阅读说明：本技术 确定肿瘤细胞的侵袭潜能的方法 (Method for determining the invasive potential of tumor cells ) 是由 玛纽尔·特里 约兰·马格伦 奥黛尔·菲尔霍尔-科切特 于 2018-05-30 设计创作，主要内容包括：本发明涉及用于预测肿瘤细胞的侵袭潜能和预测治疗效率的方法。所述方法基于至少三个参数即核间距离、极化和细胞-细胞连通度的组合。(The present invention relates to methods for predicting the invasive potential of tumor cells and predicting the efficiency of treatment. The method is based on a combination of at least three parameters, namely internuclear distance, polarization and cell-cell connectivity.)

1. A method for predicting the invasive potential of a tumor cell, the method comprising:

-providing a culture device comprising a planar substrate and two adhesive patterns on said substrate for culturing at least two sets of adherent cells,

wherein the two adhesive patterns are sufficiently separated from each other by a substantially non-adhesive insertion region to prevent cells on a first adhesive pattern from reaching the other adhesive pattern; and, the area covered by the two adhesive patterns and the insertion region is sufficient to adhere two animal cells, and wherein the two adhesive patterns of each set are separated by a surrounding region of lipophobicity;

-seeding at least one tumor cell on the set of two adhesive patterns;

-culturing said tumor cells so as to obtain one cell doublet from one cell division on said set of two adhesion patterns;

-measuring at least the following three parameters on the set of two adhesion patterns for the tumor cells:

-the internuclear distance of the two adhesion patterns of each group;

-cell connectivity of the two adhesion patterns of each group; and

-a percentage of highly polarized cells to the total cell population on the set of two adhesion patterns;

-determining the invasive potential of the tumor cell based on the at least three measured parameters.

2. The method of claim 1, wherein the distance between the two adhesive patterns is between about 2/3D and about 4/3D, D being the diameter of the surface S covered by the cells on the support without any constraint, and the area covered by the adhesive patterns is less than 80% of the cell surface S covered by the cells on the support without any constraint for each cell.

3. The method of claim 1 or 2, wherein the set of two adhesive patterns has an H-form.

4. The method of any one of claims 1 to 3, wherein the highly polarized cells are cells with centrosomes located near the cell-cell junction in triangular regions as defined below:

|y|≤(x-0.5)*tan 60

wherein x is more than or equal to 0.5 and less than or equal to 2

Wherein x is the position of the centrosome on the nucleus-nucleus axis pointing to the intercellular junction and y is the position of the centrosome on an axis perpendicular to said nucleus-nucleus axis, in units of the radius of the nucleus;

the percentage of highly polarized cells is the ratio of the number of highly polarized cells to the total number of cells x100 as defined above.

5. The method of any one of claims 1 to 4, wherein said cell-cell connectivity is the difference between the area of the convex envelope around said cell doublet and the area of said cell doublet.

6. The method of any one of claims 1 to 5, wherein, in addition to the tumor cells to be tested, reference cells having a predetermined invasive potential are seeded on the device, cultured and the parameter measured.

7. The method of claim 6, wherein the reference cell is selected from the group consisting of: cells with very low or no invasive potential, cells with low invasive potential, cells with moderate invasive potential, cells with high invasive potential and cells with very high invasive potential.

8. The method of any one of claims 1 to 7, wherein the invasive potential of the tumor cell is determined by comparison to a reference cell.

9. The method of any one of claims 6 to 8, wherein a statistical analysis is performed to determine the reference cell with the highest similarity to the tested tumor cells.

10. The method of claim 9, wherein the statistical analysis is performed by principal component analysis.

11. A method for providing information useful for the diagnosis or prognosis of a subject, the method comprising performing the method of any one of claims 1 to 10, thereby determining the invasive potential of tumor cells from the subject.

12. A method for assessing the efficiency of a molecule for treating a tumor in a subject, the method comprising contacting the molecule with tumor cells from a tumor of the subject, determining the invasive potential of the tumor cells treated with the molecule, comparing the determined invasive potential of the tumor cells treated with the molecule to the invasive potential of untreated tumor cells, and selecting a molecule that decreases the invasive potential, wherein the invasive potential of the tumor cells is determined by the method of any one of claims 1 to 10.

13. The method of any one of claims 1 to 12, wherein the tumor is a solid tumor.

14. The method of claim 13, wherein the solid tumor is a carcinoma.

15. The method of any one of claims 1 to 14, wherein the tumor is selected from the group consisting of breast cancer, colorectal cancer, lung cancer, kidney cancer, thyroid cancer, prostate cancer, liver cancer, pancreatic cancer, head and neck cancer and ovarian cancer, preferably from the group consisting of breast cancer, colon cancer, lung cancer, kidney cancer and thyroid cancer.

Technical Field

The present invention relates to the field of medicine, in particular oncology. More specifically, the invention relates to a method for classifying tumors according to aggressiveness.

Background

Cancer invasion is a pathophysiological process that occurs locally from the metastatic spread of primary tumor tissue into distant organs.

Cancer diagnosis from tumors is in fact mainly achieved from histological examination of tumor biopsies. Based on qualitative estimates of sample composition (cell and nucleus morphology, mitotic frequency), clinicians can classify tumors on a tree scale by using the Nottingham histological scoring system (also known as the Elston-Ellis modified version of the Scarff-Bloom-Richardson grading system) (Elston CW et al, (1991) Histopathology 19: 403-. Based on this classification, these methods can be used to assess cancer aggressiveness and are crucial for selecting the optimal treatment regimen. However, since these methods are based on subjective and qualitative assessment of histological features of tissue samples damaged by surgery, they do not allow reliable quantification of the cell characteristics of tumor origin and consequently accurate classification of the invasive potential of tumors.

The invading tumor cell initially involves deep reorganization of the overall morphology of its intracellular tissue, including reorientation of its polarity. Indeed, a defect in the ability of cells to adapt their internal polarity to external events is a hallmark of cancer (Wodarz A et al, (2007) Nature-Cell biology (Nat Cell Biol.)9: 1016-. The main program of molecular events that trigger invasion of cells from a quiescent state is called epithelial-mesenchymal transition (EMT). This multi-step process (1) involves reversal of epithelial polarity, (2) partial to complete cell personalization, and (3) initial invasion of cells into surrounding tissues (Yeung KT et al, (2017) molecular oncology (Mol Oncol.)11: 28-39). In recent studies, Th ery and colleagues have demonstrated that polarity reversal is a key event in EMT (BureteM et al, (2017.) developmental cells (DevCell.)40, 168-184).

As a prerequisite for cellular polarity, the asymmetry of the internal Cell tissue is conferred by an anisotropic distribution of microtubules, which emanate from the nucleation organelles, the centrosomes (Bornensm M, (2008), Nature review-molecular Cell Biol 9: 874-886). The central body is described as the main microtubule tissue center, which adopts an eccentric positioning that is highly correlated with the internal cell polarity. As a marker of cellular polarity, centrosomes are associated with more than a hundred regulatory proteins, making them spatial organizers for organelle localization and signaling pathways (Doxsey S et al, (2005) Annu Rev Cell DevBiol 21: 411-434). Notably, in epithelial cells, the centrosomes are located near the cell-cell junction at the apical pole, but this localization is disrupted after the cells are dissociated and plated onto the infinite surface of a petri dish, traditionally used for culturing adherent mammalian cells. This suggests that classical culture of adherent cells on an immortal adhesive surface is unable to capture this important feature of centrosomal localization modulation and is then not suitable for revealing subtle changes that occur during the onset of tumor invasion.

In order to promote the adhesion of cells at specific and defined locations, EP1664266 and EP2180042 disclose a culture device comprising an adhesive surface with defined shapes of adherable cells, called adhesive micropatterns. These patent applications also disclose a method of controlling cell shape and internal tissue, and the use of such a device in screening for biologically active compounds. In such devices, once cells are attached to such a surface, they adopt the overall shape of the surface using the provided adhesion triggers, while cellular organelles are arranged in a controlled tissue manner. The cells can then be normalized according to the spatial configuration of the cells using the adhesive surface geometry.

By generalizing the advanced mechanistic clues involved in the initiation and maintenance of epithelial cell polarity, a recent breakthrough by the m.th é ry team disclosed a minimal epithelial model consisting of a doublet of MCF10A cells cultured on H micropatterns that fully reflected the in vivo phenotype of centrosomes in terms of localization (Burute M et al, (2017) developmental cells (DevCell.)40,168-. This study showed that MCF10A cells cultured on H micropatterns were split within 24 hours and the cell doublets obtained adopted a standardized shape with highly normalized internal tissue: centrosomes are off-center, near intercellular junctions; and the nuclear-central body axis is oriented toward the junction. After automated image acquisition and processing of centrosomes and nuclei stained with centromeric pericentrin antibody and Hoechst, respectively, quantification of cellular polarity descriptors showed very reliable results, separating polarized cells into junctions (72% in MCF 10A). Addition of the EMT inducer TGF- β can greatly reduce polarized MCF10A cells to 32%, cause a shift in polarity index, and increase the distance between nuclei and the central body. Although the disclosed descriptors can effectively detect EMT in healthy cell types in black and white, the focus of this study is not focused on assessing tumor invasion. Furthermore, the method adopted by Burute M et al is not suitable for the fine comparison of progressively aggressive cancer cells, as will be explained in the examples section.

Tumor diagnosis is subjective in nature. In fact, since the histological examination is based on the subjective and qualitative estimation of the characteristics of tissue samples damaged by surgery, it does not allow a reliable quantification of the characteristics of the cells of tumor origin, and consequently an accurate classification of the invasive potential of the tumors. In addition, tumor diagnosis is not directly related to treatment options. Therefore, there is still a strong need for a method of classifying tumors according to their invasive potential, in particular for prognosis and concomitant testing as an effective drug in predictive personalized medicine.

Disclosure of Invention

The present invention relates to a method for classifying tumors according to invasiveness and drug response. Methods based on this classification can be used for cancer diagnosis, which elucidates the invasive potential of tumor cells and outlines prognosis, and can be used as a companion test dedicated to personalized medicine to predict anti-tumor effects from molecules that reduce tumor invasion or growth.

Accordingly, the present invention relates to a method for predicting the invasive potential of a tumor cell, said method comprising:

-providing a culture device comprising a planar substrate and two adhesive patterns on said substrate for culturing at least two groups of adherent cells,

wherein the two adhesive patterns are sufficiently separated from each other by a substantially non-adhesive insertion area to prevent cells on a first adhesive pattern from reaching the other adhesive pattern; and, the area covered by the two adhesive patterns and the insertion region is sufficient to adhere two animal cells, and wherein the two adhesive patterns of each set are separated by a surrounding region of lipophobicity;

-seeding at least one tumor cell on the set of two adhesive patterns;

-culturing said tumor cells so as to obtain one cell doublet from one cell division on said set of two adhesion patterns;

-measuring at least the following three parameters on the set of two adhesion patterns for the tumor cells:

-the internuclear distance of the two adhesion patterns of each group;

-cell connectivity of the two adhesion patterns of each group; and

-a percentage of highly polarized cells to the total cell population on the set of two adhesion patterns;

-determining the invasive potential of the tumor cell based on the at least three measured parameters.

Preferably, the distance between two adhesive patterns is between about 2/3D to about 4/3D, D being the diameter of the surface S covered by the cells on the support without any constraint, and the area covered by the adhesive patterns is less than 80% of the cell surface S covered by the cells on the support without any constraint for each cell. In particular, the set of two adhesive patterns has an H-form.

Preferably, the highly polarized cells are cells with centrosomes located near the cell-cell junction in the triangular regions defined as follows:

|y|≤(x-0.5)*tan 60

wherein x is more than or equal to 0.5 and less than or equal to 2

Where x is the position of the centrosome on the nucleus-nucleus axis pointing to the intercellular junction and y is the position of the centrosome on an axis perpendicular to the nucleus-nucleus axis, in units of the radius of the nucleus.

Preferably, cell-cell connectivity is the difference between the area of the convex envelope around the cell doublet and the area of the cell doublet.

Preferably, in addition to the tumor cells to be tested, reference cells with a predetermined invasive potential are seeded on the device, cultured and the parameters measured. In particular, the reference cell is selected from the group consisting of a cell with very low or no invasive potential, a cell with low invasive potential, a cell with medium invasive potential, a cell with high invasive potential and a cell with very high invasive potential.

Preferably, the invasive potential of the tumor cell is determined by comparison with a reference cell. In particular, statistical analysis was performed to determine the reference cells with the highest similarity to the tested tumor cells. More preferably, the statistical analysis is performed by principal component analysis.

The invention also relates to a method for providing information useful for the diagnosis or prognosis of a subject, said method comprising performing a method for predicting the invasive potential of a tumor cell according to the invention.

The invention also relates to a method for assessing the efficiency of a molecule for treating a tumor in a subject, the method comprising contacting the molecule with tumor cells from the tumor of the subject, determining the invasive potential of the tumor cells treated with the molecule, comparing the determined invasive potential of the tumor cells treated with the molecule with the invasive potential of untreated tumor cells, and selecting a molecule that reduces the invasive potential, wherein the invasive potential of the tumor cells is determined by the method for predicting the invasive potential of tumor cells according to the invention.

Preferably, the tumor is a solid tumor. In particular, the solid tumor is a carcinoma. Preferably, the tumor is selected from the group consisting of breast, colorectal, lung, kidney, thyroid, prostate, liver, pancreatic, head and neck and ovarian cancers, preferably from the group consisting of breast, colon, lung, kidney and thyroid cancers.

Drawings

Fig. 1-fig. 1A and 1B present the relevant morphological descriptors and statistical programs for differentiating the invasion potential of tumor cells according to the present invention. FIG. 1A. As a structural descriptor, the percentage of highly polarized cells is quantified in a single cell as the relative coordinates of the centrosomes with respect to the centroid of the nucleus. Unpolarized centroids (gray dots) were removed and the percentage of the total number of centroids that face the junction was calculated as filtered centroids (black dots). Other descriptors used according to the invention are the internuclear distance and the cell-cell connectivity, which is defined as the difference between the area of the convex envelope around the cell doublet (dashed line) and the area of the cell doublet. Results are shown for descriptors of 13 breast cancers with progressive invasive potential. FIG. 1B. The feature vector representation of the descriptors determined from the principal component analysis of the results shows the correlation between the descriptors according to the invention.

Fig. 2-2A-2D show results of studies performed in accordance with the present invention. Fig. 2A. Scatter plots of the mean-centered data of the samples were repeated for each cell line and the eigenvectors were overlaid. Fig. 2B. The study resulted in a dendrogram showing clustering of 13 cell lines using three descriptors according to the invention. Each replicate sample was classified separately. Pearson correlation coefficient, n >300 cells analyzed. Fig. 2C. Duplicate samples were scored according to their classification. MCF10A and MDA-MB-231 were arbitrarily set as low-invasive (score 1) and high-invasive (score 4) references, respectively. Fig. 2D. The invasive potential of the 13 cell lines was then finally classified according to the mean of the replicate sample scores.

FIG. 3-comparative data-FIGS. 3A to 3C illustrate the method used by Burute M et al ((2017.) developmental cells 40,168-184), and why this method is not suitable for classifying tumors according to their invasive potential. FIG. 3A, descriptor of Burute M et al ((2017.) developmental cells 40,168-184) for the assessment of EMT in MCF10A cells: percentage of polarized cells, polarity index, inter-nuclear distance and inter-centriole distance. FIG. 3B, a dendrogram result of the study, shows clustering of 13 Cell lines using the descriptors shown in Burute M et al (2017.) (Dev Cell.)40, 168-184). Pearson correlation coefficient, n >300 cells analyzed in triplicate. Fig. 3C. The invasive potential of the 13 cell lines was subsequently classified.

Figure 4-comparative data-figures 4A and 4C illustrate the importance of the percentage descriptors of highly polarized cells to effectively differentiate the invasive potential of tumor cells. FIG. 4A, the percentage of polarized cells and the internuclear distance were quantified according to the method of Burute M et al (2017. developmental cells 40,168-184), and Cell-Cell connectivity descriptors were determined as defined in the present invention. Fig. 4B, a generated dendrogram showing clustering of repeated samples according to this method. FIG. 4C, the invasive potential of the 13 cell lines was finally classified according to this method. MCF10A and MDA-MB-231 were arbitrarily set as a low-and high-invasiveness reference, respectively.

Detailed Description

The invention provides a method for accurately classifying tumors according to the invasive potential of the tumors. The method is based on the combination of three parameters that are necessary and sufficient for classifying tumors according to their invasive potential. The three basic parameters are the internuclear distance, cell-cell connectivity, and the percentage of highly polarized cells to the total cell population over several sets of two adhesion patterns. Of course, other parameters may be combined with these three basic parameters.

An advantage of the present invention is that the results obtained from such a classification not only correspond to the physiopathology of the tumor used for diagnosis, but also provide a prediction of the degree of potential malignancy and subsequent prognosis for the patient.

Another advantage of the present invention is that the classification method enables extrapolation from the efficiency with which a compound affects the behaviour of cells cultured from a tumour to the cure or remission efficiency with which the compound treats that particular tumour.

Accordingly, the present invention relates to a method for predicting the invasive potential of a tumor cell, comprising:

-providing a culture device comprising a set of two adhesive patterns for culturing adherent cells on said substrate,

-seeding at least one tumor cell on the set of two adhesive patterns;

-culturing said tumor cells so as to obtain one cell doublet from one cell division on said set of two adhesion patterns;

-measuring the following three parameters on the set of two adhesive patterns:

-the internuclear distance of the two adhesion patterns of each group;

-cell connectivity of the two adhesion patterns of each group; and

-on said set of two adhesion patterns, highly polarized cells as a percentage of the total cell population;

-determining the aggressiveness of said tumor cells based on said three measured parameters.

Culture device

The culture device is the device disclosed in patent application WO 2010/046459. In particular, it comprises sets of adhesive patterns of specific geometries that prevent rotational movement of cells and impart them with spatially adhesive conditions, thereby achieving mechanical equilibrium and a fixed and reproducible conformation. WO2010/046459 provides rules that allow designing a device for adhering two cells in a multicellular arrangement in a mechanically stable and reproducible conformation. The rules are as follows: each cell of the multicellular arrangement has an adhesive pattern. The area between two adhesive patterns is called an insertion area. The adhesive patterns are sufficiently spaced from each other by substantially non-adhesive intervening regions to prevent cells on a first adhesive pattern from reaching another adhesive pattern. Of course, the adhesion patterns must also be close enough to allow interactions between cells. It was determined that the insertion region was about one time the cell surface diameter (D). The adhesion pattern is preferably smaller than the surface area (S) of the cells. These small adhesion patterns are important because they provide enough freedom for the cells (through the non-adhesive regions) to allow them to wedge into each other. The balance between these two constraints (the adhesion pattern and the other interacting cells) stabilizes the interacting cells and keeps them in a natural position, and therefore stable and reproducible. To provide sufficient freedom to the cells, the insertion region is substantially non-adhesive. However, the insertion region may comprise an adhesive region, more preferably a single adhesive region, to help the cells meet each other and establish an interactive contact. The width of this adhesive region of the insertion region must be narrow (less than 1/2 for the cell diameter). In fact, if the area is too large, the cells may be allowed to reach another adhesion pattern. The area of the two adhesive patterns with the insertion regions is sufficient to adhere two animal cells. Of course, to avoid interference of surrounding cells (not involved in the multicellular arrangement) with the multicellular arrangement, each set of patterns is surrounded by a region of the acellular region.

The apparatus of the present invention comprises: a plate defining a surface; and at least two sets of two adhesive patterns sufficiently spaced apart from each other by a substantially non-adhesive insertion region to prevent cells on a first adhesive pattern from reaching the other adhesive pattern, and the two adhesive patterns having insertion regions of sufficient area to adhere two animal cells. In other words, the area covered by the two adhesive patterns and the insertion region is sufficient to adhere two animal cells, and each cell adheres to one individual adhesive pattern. In particular, the plate defines a planar surface. In particular, the area covered by the two adhesive patterns and the insertion region is too large to be covered by a single cell. Therefore, the area is much larger than 1S. For example, for a multicellular arrangement of 2 cells, this area will be about 2S.

The area defined by the convex envelope of each adhesion pattern is between about 1/2S and about 3/2S, S being the area covered by the cells without any constraint. In a preferred embodiment, the area defined by the convex hull of each adhesive pattern is between about 3/4S and about 5/4S, more preferably about S. For example, S may be in the range of 1-2,500. mu.m²More preferably 1 to 1,000 μm²Still more preferably in the range of 1 to 500. mu.m²Or 500-900 μm²In the meantime. S may depend on the cell type. However, the area defined by the convex envelope of each adhesive pattern comprises a high percentage of non-adhesive areas, for example at least 20%, 30%, 40% or 50%, preferably between 20-70%, 30-60% or 40-50% of non-adhesive areas. Alternatively, the area of each cell covered by the adhesive pattern is smaller than the fine80%, 70%, 60% or 50% of the cell surface S. For example, the area of each adhesion pattern is between 30-80%, 40-70%, or 50-60% of the cell surface S. Once mechanical equilibrium is reached, the area can also be defined as the surface covered by a cell.

The convex hull of each adhesive pattern may have any kind of form (e.g., disk, square, rectangle, trapezoid, disk fragment, oval, etc.). On this convex envelope, a major axis and a minor axis can be defined.

The adhesive pattern may be formed by a single attached adhesive surface and/or several non-attached adhesive surfaces. "Single attached adhesive surface" preferably refers to a solid line or a curved line. Preferably, the "non-linked adhesive surface" preferably refers to a dotted or dashed line or curve, or discrete points or areas. In a preferred embodiment, the adhesive pattern consists of a combination of adhesive elements selected from the group consisting of lines, curves and dots. The width of the adhesive dots, lines, curves or edges is preferably 0.1 to 100 μm, more preferably 1 to 50 μm, still more preferably about 8 μm.

In a preferred embodiment, the distance between two adhesive patterns within a group is between about 2/3D and about 4/3D. In a preferred embodiment, the two adhesive patterns within a group are separated by a distance of about 3/4D to about 5/4D, more preferably about D.

In a preferred embodiment, the substantially non-adhesive insertion region comprises a single adhesive region. The single adhesive zone is adapted to not allow cells to reach another adhesive pattern. That is, the single adhesion zone must be narrow. In fact, a long and thin line of adhesion parallel to the major axis of the adhesion pattern is not suitable as it allows cells to reach another adhesion pattern. For example, the adhesive region is a region located between two adhesive patterns of a set and having a width of less than D, preferably less than 1/2D, preferably less than 1/3D, more preferably less than 1/4D. In one embodiment, the adhesive zone may connect two adhesive patterns, for example as a line or a curve. In another embodiment, the adhesive region may be between two adhesive patterns without any connection to them. While not wishing to be bound by theory, it is believed that this adhesive region helps two cells to establish cell-cell interactions. In a preferred embodiment, 50% of the substantially non-adhesive insertion region is non-adhesive, more preferably 60%, 75%, 80%, 85% or 90%. A single adhesive zone may comprise a single attached adhesive surface and/or several non-attached adhesive surfaces forming the adhesive zone.

In a most preferred embodiment, the device of the present invention is such that the distance between sets of two adhesive patterns is between about 2/3D to about 4/3D, preferably between about 3/4D to about 5/4D, more preferably about D; the substantially non-adhesive insertion region comprises a single adhesive zone located between two adhesive patterns and having a width less than D, preferably less than 1/2D, preferably less than 1/3D, more preferably less than 1/4D; and the adhesion pattern of each cell covers less than 80%, 70%, 60% or 50% of the cell surface S.

In one particular embodiment, the set of two adhesive patterns may take the form of: C. x, H, Z and Π. In a most preferred embodiment, the set of two adhesive patterns has the form of an H.

Preferably, the device comprises a plurality of sets of two adhesive patterns separated from each other by a region of the acellular region to which the cells do not adhere. More particularly, the device comprises at least two sets of two adhesive patterns, preferably at least 5, 10, 100, 1000, 10000 or 100000 sets of two adhesive patterns. In a preferred embodiment, the device comprises two adhesive patterns per cm in the set 5 to 25000²More preferably two adhesive patterns/cm in the group 5000 to 10000²And still more preferably about set 7500 of two adhesive patterns/cm²。

The adhesive pattern or region includes molecules that promote cell attachment. These molecules may be non-specific, as is oxidized polystyrene in tissue culture treated polystyrene dishes. The molecule may also be a specific adhesion molecule, such as those well known to those of ordinary skill in the art, and include antigens, antibodies, cell adhesion molecules, extracellular matrix molecules such as collagen I, collagen IV, laminin, fibronectin, synthetic peptides, carbohydrates, and the like. Preferably, the adhesive pattern comprises an extracellular matrix molecule, more preferably fibronectin.

The non-adhesive surface is an inert surface. Suitable inert surfaces are those covered with polyethylene glycol derivatives.

The plate is a support suitable for confocal, optical and/or fluorescence microscopy. In a more preferred embodiment, the plate is glass, possibly covered with a thin layer of oxidized polystyrene. Suitable plates according to the invention are, for example, coverslips or slides. The support may also be a plastic modified for cell culture. It may also be a thermoformed plastic plate or a tissue culture treated petri dish. In the present invention it is understood that the plate preferably refers to a planar surface without any holes. Alternatively, the plate may be a soft material having a Young's modulus of between 5 and 100 kPa.

The device according to the invention may comprise groups of adhesive patterns spaced apart from each other on the same plate, so that each group may be incubated in a different medium. For example, one group of sets of adhesive patterns may be contacted with one test compound, while another group may be contacted with another test compound or not. This isolation may be provided by a physical barrier such as a teflon seal. See, for example, SPI Supplies' SPI

Of Aname

The slide is printed.

The device according to the invention with the adhesive pattern and the areas as well as the acellular areas are formed by micropatterning. Microcontact printing or UV patterning may be used. Standard methods are well known to those skilled in the art. For a review see Whitesides et al (annu. rev. biomed. eng.,2001, pages 335-373, more particularly pages 341-345).

Tumor cells

The tumor cell tested may be any tumor cell. Preferably, the tumor cells may be obtained from a subject having a tumor. Tumor cells can be obtained from a biological sample of a subject containing tumor cells. Examples of such biological samples include biopsies, organ, tissue or cell samples. Preferably, the biological sample is a biopsy sample.

The subject is preferably a mammal. It may be a human or an animal. The animal may be a pet. For example, the subject may be a dog, cat, horse, cow, pig, sheep, and non-human primate, and the like.

Any kind of tumor cells can be used in the present invention. The cells may be, for example, fibroblasts, endothelial cells and epithelial cells. In a preferred embodiment, the cell is an epithelial cell.

For example, the tumor cells may be from a solid or hematopoietic tumor. Preferably, the tumor cell is from a solid tumor. Examples of solid tumors are sarcomas, carcinomas and lymphomas. In a preferred embodiment, the solid tumor is a carcinoma.

The tumor may be selected from the group consisting of breast, colorectal, lung, kidney, thyroid, prostate, liver, pancreatic, head and neck and ovarian cancers, preferably from the group consisting of breast, colon, lung, kidney and thyroid cancers.

Thus, the method may comprise a prior step of providing tumor cells from the subject.

Reference cell

In addition to the tumor cells to be tested, reference cells can also be used in parallel on the culture device.

The reference cell is a cell for which invasion potential has been previously established/determined. These cells were used to determine the level of invasiveness. For example, a first reference cell may have a low invasive potential, while a second reference cell may have a high invasive potential. Optionally, several reference cells may be used, for example cells selected from: cells with very low or no invasive potential, cells with low invasive potential, cells with moderate invasive potential, cells with high invasive potential and cells with very high invasive potential. In a preferred embodiment, at least four reference cells are used in the method: cells with very low or no invasive potential, cells with low invasive potential, cells with moderate invasive potential, cells with high or very high invasive potential.

For example, cells that have no or very low invasive potential may be non-tumor cells (e.g., MCF10A or RPE 1). Cells with low invasive potential may be, for example, HCC1143, Hs578T, MDA-MB-468, HCC38, HCC 70. Cells with intermediate invasive potential may be, for example, BT-549, MDA-MB-157, BT-20, HCC 1937. Cells with high invasion potential may be MDA-MB-231 or MDA-MB-436. In particular, the MDA-MB-231 cell line is referred to as a very highly invasive cell.

The invasive potential of a reference cell can be determined by any method and classification available to those skilled in the art. Invasive potential can be determined according to the Nottingham histological scoring system (Elston-Ellis modified version of the Scarff-Bloom-Richardson grading System) or the WHO/IARC classification. Invasive potential refers to the ability of a cell to invade adjacent or distant tissues. The motility of the cells can be measured, for example, according to the gold colloid method. The gold colloid method is a method in which cells are seeded on a cover glass to which colloidal gold is adhered, and the area of a gold-deficient portion formed on the movement locus thereof is calculated, thereby measuring the motility of the cells. The motility and invasive potential of cells can be assessed by Boyden chamber assay or the like. In the Boyden chamber assay, the motility is first assessed by the number of cells passing through the uncoated porous PET film, then using the PET film coated with a basement membrane, measuring the number of cells that have degraded the basement membrane, have penetrated through the pores and migrate to below the membrane, and determining the ratio of the number of cells that have migrated, thereby assessing the invasive potential of the cells.

Alternatively, the invasive potential of the cells can be measured by time-lapse microscopy of migrating cells on a 2D or 3D matrix, a scratch/wound healing assay or a soft agar invasion assay.

Culturing, fixing, staining and image acquisition and processing

The method comprises the following steps: a step of seeding at least one tumor cell on the set of two adhesive patterns; and culturing said tumor cells to obtain a cell doublet from one cell division on said set of two adhesion patterns. More particularly, tumor cells are cultured on the device for a time sufficient for the tumor cells to divide into two cells on the set of adhesive patterns to adhere to each of the adhesive patterns and reach mechanical equilibrium. The time period may depend on the tumor cells and is well known to those skilled in the art. The incubation time is usually 1 to 2 days. Generally, a 24 hour period is sufficient.

In a preferred embodiment, several reference cells with different invasive potential levels are also seeded and cultured on the culture device simultaneously with the cellular tumor to be tested.

For example, cells can be seeded at a density of 10,000 to 100,000 cells per square millimeter chip, e.g., about 25,000 cells per square millimeter chip. The method may comprise a rinsing step to remove unattached cells.

At the end of this incubation time, the cells may be fixed by any method known in the art, for example by using cold methanol. The cells were then stained to visualize centrosomes and nuclei. For example, centrosomes are immunostained with antibodies against centrosomal protein (centrin), pericentrin, gamma-tubulin, neiin, Cep164, Cep192, Odf1, cenexin, centriolin and/or PLK4, preferably pericentrin. For example, nuclei can be stained by Hoechst labeling or DAPI (4', 6-diamidino-2-phenylindole).

Images of the cells on the set of two adhesion patterns are then collected with any suitable microscope. Image acquisition may be automated.

Image analysis the internal organization of the cells was monitored from a doublet. It preferably includes binarization of nuclei and centrosomes, calculation of the centroids of the nuclei and centrosomes, calculation of the length and orientation of the nucleus-nucleus vectors of the doublets and the nucleus-centrosome vectors within individual cells of the doublets, as well as the area of the whole cell doublet and its convex envelope.

In a preferred embodiment, the image processing comprises personalization of the set of two adhesion patterns, for example using a template matching method; detection of nuclei and centrosomes, e.g., threshold and size filtering within each cell based on doublets; for each doublet, the coordinates of the centrosome and the nucleus centroid, as well as the area of the entire cellular doublet and its convex envelope, were calculated.

In a preferred embodiment, at least 50 cell doublets per cell type are required for analysis. However, the method can be carried out using 50 to 1000 cell doublets, preferably about 100 cell doublets, for each type of cell.

Then, three parameters necessary and sufficient to classify tumor cells according to their invasive potential, namely the internuclear distance, the cell-cell connectivity and the percentage of highly polarized cells to the total cell population, were measured. These parameters are measured against the tumor cells to be tested and the reference cells.

Distance between nuclei

The internuclear distance is the length between the nuclei of two cells on the set of two adhesion patterns. More specifically, the internuclear distance is the length between the nuclear centroids of two cells of the doublet.

Degree of cell-cell connectivity

Cell-cell connectivity is defined as the difference between the area of the convex envelope around the cell doublet and the area of the cell doublet. The cell-cell connectivity is shown in fig. 1A. The dotted line is the convex envelope around the cell doublet. The gray line defines the area of the cell doublet, and the area of the slashed line is the difference between the area of the convex envelope around the cell doublet and the area of the cell doublet. The average area of cell doublets is typically 1100. mu.m²。

Percentage of highly polarized cells

As shown in fig. 1A, cellular polarization is determined based on the location of centrosomes relative to the nucleus-nucleus axis and the vertical axis. Preferably, the orientation of the nucleus-centrosome vector is measured relative to the nucleus-nucleus axis pointing to the intercellular junction. The nucleus-centrosome distance is normalized to the nucleus radius. Thus, the positive coordinate corresponds to the nuclear-centrum axis pointing to the adjacent cell, while larger values correspond to centrum positions that are highly off-center.

Highly polarized cells are quantified as the relative coordinates of the centrosome with respect to its nucleus on the nucleus-nucleus reference axis, where the maximum radius of the nucleus is an arbitrary unit.

Highly polarized cells are then defined as cells with centrosomes located near the cell-cell junction in the triangular regions defined as follows:

|y|≤(x-0.5)*tan 60

wherein x is more than or equal to 0.5 and less than or equal to 2

Where x is the position of the centrosome on the nucleus-nucleus axis pointing to the intercellular junction and y is the position of the centrosome on an axis perpendicular to the nucleus-nucleus axis, in arbitrary nucleus radii. Thus, 1 corresponds to the length of the radius of the nucleus.

For example, in FIG. 1A, the dashed lines define regions corresponding to highly polarized cells.

The percentage of highly polarized cells to the total cell population is the number of cells with high polarization as defined above divided by the total number of cells tested for each cell (e.g., tumor cells tested, reference cells). More specifically, it is the ratio of the number of cells with high polarization divided by the total number of cells x100 as defined above.

The percentage of highly polarized cells indicates the proportion of cells in the tumor that are less invasive. It is inversely proportional to the invasive potential of the tumor.

Additional descriptors

To determine the invasive potential of a cellular tumor, additional descriptors or parameters may be further considered. For example, such additional parameters may be selected in the group consisting of: nuclear-centrosome distance, nuclear diameter, ratio of nuclear-centrosome distance to nuclear diameter, duplex area, angle between nuclear-centrosome and nuclear-nuclear axis, duplex compactness (ratio of convex envelope to duplex area), duplex compactness variation coefficient, duplex intracorporeal centrosome-centrosome distance, polarity index (especially as defined in Burute et al, 2017; see also fig. 1A), and percentage of cells with complete polarity reversal relative to total population.

Determination of invasiveness

Based on the at least three parameters, the invasive potential of the tested tumor cells can be determined, in particular by comparison with reference cells having a predetermined invasive potential.

In a first aspect, as shown in fig. 1B and 2A, each cell may be separately mapped in a feature vector representation based on the at least three parameters.

In a second aspect, statistical analysis is performed to determine the reference cells with the highest similarity to the tested tumor cells. Preferably, the statistical analysis is performed by principal component analysis. In this analysis, each parameter or descriptor may be weighted to adjust its importance.

Statistical analysis allows the determination of scores. Statistical analysis has led to the classification of different classes of invasive potential. In a particular embodiment, the method allows for the classification of tumor cells into one of four classes, which are 1) no or very low invasive potential, 2) low invasive potential, 3) moderate invasive potential, and 4) high invasive potential.

Diagnostic use

Determination of the invasive potential of a cellular tumor can be used for diagnostic purposes, particularly in the field of oncology. Thus, the method according to the invention may be used for diagnostic or prognostic purposes. In particular, it may be used to classify a tumor in a subject.

Use for predicting drug efficacy

The method according to the invention can be used to assess the efficacy of a molecule in treating a tumor in a subject. Indeed, it can be used to measure the ability of molecules to reduce the invasive potential of tumor cells. Molecules are also considered to be combinations of different molecules.

Then, when performing the method for assessing the efficiency of a molecule to alter the invasive potential of a tumor cell, the tumor cell may be contacted with the molecule before seeding on the set of two adhesion patterns, after seeding on the set of two adhesion patterns, or before and after seeding. In addition, in this case, one group of tumor cells is contacted with the molecule, while the other group is not (control group). Optionally, another population of tumor cells can be contacted with a molecule known to reduce the invasive potential of the tumor cells (as a positive control). Optionally, the tumor cells can be contacted with a combination of molecules or with different molecules (one population of tumor cells per molecule).

The method may further comprise the steps of: selecting the molecule as being effective for treating a subject if the molecule reduces the invasive potential of tumor cells from the subject.

Thus, a method for assessing the efficiency of a molecule for treating a tumor in a subject comprises contacting the molecule with tumor cells from the tumor of the subject, determining the invasive potential of the tumor cells treated with the molecule, comparing the determined invasive potential of the tumor cells treated with the molecule to the invasive potential of the untreated tumor cells, and selecting a molecule that reduces the invasive potential.

In another aspect, the invention also relates to a method for assessing the efficiency of a molecule for treating a tumor in a subject, wherein said method comprises determining the invasive potential of a tumor cell of said patient by a method according to the invention, and selecting a molecule known to be effective against tumor cells classified in the same class of invasive potential as the tumor cell of said patient.

Definition of

By "invasive potential" is meant the ability of a cell to invade adjacent or distant tissues.

By "convex envelope" is meant the smallest convex polygon containing the set of two adhesive patterns.

By "area defined by the convex envelope" is meant the area covered by the region contained in the convex envelope.

"S" is intended to mean a surface covered, without any constraints, by eukaryotic cells on a support (e.g., cells on petri dishes, plastic or glass coverslips).

"D" is intended to mean the diameter of the disc having said surface S.

"distance between two adhesive patterns" means the distance between two closer points of two adhesive patterns.

By "about" is meant a value greater than or less than 5%.

"doublet" means two cells immobilized on two adhesive patterns in a set.

"descriptor" means a parameter that quantifies a characteristic of the cell being analyzed.