implemented method for identifying cell compartments from AFM force data, comprising: deriving one or more nanomechanical parameters from AFM force measurements taken at a plurality of locations on a biological sample; applying a rule-based, statistical, machine-learning, or hybrid segmentation procedure to the one or more nanomechanical parameters and / or to representations derived therefrom; and generating a compartment map assigning the plurality of locations to nuclei-associated, extra-nuclear, whole-cell, cytoplasmic, membrane-associated, junction-associated, or other cellcompartment classes.

[0098] Item 20. The method of item 19, wherein the segmentation procedure uses spatial context, force-curve shape, parameter distributions, image guidance, supervised learning, unsupervised clustering, graph-based methods, atlas matching, or any combination thereof. Item 21. The method of item 19 or 20, wherein an EXTRA_NUCLEI representation is generated by subtracting, excluding, masking, or otherwise removing NUCLEI-associated contributions from a WHOLE_CELL representation.

[0099] Item 22. The method of any one of items 19 to 21, wherein the compartment map is used to attribute substance burden and / or biomechanical response to one or more cell compartments. Item 23. A method of training or calibrating an AFM analysis model, comprising: receiving AFM force data or derived features from a plurality of substance conditions; associating the AFM force data or derived features with training labels corresponding to exposed versus control states, burden states, localization states, stress states, toxicity states, process-control states, or comparability states; and training or calibrating a model to output one or more scores according to any one of items 1 to 18.Item 24. The method of item 23, wherein the training labels are generated without requiring any orthogonal assay and are based on exposed-versus-control definitions, time-resolved labels, known condition metadata, process metadata, or combinations thereof.

[0100] Item 25. The method of item 23 or 24, wherein orthogonal assay data are optionally used as supplemental labels, constraints, or calibration signals, the orthogonal assay data comprising confocal microscopy, fluorescence microscopy, hyperspectral imaging, electron microscopy, flow cytometry, spectroscopy, or combinations thereof.

[0101] Item 26. The method of any one of items 23 to 25, wherein separate submodels are trained for different particle sizes, particle chemistries, concentrations, time points, cell types, or combinations thereof.

[0102] Item 27. The method of any one of items 23 to 26, wherein a universal panel is selected by ranking features according to cross-particle stability, cross-time stability, cross-compartment separability, or combinations thereof.

[0103] Item 28. A method of screening, ranking, equivalence testing, or comparability assessment for candidate substances, nanoforms, lots, formulations, process variants, or supplier materials, comprising applying the method of any one of items 1 to 18 and outputting a rank, pass / fail call, equivalence call, release-support call, or process-change decision.

[0104] Item 29. The method of item 28, wherein the candidate substances differ in size, composition, concentration, coating, surface chemistry, charge, aggregation state, corona, manufacturing process, storage state, or combinations thereof.

[0105] Item 30. The method of item 28 or 29, wherein the method is used for nanoform stewardship, lot release support, process-change control, customer assurance, toxicology support, formulation screening, regulatory support, or combinations thereof.

[0106] Item 31. An AFM analysis system comprising: (i) an atomic force microscope or an interface configured to receive force-spectroscopy or indentation data generated by an atomic force microscope; (ii) one or more processors; and (iii) memory storing instructions that, when executed by the one or more processors, cause the one or more processors to perform the method of any one of items 1 to 30.

[0107] Item 32. The system of item 31, wherein the memory further stores a reference atlas, one or more trained models, one or more compartment definitions, one or more decision thresholds, and a report generator.

[0108] Item 33. The system of item 31 or 32, wherein the atomic force microscope uses a probe having a tip radius from 5 nm to 100 nm and a peak force from 0.1 nN to 10 nN.Item 34. The system of item 33, wherein the probe has a tip radius of about 20 nm and the peak force is about 1.8 nN.

[0109] Item 35. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform the method of any one of items 1 to 30.

[0110] Item 36. The method of any one of items 1 to 30, wherein the biological sample comprises live cells, fixed cells, primary cells, immortalized cells, tumor cells, epithelial cells, immune cells, stem cells, organoids, tissues, co-cultures, cell monolayers, or combinations thereof.

[0111] Item 37. The method of any one of items 1 to 30, wherein the target substance comprises polymeric particles, metal particles, fluorescent particles, silica particles, carbonaceous particles, lipids, proteinaceous particles, toxins, or combinations thereof.

[0112] Item 38. The method of any one of items 1 to 30, wherein the output comprises both a burden-aligned extra-nuclear score and a nuclei remodeling score.

[0113] Item 39. The method of any one of items 1 to 30, wherein the output comprises a signed-effect representation and an effect-magnitude representation for the same nanomechanical feature or panel.

[0114] Item 40. The method of any one of items 1 to 30, wherein the method is performed on AFM data acquired locally, on AFM data supplied by another party, or on a combination thereof. Exemplary embodiments are described below in conjunction with the Figures. The Figures are appended to the claims and are accompanied by text explaining individual features of the shown embodiments and aspects of the present invention. Each individual feature shown in the Figures and / or mentioned in the text of the Figures may be incorporated (also in an isolated fashion) into a claim relating to the first, second, third, fourth and / or fifth aspect according to the present invention.

[0115] Fig. 1 shows typical force data generated from indenting a sample taken from a human or animal with an indenter as a function of a displacement of the indenter indenting the sample;

[0116] Fig. 2 shows a schematic illustration of an embodiment of the method indicating the generation of the force data (force curves) and determination of mechanical properties of the analyzed cells (here, e.g. dissipation, adhesion, and modulus (stiffness));

[0117] Fig. 3 illustrates treating SH-SY5Y cells with retinoic acid and 12-0- tetradecanoyl-phorbol-13-acetate (TPA) to achieve differentiated (treated) cells,Fig. 4 shows a representative heat map of the stiffness component for the SH-SY5Y control cell line (A), and the representative heatmap of the stiffness component for the SH-SY5Y cells treated with retinoic acid and 12-O-tetradecanoyl-phorbol-13-acetate (TPA);

[0118] Fig. 5 indicates the effect of differentiation by treatment with retinoic acid and 12-O-tetradecanoyl-phorbol-13-acetate (TPA) on the mechanical properties of SH-SY5Y cells in comparison to control (non-treated SH- SY5Y cell line) based on differences in stiffness, adhesion, and dissipation (A) and shows relative expression of dopaminergic and neuronal markers (B) as well as differences in nanomechanical signature (C),

[0119] Fig. 6 Fig. 6A shows forward stiffness (upon loading the indenter) and Fig. 6B backward stiffness (upon unloading the indenter) for the control (untreated differentiated SH-SY5Y) and differentiated SH-SY5Y,

[0120] Fig. 7 shows adhesion (A) and dissipation (B) for the control (untreated SH- SY5Y) and treated (differentiated) SH-SY5Y,

[0121] Fig. 8 shows the toxicological effect and uptake of polystyrene (PS) nanoparticles on the stiffness of HeLa cells. The inlay shows the dependency of stiffness component with PS nanoplastic concentration.

[0122] Fig. 9 Figs. 9A-C show the effect of polystyrene (PS) nanoparticles on different cell compartments of HeLa9903 cell culture,

[0123] Fig. 10 shows Hela9903 cells interacting with labeled PS nanoparticles (concentration 10ppm and size 350nm),

[0124] Fig. 11 shows application of three different nanoparticles, i.e., polyethylene (PE), polystyrene (PS) and polypropylene (PP) to Hela9903 cells for a concentration of 10ppm of the respective nanoparticle,

[0125] Fig. 12 shows segmented cells (A) and the reduced elastic modulus (B)

[0126] determined from the force data generated on the surface of a tissue sample, wherein different cell compartments of the sample show different characteristics of the elastic modulus,

[0127] Fig. 13 shows different bulk forward elastic moduli (upon loading the indenter) measured with an AFM tip on HeLa9903 cells administered withdifferent types of nanoparticles, including polystyrene nanoparticles (top), fluorescent polystyrene nanoparticles (center) and gold nanoparticles (bottom), for two different sizes (50 nm and 300 nm), two different with concentrations (10 ppm and 100 pm) and two different incubation times (3h and 24h),

[0128] Fig. 14 shows different nuclear forward elastic moduli (upon loading the indenter) measured with an AFM tip on HeLa9903 cells administered with different types of nanoparticles, including polystyrene nanoparticles (top), fluorescent polystyrene nanoparticles (center) and gold nanoparticles (bottom), for two different sizes (50 nm and 300 nm), two different with concentrations (10 ppm and 100 pm) and two different incubation times (3h and 24h),

[0129] Fig. 15 show comparisons of bulk and nuclear forward elastic moduli for nanoparticles sized 50 nm and with an incubation time of 3h (Fig. 15A), sized 300 nm with an incubation time of 3h (Fig. 15B), sized 50 nm with an incubation time of 24h (Fig. 15C) and sized 300 nm with an incubation time of 24h (Fig. 15D),

[0130] Fig. 16 compares the measured bulk forward elastic moduli for HeLa9903 cells treated with the different nanoparticles of different sizes, in different concentrations and for different incubation periods (red squares interconnected by dashed lines) to the percentage of live cells determined by flow cytometry (blue dots interconnected by solid lines),

[0131] Fig. 17 compares the measured bulk forward elastic moduli for HeLa9903 cells treated with the different nanoparticles of different sizes, in different concentrations and for different incubation periods (red squares interconnected by dashed lines) to the early apoptosis of cells determined by flow cytometry (blue dots interconnected by solid lines),

[0132] Fig. 18 compares the measured bulk forward elastic moduli for HeLa9903 cells treated with the different nanoparticles of different sizes, in different concentrations and for different incubation periods (red squares interconnected by dashed lines) to the late apoptosis of cells determined by flow cytometry (blue dots interconnected by solid lines),

[0133] Fig. 19 compares the measured bulk forward elastic moduli for HeLa9903 cells treated with the different nanoparticles of different sizes, in differentconcentrations and for different incubation periods (red squares interconnected by dashed lines) to the percentage of necrosis of cells determined by flow cytometry (blue dots interconnected by solid lines),

[0134] Fig. 20 shows representative confocal images of a HeLa9903 sample treated with 300nm fPS nanoparticles at a concentration of 100 ppm after 24h incubation time,

[0135] Fig. 21 shows representative images of a HeLa9903 sample treated with 300 nm AuNPs (10 ppm, 3 h incubation), as a z-stack maximum-intensity projection with synchronized pixel intensity values (LLITs), the upperleft panel displays the merged image of all four channels: Hoechst (blue), phalloidin (green), nanoparticles (red), and CytoLiner (violet),

[0136] Fig. 22 compares the localization of fluorescent nanoparticles (top) to gold nanoparticles (bottom), for different cell compartments including nucleus, membrane and cytoplasm and for the different sizes, concentrations and incubation times,

[0137] Fig. 23 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 50 nm, with a concentration of 10 ppm and an incubation time of 3h,

[0138] Fig. 24 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 50 nm, with a concentration of 100 ppm and an incubation time of 3h,

[0139] Fig. 25 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 50 nm, with a concentration of 10 ppm and an incubation time of 24h,

[0140] Fig. 26 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 50 nm, with a concentration of 100 ppm and an incubation time of 24h,

[0141] Fig. 27 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 300 nm, with a concentration of 10 ppm and an incubation time of 3h, andFig. 28 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 300 nm, with a concentration of 10 ppm and an incubation time of 24h.

[0142] Fig. 1 shows an example of typical force data generated from indenting an in vitro or ex vivo sample that has been e.g. obtained from a subject, particularly from a human or animal, with an indenter as a function a displacement of the indenter indenting the sample. These force data can be used for the ex-vivo detection of uptake of target substances in the sample (e.g. cells).

[0143] In the present example of Fig. 1, the force data are determined by means of an atomic force microscope (AFM) with a cantilever. To generate the force data, the sample is placed below the cantilever such that the cantilever can be moved with respect to the sample in order to indent it with a tip of the cantilever. For example, the sample is a cell line such as SH-SY5Y cells, particularly differentiated by treatment with retinoic acid and 12-O-tetradecanoyl phorbol-13-acetate (TPA), or other cells as disclosed herein.

[0144] The cantilever can be moved with respect to the sample by means of an actuator, in the present example by a piezo actuator configured to move the cantilever at least along a distance direction defined by a displacement between the cantilever and the sample. The cantilever can also be configured to be moved in two or three orthogonal spatial directions, which allows not only to vary a distance between the sample and the cantilever, but also to obtain force data from different locations on a surface of the sample. Alternatively, the cantilever can be fixed in space and the sample can be moved relative to the sample in one, two or three orthogonal directions.

[0145] Fig. 1 shows force data 1 generated for a single location on the surface of the sample. The force data 1 can be determined from force signals that correspond to a deflection of a cantilever caused by a force interaction between the sample and the cantilever, while moving the cantilever along the displacement direction and recording the displacement. The deflection of the cantilever can be determined optically, by measuring an intensity of a light beam projected onto a photodetector from a backside of the cantilever facing away from the sample in terms of a voltage after amplification of a photocurrent triggered in the photodetector. This quantity is shown on the left y-axis of Fig. 1. The force signals, i.e. the deflection of cantilever can be converted to a force acting on the cantilever due to the force interaction between the sample and the cantilever. Force data 1 can be generated from the determined forces. In particular, the force data 1 can be determined from, at a predetermined location on the surface of the sample, moving the cantilever toward, onto, into, out of and away from the sample, while recording the deflection of the cantilever and the displacement, i.e. the driveway of theA / 0

[0146] cantilever driven by an actuator during the displacement. The relative movement between the cantilever and the sample can be described in a sequence comprising the following phases:

[0147] - An approach phase, in which the cantilever approaches the sample. In the approach phase, the cantilever deflection and the force determined from the cantilever deflection are essentially zero, as the cantilever is not in contact with the sample.

[0148] A contact event, in which contact is established between the surface of the sample and the indenter. The contact event can be determined from an abrupt onset of increasing cantilever deflection and corresponding forces.

[0149] - A loading indentation phase, in which the sample is indented with the cantilever until a predefined indentation depth of the indenter on the sample is reached. The indentation causes a deformation of the sample, particularly of the surface of the sample in contact with the cantilever. Depending on the sample properties, the deformation may be particularly a reversible, elastic deformation or a non-reversible, plastic deformation. The indentation depth can be determined from the difference of the displacement of the cantilever with respect to a contact point associated with the contact event. In particular, when a cantilever is used, an indentation correction algorithm as disclosed herein may be used to account for the finite deflection of the cantilever for determining the indentation depth.

[0150] In particular for elastic samples and elastic deformations of the sample, an unloading indentation phase, in which the cantilever is moved away from the sample indented by the indenter, wherein the sample and the indenter remain in contact with each other and an elastic deformation of the sample caused by the indenter is at least partially reversed.

[0151] A separation phase, separating the sample and the indenter from each other.

[0152] The force data 1 generated in the above-described manner comprise valuable information of the tissue structure of the sample. In particular, mechanical properties of the sample can be determined from the force data 1, for example an elastic modulus of the sample, a dissipation between the sample and the indenter, a stiffness of the sample and / or an adhesion between the sample and the indenter. These mechanical properties reveal even more information about the tissue structure. In the present invention, the information comprised in the force data 1 can be used for the ex-vivo diagnosis of cancer in the sample, for predicting outcome of a cancer treatment for the subject from who the sample has been obtained and / or for determining a cancer treatment recommendation for the subject, by means of a machine learning algorithm 7.

[0153] Fig. 1 shows force data 1 associated with a single location on the surface of the sample, reflecting a one-dimensional (1 D) force-displacement curve. This force-displacement curvecan comprise the approach phase, the contact event, the indentation phase, corresponding to a movement of the indenter into the sample, but it may also additionally include the unloading phase and the separation phase described above. However, in the context of the present invention, the force data 1 can comprise a plurality of 1D force-displacement curves determined at different locations on the surface of the sample. In this way, the tissue structure of the sample can be thoroughly investigated, which improves the reliability of the diagnosis, prediction and / or treatment recommendation according to the invention.

[0154] Fig. 2. shows a schematic illustration of an embodiment of the method according to the invention indicating the generation of the force data (force curves) and determination of mechanical properties of the analyzed cells (here, e.g. dissipation, adhesion, and modulus (stiffness)). Results of the method are shown e.g. in Figs. 8 to 11.

[0155] Fig. 3 shows an exemplary process of differentiating (treating) SH-SY5Y by treatment with retinoic acid and 12-O-tetradecanoyl-phorbol-13-acetate (TPA).

[0156] Fig. 4A-B show differentiated (treated) SH-SY5Y cells (Fig. 4B) compared to the control (nontreated SH-SY5Y cells (Fig. 4a). Fig. 5A states values for the mechanical properties with respect to the treated and non-treated (control) SH-SY5Y cells. Differentiation of SH-SY5Y cells using RA and TPA successfully induces dopaminergic properties. Relative expression of dopaminergic and neuronal markers in SH-SY5Y cells confirms differentiation (cf. Fig. 5B). control and treated cells SH-SY5Y can be distinguished by the method according to the present invention (e.g. Fig. 5C). Difference in the mechanical properties are observed with statistical differences. The higher stiffness in differentiated cells likely indicates reorganization of cytoskeletal components, which play essential roles in neuronal stability, signaling, and intracellular transport.

[0157] Figs. 6A to 7B show the individual mechanical properties for control and treated SH-SY5Y cells.

[0158] Further Fig. 8 shows the toxicological effect and uptake of polystyrene (PS) nanoparticles on the stiffness of HeLa cells. It indicates the stiffening upon the interaction with PS nanoplastics at different non-toxic concentrations. The inlay shows the dependency of Stiffness component with PS nanoplastic concentration.

[0159] Finally, Figs. 9A-C show the effect of polystyrene (PS) nanoparticles on different cell compartments of HeLa9903 cell culture. The administration of PS_(size 350nm_and concentration 10ppm) increases the overall stiffness of the HeLa cells which subsequently complicates the discernment between the nuclear stiffness peak and the perinuclear stiffness peak. The administration of higher doses of PS (350nm, 100 ppm), also shows the stiffeningA / 0 77

[0160] on all the compartments of the cells. However, this impact is less pronounced than the rigidity observed at the 10ppm concentration.

[0161] Fig. 8 also indicates the uptake of the target substances (PS nanoparticles), as the application of PS nanoparticles (350nm at 10 ppm and 100 ppm) results in alterations in cellular properties, indicating a concentration-dependent cellular response to nanoparticle exposure. Internalized nanoparticles can increase cytoskeletal stiffness or redistribute mechanical properties across the cell (observed at different cell compartments Fig. 9).

[0162] Furthermore, Fig. 10 shows the Hela cells interacting with labeled PS nanoparticles (concentration 10ppm and size 350nm) to observe / confirm the internalization of the particles. Fig. 11 finally compares the uptake of three different nanoparticles, i.e., polyethylene (PE), polystyrene (PS) and polypropylene (PP) into Hela9903 cells for a concentration of 10ppm of the respective nanoparticle.

[0163] Further, Fig. 12A shows different cell compartments of a tissue sample by means of a cell nuclei, cytoplasm and cell junctions. The cell nuclei are each surrounded by cytoplasm. The different cell nuclei and their respective surrounding cytoplasm are separated from each other by cell junctions. The different cell compartments can be identified from the force data 1, particularly from said force maps, using the machine learning algorithm. Typically, different cell compartments comprise different mechanical properties. For example, a cell nucleus of the sample is typically stiffer than cytoplasm. That is, the machine learning algorithm can identify and distinguish between different cell compartments by processing the force data 1, particularly force data 1 determined at a plurality of locations on the surface of the sample. Different cell compartments can be visualized in real space as in Fig. 12A using the force data of the force maps, for example parameterized by a mechanical property such as stiffness, or abstract features determined by the machine learning algorithm. This segmentation can be used to map the uptake of a target substance with respect to the cell compartments.

[0164] Fig. 12B shows distribution densities of the determined reduced elastic modulus in the area of the cytoplasm, cell junction and nucleus, demonstrating their different nanomechanical properties that can be used by the machine learning algorithm to identify the different cell compartments. The differentiation between force data features determined at different compartments allows for a more reliable determination of the diagnosis of cancer, as the different nanomechanical properties of different cell compartments under investigation are taken into account by the machine algorithm.

[0165] In particular, the method and system according to the present invention enable a more direct, objective, and reproducible assessment of intracellular uptake of target substances, such asATDI I ^A / n 77

[0166] nanoparticles or neurotoxins, thereby overcoming the limitations of conventional viabilitybased assays that rely on indirect or downstream cellular responses.

[0167] Furthermore, the analysis is significantly enhanced in terms of sensitivity, consistency, and scalability, particularly when applied to heterogeneous cell populations or specialized cell types such as dopaminergic neurons. The method also reduces dependence on expert interpretation, increases experimental throughput, and supports integration into research and preclinical workflows, rendering it particularly suitable for applications in toxicology, neurodegeneration research, and drug screening.

[0168] Fig. 13 shows different bulk forward elastic moduli (upon loading the indenter) measured with an AFM tip on HeLa9903 cells administered with different types of nanoparticles, including polystyrene nanoparticles (top), fluorescent polystyrene nanoparticles (center) and gold nanoparticles (bottom). In particular, for each type of nanoparticles, the effect of different sizes of the respective nanoparticle has been investigated for two different sizes (50 nm and 300 nm), two different with concentrations (10 ppm and 100 pm) and two different incubation times (3h and 24h). The data can be compared to an untreated HeLa9903 control sample.

[0169] The administration of nanoparticles 50 nm and 300 nm initially (lower incubation times and low concentrations) increases the overall stiffness of HeLa cells. A saturation-like effect occurs with increasing incubation time and particle concentration, particularly when plain polystyrene (PS) nanoparticles without fluorescent properties are used. In the case of fluorescent polystyrene (fPS) and gold nanoparticles (AuNPs), the initial stiffening is reversed to values similar to those ones observed in the controls.

[0170] Fig. 14 shows different nuclear forward elastic moduli (upon loading the indenter) measured with an AFM tip on nuclear regions of HeLa9903 cells administered with the different types of nanoparticles, including polystyrene nanoparticles (top), fluorescent polystyrene nanoparticles (center) and gold nanoparticles (bottom), for the two different sizes, concentrations and incubation times.

[0171] Figures 15A-D show comparisons of bulk and nuclear forward elastic moduli for nanoparticles sized 50 nm and with an incubation time of 3h (Fig. 15A), sized 300 nm with an incubation time of 3h (Fig. 15B), sized 50 nm with an incubation time of 24h (Fig. 15C) and sized 300 nm with an incubation time of 24h (Fig. 15D). In each case the effect can be understood from comparing the data to the reference data of the untreated HeLa9903 control cells (left datapoint in each plot). Fig. 16 compares the measured bulk forward elastic moduli for HeLa9903 cells treated with the different nanoparticles of different sizes, in different concentrations and for different incubation periods (red squares interconnected by dashed lines) to the percentage of live cells determined by flow cytometry (blue dots interconnected by solid lines).Fig. 17 compares the measured bulk forward elastic moduli for HeLa9903 cells treated with the different nanoparticles of different sizes, in different concentrations and for different incubation periods (red squares interconnected by dashed lines) to the early apoptosis of cells determined by flow cytometry (blue dots interconnected by solid lines).

[0172] Fig. 18 compares the measured bulk forward elastic moduli for HeLa9903 cells treated with the different nanoparticles of different sizes, in different concentrations and for different incubation periods (red squares interconnected by dashed lines) to the late apoptosis of cells determined by flow cytometry (blue dots interconnected by solid lines).

[0173] Fig. 19 compares the measured bulk forward elastic moduli for HeLa9903 cells treated with the different nanoparticles of different sizes, in different concentrations and for different incubation periods (red squares interconnected by dashed lines) to the percentage of necrosis of cells determined by flow cytometry (blue dots interconnected by solid lines).

[0174] Fig. 20 shows representative confocal images of a HeLa9903 sample treated with 300nm fPS nanoparticles at a concentration of 100 ppm after 24h incubation time. Shown is a z-stack maximum-intensity projection with synchronized pixel intensity values (LLITs). The upper-left panel displays the merged image of all four channels: Hoechst (blue), phalloidin (green), nanoparticles (red), and CytoLiner (violet).

[0175] Fig. 21 shows representative images of a HeLa9903 sample treated with 300 nm AuNPs (10 ppm, 3 h incubation). Shown is a z-stack maximum-intensity projection with synchronized pixel intensity values (LLITs). The upper-left panel displays the merged image of all four channels: Hoechst (blue), phalloidin (green), nanoparticles (red), and CytoLiner (violet).

[0176] The images of Fig. 20 and 21 have been recorded using confocal microscopy.

[0177] Fig. 22 compares the localization of fluorescent nanoparticles (top) to gold nanoparticles (bottom), for different cell compartments including nucleus, membrane and cytoplasm and for the different sizes, concentrations and incubation times. The localization was determined by confocal microscopy.

[0178] Figures 23 to 28 show hyperspectral dark field images confirming the intracellular localization of AuNPs 50nm and AuNPs 300nm in HeLa9903 cells following 3h or24h exposure at both 10 ppm and 100ppm, with a predominantly cytoplasmic distribution. Clear spectral signatures match AuNPs 50 nm and AuNPs 300 nm: The accumulation of nanoparticles is dosedependent. Most of AuNPs appear to localize in the cytoplasm. The scale bar in the hyperspectral images is 16 pm long.

[0179] Fig. 23 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 50 nm, with a concentration of 10 ppm and an incubation time of 3h. The hyperspectral images canA / 0 5

[0180] be compared to the untreated HeLa9903 control sample. Red signal indicates AuNPs 50nm localization based on spectral matching.

[0181] Fig. 24 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 50 nm, with a concentration of 100 ppm and an incubation time of 3h. The hyperspectral images can be compared to the untreated HeLa9903 control sample. Colored signal indicates AuNPs 50nm localization based on spectral matching.

[0182] Fig. 25 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 50 nm, with a concentration of 10 ppm and an incubation time of 24h. The hyperspectral images can be compared to the untreated HeLa9903 control sample. Green signal indicates AuNPs 50nm localization based on spectral matching.

[0183] Fig. 26 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 50 nm, with a concentration of 100 ppm and an incubation time of 24h. The hyperspectral images can be compared to the untreated HeLa9903 control sample. Red signal indicates AuNPs 50nm localization based on spectral matching.

[0184] Fig. 27 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 300 nm, with a concentration of 10 ppm and an incubation time of 3h. The hyperspectral images can be compared to the untreated HeLa9903 control sample. Red signal indicates AuNPs 50nm localization based on spectral matching.

[0185] Fig. 28 shows hyperspectral dark field images of HeLa9903 cells treated with AuNPs of 300 nm, with a concentration of 10 ppm and an incubation time of 24h. The hyperspectral images can be compared to the untreated HeLa9903 control sample. Red signal indicates AuNPs 50nm localization based on spectral matching.

[0186] References

[0187] [1] Kalia LV, Lang AE. The Lancet (2015) 386(9996): 896-912.

[0188] [2] Presgraves SP, Ahmed T, Borwege S, Joyce JN. Neurotox Res. (2003) 5(8):579-598.

Claims

Patent claims1. A method for detecting the uptake of a target substance in cells in vitro or ex vivo, the method comprising:i) contacting a population of cells in vitro or ex vivo with the target substance;ii) indenting the sample with an indenter,iii) detecting force signals indicative of a force interaction between the cells and the indenter,iv) generating force data from the force signals,v) optionally processing the force data to generate processed force data, and vi) detecting the cellular uptake of the target substance based on said force data and / or said processed force data.

2. The method according to one of the preceding claims, wherein processing of the force data comprises determining one or more mechanical properties of the cells from the force data and particularly comprises generating mechanical property data encoding the one or more mechanical properties.

3. The method according to claim 2, wherein the mechanical properties are selected from the list comprised of: an elastic modulus of the cells, a dissipation between the cells and the indenter, a stiffness of the cells, a topology of the cells, an adhesion between the cells and the indenter.

4. The method according to claim 2 or 3, wherein detecting the intracellular uptake of the target substance comprises comparing the mechanical property data with a control.

5. The method according to claim 4, wherein the control is a mechanical property reference data obtained for cells not exposed to the target substance.

6. The method of one of the preceding claims, wherein the target substance comprises one or several nanoparticles.

7. The method according to claim 6, wherein the respective nanoparticle is selected from the group comprised of: metallic nanoparticles, metal oxide nanoparticles, plastic nanoparticles, polystyrene (PS) nanoparticles, polyethylene (PE) nanoparticles, polypropylene (PP) nanoparticles.

8. The method of one of the preceding claims, wherein the cells are selected from the group consisting of culture cells, human cells, mammalian cell lines, human cell lines, HeLa9903 cells.

9. The method according to one of the claims 1 to 5, wherein the target substance comprises a toxic compound.

10. The method according to claim 9, wherein the toxic compound comprises 1-methyl-4- phenylpyridinium (MPP+).

11. The method according to one of the preceding claims, wherein the cells are selected from the group consisting of: culture cells, human cells, mammalian cell lines, human cell lines, SH-SY5Y cells, SH-SY5Y cells differentiated by treatment with retinoic acid and 12-O-tetradecanoyl-phorbol-13-acetate (TPA), C4-2 cells, LNCaP cells, V79 cells, HL- 60, SK-N-BE(2), HeLa9903 cells.

12. The method according to one of the preceding claims, wherein cell compartments of the cells are identified from the force data using a machine learning algorithm, yielding segmented spatial regions corresponding to cell nuclei, cell cytoplasms, and cell junctions.

13. The method according to claim 12, wherein a spatial distribution of the uptaken target substance is determined with respect to the segmented spatial regions, to identify the localization of the target substance within one or more cell compartments.

14. The method according to one of the preceding claims, wherein step vi) comprises detecting the cellular uptake of the target substance based on said force data and / or said processed force data into a specific cell compartment of the cells.

15. The method according to one of the preceding claims, wherein step vi) comprises detecting the cellular uptake of the target substance based on said force data and / or said processed force data into the cells or into compartments of the cells and measuring an effect on the function of the cells or compartments.

16. The method according to one of the claims 12 to 15, wherein the cell compartment is selected from the group consisting of: a cell nucleus, a cell cytoplasm, a cell junction.

17. The method according to one of the preceding claims, wherein an interaction state of the target substance with the cells is determined based on said force data and / or said processed force data.

18. The method according to one of the preceding claims, wherein detecting the cellular uptake comprises evaluating at least one nuclei-associated backward modulus-family parameter together with at least one extra-nuclear parameter.A / 0 g19. The method according to one of the claims 13 to 18, wherein identifying the localization comprises evaluating at least one nuclei-associated backward modulus-family parameter together with at least one extra-nuclear parameter.

20. The method according to claim 18 or 19, wherein the nuclei-associated backward modulus-family parameter is selected from Young's modulus (backward), reduced modulus (backward), precise modulus (backward), and precise reduced modulus (backward).

21. The method according to one of the claims 18 to 20, wherein the extra-nuclear parameter comprises Backward convexity, dissipation (contact only), contact-area dissipation, an extra-nuclear backward modulus-family parameter, or any combination thereof.

22. A system for detecting the uptake of a target substance by cells in vitro or ex vivo, comprising:an indenter for indenting the cells,a detector for detecting force signals indicative of a force interaction between the cells and the indenter, anda computer, adapted to execute the steps iv), v), and vi) of the method according to one of the preceding claims.

23. Computer program comprising instructions to cause the computer of claim 22 to execute the steps of the method according to one of the claims 1 to 21.

24. A method for detecting cell compartments in cells in vivo, the method comprising:i) indenting the sample with an indenter,ii) detecting force signals indicative of a force interaction between the cells and the indenter,iii) generating force data from the force signals,iv) optionally processing the force data to generate processed force data, and v) identifying cell compartments based on said force data and / or said processed force data.

25. The method according to claim 24, wherein step v) corresponds to identifying cell compartments based on said force data and / or said processed force data using a machine learning algorithm.

26. The method according to claim 24 or 25, wherein step v) yields segmented spatial regions corresponding to cell nuclei, cell cytoplasms, and cell junctions.l. The method according to one of the claims 24 to 26, wherein prior to step i) the method further comprises the step of contacting the cells in vitro with a target substance.

28. The method according to claim 27, wherein the target substance is selected from the group consisting of: nanoparticles, metallic nanoparticles, metal oxide nanoparticles, plastic nanoparticles, polystyrene (PS) nanoparticles, polyethylene (PE) nanoparticles, polypropylene (PP) nanoparticles, a neurotoxin, 1-methyl-4-phenylpyridinium (MPP+).

29. The method according to one of the claims 24 to 28, wherein the cells are cancerous cells.

30. A system for detecting cell compartments in cells, comprising: an indenter for indenting the cells, a detector for detecting force signals indicative of a force interaction between the cells and the indenter, and a computer, adapted to execute the steps iii), iv) and v) of the method according to one of the claims 24 to 29.

31. Computer program comprising instructions to cause the computer of claim 30 to execute the steps of the method according to one of the claims 24 to 29.