Methods and tools for detecting effects of compounds on protein, RNA or DNA interactions

Abstract

The present invention encompasses an in vitro method for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more baits and one or more candidate preys in eukaryotic cells. In addition, the present invention encompasses compounds for use as medicaments identified by an in vitro method according to any of the preceding claims.

Description

[0001] This invention covers an in vitro method for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more candidate preys in eukaryotic cells. Furthermore, this invention covers compounds used as pharmaceuticals identified by the in vitro method according to any one of the preceding claims. Technical Field

[0002] This invention relates to methods, uses, and tools for studying ligand interactions in eukaryotic cells and the effects of compounds on said interactions, preferably for identifying new drug candidates. Background Technology

[0003] Protein-protein interactions (PPIs) are essential for cellular function and are present in all types of organisms, with an estimated >500,000 interactions in humans (Li et al., 2017; Rolland et al., 2014). PPIs are involved in the regulation of biological pathways, consistent with their involvement in the development of many diseases. Therefore, they represent attractive targets for developing novel therapeutics for a wide range of human diseases (Lu et al., 2020; Scott et al., 2016). Most PPI interactions were initially considered undrugable due to their large interaction surface area, but have since become promising potential targets with the identification of several effective PPI inhibitors associated with human diseases (Arkin and Wells, 2004; Arkin et al., 2014).

[0004] Targeting RNA:protein interactions (RPIs) that are key to pathological mechanisms is also a promising strategy for discovering novel classes of drug candidates that remain largely unexplored (Einstein, JM et al., 2021). Cellular RPIs are highly diverse, encompassing interactions with messenger RNA (mRNA) (Baltz, AG et al., 2012), ribosomal RNA (rRNA) (Simsek, D. et al., 2017), and non-coding RNA (ncRNA) (Lu, S. et al., 2019), which are crucial for fine-tuning spatiotemporal gene expression. As revealed by genomic approaches (VanNostrand, EL et al., 2020 and Castello, A. et al., 2012), the human genome contains over 1000 transcripts encoding RNA-binding proteins (RBPs), providing a wide range of interactions with both coding and non-coding RNAs. However, while the diversity of RNA:protein interfaces allows for the development of RPI repressive molecules (Wu, P. 2020), only a small number of studies have been conducted and are limited to a few complexes, such as Lin28-let7 interactions (Roos, M. et al., 2018 and Wang, L. et al., 2018), MSI-RNA (Minuesa, G. et al., 2019), and hnRNP A18-RNA (Solano-Gonzalez, E. et al., 2021).

[0005] As mentioned, targeting PPIs and RPIs for drug discovery is an attractive strategy for the pharmaceutical industry. However, despite years of extensive research into developing new PPI or RPI modulators, only a few small molecules have reached the market. While the large surface area of ​​PPIs and RPIs may hinder drug development, the availability of methods for accurately assessing the effectiveness of candidate modulators in the cellular environment using high-content screening approaches is lacking.

[0006] Especially regarding PPIs, the need for methods for large-scale screening of putative PPI inhibitors is particularly urgent due to the aforementioned obstacles in the discovery of PPI inhibitors. Several in vitro and cell-based methods have been developed to this end (Arkin et al.; Titeca et al., 2019; Wade and Arkin). One of the most popular assays for screening small molecules targeting specific protein-protein interactions is homogeneous time-resolved fluorescence (HTRF) (Degorce, 2009). This highly sensitive technique, based on energy transfer between donor and acceptor molecules, has been adapted for automation in HTS. However, a major drawback of HTRF, as with other in vitro methods, is the necessity of purifying the protein of interest, which is often limited to the use of specific domains, and their lack of information about the fate of small molecules and their efficiency in inhibiting PPIs in the cellular environment. While HTRF methods can still monitor protein-protein interactions in cells, the adaptation of HTRF to cell-based assays requires subsequent steps of protein labeling with antibodies or chemical bonds, which reduces the interest in using this method in cell-based assays (Degorce, 2009). While FRET technology is primarily used to assess interactions between two proteins in cells, its lack of sensitivity and low signal-to-noise ratio make it unsuitable for accurately measuring disruptions of interactions in the cellular environment (Piston and Kremers, 2007). On the other hand, with the recent development of nanoluciferases, BRET, still based on energy transfer but using bioluminescence instead of fluorescence, has become a subject of renewed interest. This small, bright, and stable donor greatly expands the potential applications of BRET assays through its use (Dale et al., 2019; Machleidt et al., 2015). NanoBRET technology is based on the expression of the full-length protein labeled with the bright luciferase Nanoluc, which acts as the energy donor, and a fluorophore called HaloTag acts as the acceptor. The reading is the average luminescence signal per well recorded on a microplate reader. This technology can only detect PPIs in the native cellular environment when the donor and acceptor probes are sufficiently close to each other to allow energy transfer. Furthermore, even with good sensitivity, the reproducibility of NanoBRET assays depends on donor expression, as the BRET signal is dependent on the donor / receptor ratio, as described in donor saturation assays (Couturier and Deprez, 2012). To overcome this limitation and ensure the stability and reproducibility of the BRET signal, stable cell lines expressing the donor are preferred for developing BRET-based screening assays, which increases the complexity of establishing such assays (Couturier and Deprez, 2012).Therefore, there is a need for reliable and simple cell-based assays that can be used to screen small molecule regulators of protein-protein interactions.

[0007] Regarding RPi, previous methods for finding effective inhibitors of the RNA:protein interface have employed in vitro assays such as fluorescence polarization assays, supplemented by pull-down assays of cell lysates or RNA ELISA to test the effectiveness or selectivity of small hits (Roos, M. et al., 2016 and Minuesa, G. et al.). While in vitro methods are important for defining putative hits and validating effective compounds, deciphering the effectiveness of selected molecules in a cellular context often relies on indirect measurements using techniques such as cell binding thermal migration assays (CETSA) or functional assays, where the putative consequences of RPi disruption on cellular function are subject to considerable uncertainty. In fact, multiple functions are associated with RBPs, making the interpretation of functional assay results challenging. Furthermore, toxicity and off-target effects are putative biases that are always difficult to eliminate, especially when using K with a low micromolar range. D For small molecules, this is generally the case for RPI inhibitors (Jung, J. et al., 2013 and Camborde, L. et al., 2017). To bridge the gap between in vitro and functional assays, cellular methods originally used to detect PPIs, such as FRET (fluorescence resonance energy transfer) or PLA (ortho-linked assay), have been adapted to detect RPIs in cells (Jung, J. et al., 2013 and Camborde, L. et al., 2017), but several technical issues have hindered their application, such as the need for RNA adaptors in FRET and PLA, the proximity of donor and acceptor proteins in FRET, and the use of antibodies in PLA.

[0008] In addition to assays that detect RPI in cells, the drug discovery process presents other challenges, such as the quality of computational models, the strategies used in computer screening, and the lack of experimental feedback and validation of computationally predicted inhibitors, which is essential for directing a rational drug design process toward the most relevant molecules.

[0009] The patent family WO2016012451, filed by the same inventors of this invention, relates to methods for detecting interactions between one or more protein decoys and one or more candidate preys in eukaryotic cells. These methods utilize polymeric tubulin binding portions, such as microtubule-binding domains (MBDs), fused to one or more decoys as tools for determining the occurrence of such interactions between the one or more decoys and one or more candidate preys in eukaryotic cells.

[0010] Starting with the technology described in this patent family, the applicant has developed an innovative new technique that allows for the assessment of the ability of compounds to disrupt or stabilize interactions between DNA, RNA, and / or proteins, and in particular, allows for the assessment of the ability of compounds to disrupt or stabilize PPIs and RPIs by single-cell fluorescence microscopy on a high-content screening (HCS) imager. This technique allows for the screening of molecular libraries for the identification of drug candidates.

[0011] This technology addresses the aforementioned challenges by introducing experimental assays applicable to HCS to score interactions in cells (especially RPI or PPI) and by introducing drug screening methods that integrate chemical, structural, and cellular data from experimental techniques to identify and develop molecules (especially small molecules) that target interactions (especially RPI or PPI). Summary of the Invention

[0012] The present invention relates to a method for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys (preferably proteins) and one or more candidate preys in eukaryotic cells, and more particularly in the living eukaryotic cellular environment.

[0013] The system consists of a detection system based on the expression of the following substances in cells: (i) a "decoy", a known protein, ribonucleic acid or deoxyribonucleic acid, which is brought to the microtubules by its fusion with the microtubule-binding domain, and (ii) the prey, which may be, for example, a protein, nucleic acid or deoxyribonucleic acid.

[0014] When the prey interacts with the decoy, the prey also positions itself along the microtubules, making it easily detectable. As developed in this invention, it allows for the determination of differences in interaction before and after contact with the test molecule and the identification of molecules of interest in various pathologies, such as cancer, viral or bacterial infections, epigenetic diseases, or protein disorders, in cases involving PPIs and RPIs.

[0015] Here, the system of WO2016012451 (MT bench (MT BENCH)) was improved to fill the gap between in vitro assays and functional assays by probing whether the interaction between the selected bait and the selected prey is affected by the compound in the cellular environment.

[0016] The present invention also relates to baits, carriers, cell lines, kits suitable for the said method, uses, and compounds identified according to the method of the present invention.

[0017] Therefore, a first aspect of the present invention is to provide an in vitro method for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more candidate preys in eukaryotic cells, the in vitro method comprising the following steps:

[0018] a. Providing eukaryotic cells expressing (i) one or more baits and (ii) one or more candidate prey, wherein the bait comprises a bait portion and a polymerized tubulin-binding portion;

[0019] b. Detect the interaction by determining the presence of an interaction between the one or more decoys and the one or more candidate preys in the eukaryotic cell, wherein the decoys bind to tubulin in the eukaryotic cell, thereby causing the one or more candidate preys to be localized along the tubulin;

[0020] c. Contact the eukaryotic cells with the compound;

[0021] d. As in step b., determine the occurrence of an interaction between the one or more decoys and the one or more candidate preys in the eukaryotic cell, and quantify such an interaction;

[0022] e. Compare the occurrence of the interactions generated in steps b and d;

[0023] f. Infer whether the compound acts to disrupt or stabilize the interaction between one or more decoys and one or more candidate prey.

[0024] In a preferred embodiment, the one or more baits are proteins.

[0025] In a preferred embodiment, one or more candidate prey are proteins or ribonucleic acid.

[0026] A second aspect of the invention relates to the use of a polymeric tubulin-binding portion fused with one or more decoys as a tool for evaluating the ability of a compound to disrupt or stabilize the interaction between the one or more decoys and one or more preys in a eukaryotic cell, wherein the decoy binds to the polymeric tubulin in the eukaryotic cell, thereby positioning the one or more candidate preys along the polymeric tubulin, and the one or more decoys and the one or more candidate preys come into contact with the compound, thereby allowing evaluation of the disruption or stabilization of the interaction.

[0027] A third aspect of the invention relates to a protein decoy comprising a polymeric tubulin-binding portion and a decoy portion, the polymeric tubulin-binding portion comprising one or more microtubule-binding domains.

[0028] A fourth aspect of the present invention relates to a ribonucleic acid decoy comprising a polymeric tubulin-binding portion and a decoy portion, the polymeric tubulin-binding portion comprising one or more microtubule-binding domains.

[0029] A fifth aspect of the present invention relates to a deoxyribonucleic acid (DNA) decoy comprising a polymeric tubulin-binding portion and a decoy portion, the polymeric tubulin-binding portion comprising one or more microtubule-binding domains.

[0030] A sixth aspect of the invention relates to a carrier containing an expression cassette suitable for expressing the protein decoy of the invention.

[0031] A seventh aspect of the invention relates to a carrier containing an expression cassette suitable for expressing the RNA decoy of the invention.

[0032] The eighth aspect of the present invention relates to a carrier containing an expression cassette suitable for expressing the deoxyribonucleic acid decoy of the present invention.

[0033] A ninth aspect of the invention relates to a carrier containing an expression cassette suitable for expressing the protein prey of the invention.

[0034] The tenth aspect of the present invention relates to a carrier containing an expression cassette suitable for expressing the RNA prey of the present invention.

[0035] The eleventh aspect of the present invention relates to a carrier containing an expression cassette suitable for expressing the deoxyribonucleic acid prey of the present invention.

[0036] The twelfth aspect of the invention relates to cell lines transfected with the vector of the invention and / or stably expressing baits suitable for the methods and uses of the invention.

[0037] The thirteenth aspect of the present invention relates to cell lines that are transfected with the vectors of the present invention and / or stably express prey suitable for the methods and uses of the present invention.

[0038] The fourteenth aspect of the invention relates to a kit for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more preys in cells, the kit comprising decoys, carriers and / or cell lines suitable for the methods and uses of the invention.

[0039] The fifteenth aspect of the invention relates to a compound identified by an in vitro method according to the invention, which is used as a drug. Attached Figure Description

[0040] [ Figure 1The principle of steps a and b of the method: a high-throughput analysis and detection system. Applicable to the interaction of endogenous mRNA with YB-1 (an mRNA-binding protein). (A) Cells cultured in microplates were transfected using an automated system. (B) After image acquisition, a “microtubule-like pattern” was identified in the fluorescence channel corresponding to the prey (messenger RNA), confirming the interaction between the decoy (YB-1-GFP-tau) and the prey (mRNA). (C) (Top; from left to right) Detection of the decoy: YB-1-GFP-Tau; Detection of the prey: mRNA; An example of detecting protein-RNA interaction shows a microtubule-like pattern in the Cy3 channel (corresponding to the Poly-T RNA Cy3-labeled probe for detecting mRNA). (Bottom) YB-1-GFP was used as a control and was evenly distributed in the cytoplasm.

[0041] [ Figure 2 The principles of the methods for cell fixation and for endogenous prey in steps a and b are as follows: Endogenous prey delivered to the microtubules is detected in two steps. The first step (step 1) involves transfecting an expression vector encoding a protein decoy containing a microtubule-binding domain and optionally a fluorescent label. The second step (step 2) involves detecting the endogenous prey using reagents such as antibodies against the native prey or hybridization with nucleic acids. The prey is then detected on the microtubules after interaction with the decoy.

[0042] [ Figure 3 The principles of the methods for steps a. and b., used for live cells and for exogenous prey, are as follows: Step 1 involves co-transfection of two expression vectors; the first plasmid expresses a protein decoy containing a microtubule-binding domain and optionally a fluorescently labeled protein. The second plasmid expresses an exogenous prey optionally labeled with a fluorescent marker. Step 2 involves detecting the exogenous prey along dynamic microtubules or in the cytoplasm using a fluorescent marker specific to the prey and / or an antibody.

[0043] [ Figure 4[High-content screening conditions for a case study of PPI (p53 / MDM2) in a cellular context were developed using a microtubule scaffold (MT scaffold). (A) Schematic diagram of the MT scaffold technique for detecting PPI regulation with small molecules. The decoy (p53) is covalently bound to both RFP (red fluorescent protein) and MBD (microtubule-binding domain). Thus, the RFP signal follows the microtubule network. The prey (MDM2) is bound to GFP (green fluorescent protein) but is free in the cytoplasm. The identical microtubule network shown by both signals implies that the decoy and prey interact and can be modulated by inhibitors such as RG7112. This interaction can be monitored by HCS (Opera Phenix from Perkin Elmer), which generates a large number of single-cell events per well (represented here as 10% of the surface of 96 wells).] (B) Cell images of fixed U-2 OS cells overexpressing p53-FL_C_RFP-MBD (used as a bait protein) for the positive control (control+), or overexpressing the triple mutant p53-F19AW23AL26A_RFP-MBD (used as a bait protein) and overexpressing MDM2-FL_C_GFP (used as a prey protein) for the negative control (control-). The control+ image shows bait / prey interaction (both fluorescence signals show the same microtubule network), while the control- image shows no interaction (neither fluorescence signal shows the same cell compartment), in which case MDM2 is localized to the nucleus. Inhibition of the p53 / MDM2 interaction was induced by treatment with gradually increasing concentrations of RG7112 (a well-known MDM2 inhibitor (Vu et al., 2013)) ([0.05 μM], [0.5 μM], [5 μM] and [50 μM]), demonstrating disruption of the p53 / MDM2 interaction and dose-dependent relocation of MDM2 in the nucleus.

[0044] [ Figure 5The effect of point mutations at p53 residues at their MDM2 binding sites was determined using the MT bench technique. The PCC scores of wild-type p53 and different mutant constructs were normalized for comparison as decoys with single mutations (p53-F19A_C_RFP-MBD, p53-W23A_C_RFP-MBD), double mutations (p53-F19AW23A_C_RFP-MBD, p53-W23AL26A_C_RFP-MBD), and triple mutations (p53-F19AW23AL26A_C_RFP-MBD) to allow for comparisons between experimental plates (Kevorkov and Makarenkov, 2005). PCC values ​​were normalized relative to the control pair using the mean of the negative control wells and the mean of the positive control wells corresponding to the p53-FL / MDM2 interaction, as well as the negative control wells with triple mutants.

[0045] [ Figure 6 This describes the workflow of the analysis process. The analysis used in this study was conducted using Harmony. ® It was developed with R software. The different stages of the process are: 1) cell biology, 2) image acquisition, 3) image segmentation, 4) image processing, 5) colocalization, and 6) data analysis.

[0046] [ Figure 7 Development of screening criteria. (A) Z' factor graphical representation, showing the robustness of the system developed in this paper on a 384-well plate format. Z' factors of p53 and MDM2 proteins were used as positive controls (p53-FL_C_RFP-MBD / MDM2-FL_C_GFP). (B) Immobilized U-2 OS cells, overexpressing p53-FL_C_RFP-MBD (used as bait protein, red) and MDM2-FL_C_GFP (used as prey protein) for positive controls (control+), or overexpressing G3BP1_GFP (used as prey protein) for negative controls (control-). In the positive controls, p53 / MDM2 interacted, as shown in the merged channels (both fluorescent signals showed the same microtubule network). Conversely, in the negative controls, p53 / G3BP1 did not interact. The G3BP1 protein diffused in the cytoplasm and did not relocate to the microtubule network (neither fluorescent signal showed the same microtubule network). Cells were acquired using an HCS microscopy system (Opera Phenix from PerkinElmer) with 40× water immersion (1.1 NA) objectives and confocal mode. (C) Z' factor graphical representation using non-P53-associated protein G3BP1 as a negative control. Plotting and calculations were performed using R software.

[0047] [ Figure 8The conditions for testing in the MT bench assay were as follows: U-2 OS cells were co-transfected with MDM2-FL-GFP and P53 mutants and treated with RG7112 small molecules under certain conditions.

[0048] [ Figure 9 The relative enrichment of MDM2 into microtubules was determined based on different p53 mutants. An asterisk indicates that MDM2 does not interact with p53 on microtubules under these conditions, preventing the measurement of relative enrichment. Enrichment was normalized using min-max normalization (with positive and negative controls as max and min) to allow for comparisons between conditions.

[0049] [ Figure 10 Cell images from MT bench technology. Two constructs of YB1-FL fused with GFP were used as bait and mRNA (Cy3-tagged) was used as prey. In both cases, the mRNA was located in microtubules.

[0050] [ Figure 11 Relative enrichment of mRNA into microtubules in cells transfected with two YB1-FL constructs. Enrichment was calculated as defined in Materials and Methods. *** indicates a p-value of 1.44e-5. Error bars represent the standard error calculated from 16 replicates. The negative control corresponds to the condition in which cells were transfected with GFP fused to MBD. No RNA binding occurred under these conditions.

[0051] [ Figure 12 Relative enrichment of mRNA into microtubules in cells expressing different DNA constructs. Enrichment values ​​were obtained in 384-well plates. The negative control corresponds to the condition in which cells were transfected with GFP fused to MBD. Error bars represent the standard error calculated from 16 replicates. p-values ​​are as follows:

[0052] -p value (MBD-GFP-YB1-FL / MBD-GFP-YB1-deltaCSD) = 8.59e-7

[0053] -p value (YB1-FL-GFP-MBD / YB1-deltaCSD-GFP-MBD) = 9.13e-20

[0054] [ Figure 13 The dose-response curves show that as the concentration of C8 increases, the mRNA:YB1 interaction is inhibited, thus allowing determination of the activity of C8 in the cellular environment (IC50).

[0055] definition

[0056] According to the present invention, "evaluating the ability of a compound to disrupt or stabilize an interaction" means that the method is able to determine whether the compound can affect the interaction between one or more decoys and one or more candidate preys. Specifically, the method allows for the determination of whether the compound disrupts or stabilizes such an interaction.

[0057] According to the present invention, the "decoy" comprises at least one "decoy portion" and at least one "polymerized tubulin binding portion".

[0058] According to the present invention, the “decoy portion” is a portion that is readily able to interact with one or more candidate prey in the cell.

[0059] According to the present invention, the expression "comprising" also includes "consisting of".

[0060] The “polymeric tubulin-binding moiety” comprises a peptide, protein, or nucleoprotein capable of specifically binding to polymeric tubulin. Preferably, the “polymeric tubulin-binding moiety” binds more specifically to polymeric tubulin than to non-polymeric tubulin, including its monomeric form and / or its heterodimeric form (corresponding to α / β tubulin heterodimers). Ideally, the polymeric tubulin-binding moiety binds primarily or even only to polymeric tubulin (including microtubules). The “polymeric tubulin-binding moiety” can ideally be selected from known antibodies, microtubule-binding proteins, or fragments thereof, including (non-exhaustively) microtubule stabilizers and destabilizers, molecular motors, microtubule-cleaving proteins, and terminal tracking proteins such as positive terminal tracking proteins.

[0061] Microtubules are components of the cytoskeleton, distributed throughout the cytoplasm. They are part of a structural network within the cell's cytoplasm called the "cytoskeleton." The primary function of the microtubule cytoskeleton is mechanical support, but microtubules also participate in many other processes. Therefore, microtubules are only one part of the so-called "microtubule cytoskeleton," which also includes related proteins such as microtubule-associated proteins (MAPs) and other structural components such as centromeres.

[0062] "Polymerized tubulin" or "polymerized tubulin" refers only to the assembly of monomeric tubulin molecules with different polarities in a regular manner, or alternatively, the assembly of tubulin heterodimers. Tubular polymers of tubulin can grow up to 50 micrometers in length, with an average length of 25 μm, and are highly dynamic. The outer diameter of microtubules is typically about 24-25 nm, while the inner diameter is about 12 nm. They are found in eukaryotic cells and are formed through the polymerization of dimers of two globular proteins, α-tubulin and β-tubulin. Therefore, the term "polymerized tubulin" encompasses microtubules.

[0063] Therefore, "microtubule" represents a specific rearrangement of "polymerized tubulin," which physiologically occurs in eukaryotic cells and forms the "microtubule cytoskeleton" with another pair of partners. The physiological assembly of microtubules is generally described as the first step in the regulated assembly of α-tubulin and β-tubulin heterodimers, which together form polarized protofilaments. These protofilaments are then believed to assemble as cylinders into so-called microtubules. Thus, microtubules are generally described as polymers of α-tubulin and β-tubulin dimers composed of 13 protofilaments assembled around a hollow core. However, it should be noted that so-called microtubules with varying numbers of protofilaments, such as those with 14 or 15 protofilaments, have also been described in the art. However, the physiological significance of such variations or "protofilament transitions" remains unclear.

[0064] Tubulin is one of several members of a small family of globular proteins. The tubulin superfamily includes five distinct families: α-tubulin, β-tubulin, γ-tubulin, δ-tubulin, and ε-tubulin, as well as a sixth family (ζ-tubulin) found only in zooplastic protozoa. The most common members of the tubulin family are alpha-tubulin (a-tubulin) and beta-tubulin (β-tubulin), which are the proteins that make up microtubules. The microtubule terminus corresponding to β-tubulin is called the positive terminus. The microtubule terminus corresponding to α-tubulin is called the negative terminus.

[0065] Therefore, the "polymerized tubulin binding portion" of the present invention may include α-tubulin and / or β-tubulin binding portions and / or combinations thereof, and preferably tubulin binding portions in their polymeric form rather than their monomeric form or heterodimeric form.

[0066] As is known in the art, the “heterodimeric form” of tubulin corresponds to the α / β tubulin heterodimer. Therefore, the polymeric form of tubulin also corresponds to the polymer of the heterodimer, which in turn corresponds to more than one heterodimer of α / β tubulin.

[0067] For reference, human α-tubulin has the sequence SEQ ID N°1.

[0068] For reference, human β-tubulin has the sequence SEQ ID N°2.

[0069] First, microtubules provide a large surface area within cells due to their size (25 nm in diameter and one-tenth of a micrometer in length). If we consider the 10 μm long microtubules in typical mammalian cells such as HeLa cells, the microtubule surface area is greater than 30 μm², and can be even larger in cells such as neurons or muscle cells. Therefore, such a large surface area can be used to bind a large number of decoys without saturation (if the decoy requires a 10 nm² interaction surface on the microtubule, theoretically, each cell could anchor more than 3,000,000 decoy copies to the microtubule). This is significantly higher than the typical number of overexpressed proteins in transfected mammalian cells (for most proteins, approximately 100,000 copies are already considered a large overexpression).

[0070] A second advantage of using microtubule surfaces lies in their dynamic behavior. Microtubules are inherently highly dynamic and permanently alternate between shortening and growth phases. This behavior allows decoys to detach from microtubules during the depolymerization phase and bind to another microtubule or bind after microtubule repolymerization. During this time interval, the decoy moves away from the microtubule and is then able to capture prey located in the abundant cytoplasm rather than near the microtubule.

[0071] The detection of filamentous structures (such as “microtubule-like patterns”) has long been known to be highly sensitive, relying on extrapolating straight lines in fluorescence images to allow for signal detection from uniformly distributed noise. This general principle is commonly used for imaging microtubules via fluorescence speckle microscopy (see Salmon and Waterman; How we discovered fluorescent speckle microscopy; Mol Biol Cell, 3940-2; 2011 for a review). However, this readily available detection method is not feasible in structures that are closer to spherical or other improperly defined matrices. The method of this invention is also compatible with real-time data analysis of live-cell fluorescence images, thus enabling better detection of decoys bound to microtubules in live cells. Crucially, microtubules migrate over time due to their highly dynamic structure, which further improves their detection from the large fluorescence background used for detection in live cells.

[0072] "Microtubule-binding domain" refers to one or more segments of a microtubule-binding protein responsible for its binding to polymerized tubulin, and especially microtubules. A non-restrictive list of microtubule-binding proteins is provided in Table 1 below:

[0073] Table 1

[0074]

[0075] Therefore, the "polymerized tubulin binding portion" can be selected from full-length tubulin-binding proteins, tubulin-binding protein fragments, or even isolated tubulin-binding domains, which are known to bind specifically to polymerized tubulin, such as microtubules.

[0076] The “microtubule-binding domain” (MBD) involves one or more segments of a microtubule-binding protein responsible for binding to polymerized tubulin, and especially to microtubules.

[0077] According to the present invention, "determining the occurrence of an interaction" refers to identifying contact between the one or more protein decoys and the one or more candidate preys in the microtubule binding system according to the present invention. The protein decoys bind to polymerized tubulin in eukaryotic cells, thereby causing the one or more candidate preys to be localized along the polymerized tubulin, thereby detecting the interaction (if it occurs).

[0078] According to the present invention, "comparison of the occurrence of interactions" refers to identifying potential changes (disruption or stabilization) in the interactions between steps b. and d., thereby allowing inference that the action of the compound is disruptive or stabilizing of the interactions. In addition to disruption or stabilization, the method can also demonstrate that the interactions have not changed.

[0079] According to the present invention, "disruption of interaction" means that the one or more decoys and one or more candidate prey no longer interact, or that the interaction is weaker than before the step of contacting the test compound.

[0080] According to the present invention, "stability of interaction" means that the one or more decoys and one or more candidate prey interact more strongly than before the step of contacting the test compound.

[0081] According to the present invention, "quantification of interaction" refers to evaluating the strength of the interaction that occurs between the one or more protein decoys and the one or more candidate preys. This quantification allows the data obtained in steps b. and d. to be correlated with the strength of the interaction between the one or more decoys and the one or more candidate preys in order to infer whether the effect of the compound is disruptive or stabilizing, and to evaluate the strength of such disruption or stabilization.

[0082] According to the present invention, "contacting the eukaryotic cells with the compound" means introducing the compound into the culture medium for testing eukaryotic cells.

[0083] The IC50, or half-maximal inhibitory concentration, is the most widely used and informative measure of drug efficacy. It indicates the amount of drug required to inhibit half of a biological process, thus providing a measure of the potency of antagonist drugs in pharmacological studies. In the context of this invention, IC50 measures the amount of test compound required to disrupt or stabilize the interaction between the one or more protein decoys and the one or more candidate preys.

[0084] According to the present invention, "physiological" conditions can consist of an experimental setup in which eukaryotic cells are not in a state of stress or apoptosis, including an experimental setup in which eukaryotic cells are not in contact with a culture medium having non-physiological pH or salt conditions.

[0085] According to the present invention, a "protruding domain" refers to a domain found in MAPs (such as MAP2 or Tau) and involved in microtubule bundle formation and determining the spacing between microtubules. Protruding domains can also interact with other cytoskeleton structures. An example of a protruding domain is the protruding domain of the Tau protein, having the sequence SEQ ID NO: 3.

[0086] According to the present invention, a "protein disease" refers to any disease or symptom caused by abnormal synthesis, folding, post-translational modification, or deposition of proteins in cells or tissues. Examples of protein diseases include, but are not limited to, Alzheimer's disease, Parkinson's disease, Huntington's disease, and TDP-43 protein diseases such as ALS and FTLD.

[0087] According to the present invention, “TDP-43 proteinopathy” refers to a disease characterized by the presence of abnormally phosphorylated, ubiquitinated, and cleaved DNA-binding protein TDP-43 in the affected brain and spinal cord. Examples of TDP-43 proteinopathy include, but are not limited to, amyotrophic lateral sclerosis (ALS) and frontotemporal degeneration (FTLD).

[0088] According to the present invention, a "connector" refers to a non-structured domain, particularly a non-structured domain that allows for a nm-long spacing between the surface of polymeric tubulin and the protein of interest. Examples of connectors are known in the art. In one embodiment, the connector is a prominent domain of a microtubule-associated protein (MAP).

[0089] According to the present invention, “high content screening (HCS)” is an image-based method for drug discovery to identify molecules (biologics or small molecules) that alter phenotypes by using several cellular parameters as readouts simultaneously (Fraietta and Gasparri, 2016; Mattiazzi Usaj et al., 2016).

[0090] According to the present invention, a "compound" refers to any natural or chemical substance that can act on the interaction between the one or more decoys and the one or more candidate preys. Such a compound may be a known substance or derived from a library of substances based on in vivo, in vitro, or computer data. According to embodiments of the present invention, where the interaction between the one or more decoys and the one or more candidate preys is characteristic of a specific pathology, the compound may be a drug candidate. Specifically, the compound is a small molecule, peptide, protein, RNA, or DNA.

[0091] According to the present invention, "endogenous" refers to bait and / or prey originating from within an individual organism.

[0092] According to the present invention, the term "mutation" includes the deletion, truncation, addition, or substitution of a domain, amino acid, or glycosylation. Detailed Implementation

[0093] This invention relates to an in vitro method for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more candidate preys in eukaryotic cells.

[0094] Specifically, the in vitro method for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more candidate preys in eukaryotic cells includes the following steps:

[0095] a. Providing eukaryotic cells expressing (i) one or more baits and (ii) one or more candidate prey, wherein the bait comprises a bait portion and a polymerized tubulin-binding portion;

[0096] b. Detect the interaction by determining the presence of an interaction between the one or more decoys and the one or more candidate preys in the eukaryotic cell, wherein the decoys bind to tubulin in the eukaryotic cell, thereby causing the one or more candidate preys to be localized along the tubulin;

[0097] c. Contact the eukaryotic cells with the compound;

[0098] d. As in step b., determine the occurrence of an interaction between the one or more decoys and the one or more candidate preys in the eukaryotic cell, and quantify such an interaction;

[0099] e. Compare the occurrence of the interactions generated in steps b and d;

[0100] f. Infer whether the compound acts to disrupt or stabilize the interaction between one or more decoys and one or more candidate prey.

[0101] For steps a., b., and d., the interaction between the decoy and the candidate prey is detected. Methods and uses .

[0102] According to one embodiment, the "decoy" comprises at least one "decoy portion" and at least one "polymerized tubulin-binding portion", which are not derived from the same naturally occurring entity.

[0103] The lure can be brought to the microtubule using fused polymeric tubulin binding portions, such as microtubule-associated proteins (MAPs) including tau (NCBI reference sequence: NP_005901.2 or SEQ ID NO: 14) or microtubule-binding domains (MBDs).

[0104] According to one implementation, the decoy is anchored due to repetitions of microtubule-binding domains typically present in MAPs. These repetitions have a high affinity for microtubules compared to free tubulin (the building blocks of microtubules), and using this method, only the weaker fractions will remain free in the cytoplasm.

[0105] When prey interacts with a decoy brought to a microtubule, its repositioning on the microtubule surface allows for detection via the presence of microtubule structures in the prey's fluorescent image.

[0106] The detection of filamentous structures (such as “microtubule-like patterns”) has long been known to be highly sensitive, and this sensitivity lies in extrapolating straight lines in fluorescence images, allowing the signal to be detected from uniformly distributed noise. This general principle is commonly used for imaging microtubules using fluorescence speckle microscopy (see Salmon and Waterman; How we discovered fluorescent speckle microscopy; Mol Biol Cell, 3940-2; 2011 for a review).

[0107] However, this easy detection method is not feasible in structures that are closer to spherical or other improperly defined matrices. The method of this invention is also compatible with real-time data analysis of live-cell fluorescence images, and therefore can be performed to better detect decoys that bind to microtubules in live cells. Crucially, microtubules migrate over time due to their highly dynamic structure, which further improves their detection from the large fluorescence background used for detection in live cells.

[0108] For all these reasons, the choice of microtubules is highly relevant, even compared to other types of filaments. Because the interaction between the decoy and prey can be observed in real time, the effects of test compounds can be tracked.

[0109] As illustrated in the examples, the methods and uses of the present invention have been further validated. In short, plasmids have been designed to guide the expression in mammalian cells of a known protein (the decoy) fused with tau protein via its N-terminal or C-terminal domain (the end of the prominent domain of tau).

[0110] After fusion with tau, two decoys were successfully brought to the microtubules: P53 (tumor protein 53; accession number: BAC16799) (a transcription factor) and YB-1 (nuclease-sensitive element binding protein 1 or Y box binding protein; accession number: NP_004550.2) (an mRNA binding protein), both of which were readily detected along the microtubules in HeLa cells (human cancer cells, respectively) by optical microscopy.

[0111] After fusion with tau, two decoys were successfully brought to the microtubules: P53 (tumor protein 53; accession number: BAC16799) (a transcription factor) and YB-1 (nuclease-sensitive element-binding protein 1 or Y box-binding protein; accession number: NP_004550.2) (an mRNA-binding protein). Both were readily detected along the microtubules in U2-OS small cells (epithelial morphology cell lines derived from moderately differentiated sarcomas) by optical microscopy.

[0112] The "bait portion" and the candidate prey may be the same or different, and are preferably selected from the group consisting of proteins or nucleic acids (or composed of proteins or nucleic acids), including nucleoproteins and deoxyribonucleic acid.

[0113] In a non-restrictive manner, the "bait portion" and / or "candidate prey" includes any molecule of interest, particularly proteins of interest such as antibodies, nucleic acid-binding proteins, and various other bait receptor proteins or peptides, especially those of biological relevance, including those of diagnostic and pharmacological relevance. It also includes nucleic acids such as messenger RNA, coding and non-coding RNA, transfer RNA, ribosomal RNA, interfering RNA, or silencing RNA. It also includes any type of deoxyribonucleic acid, such as antisense DNA, circular DNA, complementary DNA, spacer DNA, heterologous DNA, hybrid DNA, homologous double-stranded DNA, linker DNA, mobile DNA (transposons), ribosomal DNA, satellite DNA, single-stranded DNA, double-stranded DNA, supercoiled DNA, or triple-stranded DNA.

[0114] In the method according to the invention, the RNA, protein, or DNA is endogenous.

[0115] Since the lure interacts with tubulin via the tubulin-binding moiety, the interaction between the lure and the prey is evaluated. Therefore, it is preferable that the lure portion does not interact or co-localize with tubulin at all, or interacts or co-localizes with tubulin in a limited manner, compared to the same portion conjugated to the tubulin-binding moiety under reference and / or physiological conditions.

[0116] The bait portion specifically considered is the bait portion that diffuses and distributes in the cytoplasm of eukaryotic cells under reference and / or physiological conditions.

[0117] According to some implementation schemes, the "decoy portion" is the portion that does not interact with or co-localize with polymerized tubulin (or microtubules) in eukaryotic cells under reference or physiological conditions.

[0118] According to one exemplary implementation, the "bait portion" is a nucleic acid binding portion, such as a nucleic acid binding protein, and the "candidate prey" is a nucleic acid, such as messenger RNA.

[0119] According to another embodiment, the "bait portion" is an antibody or a fragment thereof, and the "candidate prey" is a protein or peptide that readily binds to the antibody.

[0120] According to another implementation, the "bait portion" is a nucleic acid, such as messenger RNA, and the "candidate prey" is a nucleic acid binding portion, such as a nucleic acid binding protein.

[0121] "Candidate prey" can be natural or modified, homologous or heterologous.

[0122] Although “microtubule-associated protein” (MAP) can also be a polymeric tubulin-binding protein and / or a microtubule-binding protein, the term “microtubule-binding domain” (MBD) refers to a domain that can specifically and directly bind to microtubules. For the same reason, “polymeric tubulin-binding moiety” will refer to the portion that specifically and directly binds to polymeric tubulin.

[0123] Therefore, MBD can contain all possible amino acid sequences that lead to the binding of microtubule-binding proteins to microtubules.

[0124] For reference, the microtubule-binding domain of the present invention may be derived from Tau protein, such as Tau isotype 2 (accession number: NP_005901.2), which includes sequence SEQ ID N°4, which includes sequences SEQ ID N°5 to 8.

[0125] The microtubule-binding domain of the present invention can also be derived from the MAP1A protein (accession number: NP_002364), which includes the sequence SEQ ID N°9.

[0126] The microtubule-binding domain of the present invention can also be derived from the MAP2 protein (accession number: NP_002365), which includes the sequence SEQ ID N°10.

[0127] The microtubule-binding domain of the present invention can also be derived from the MAP4 protein (accession number: AAA67361), which includes the sequence SEQ ID N°11.

[0128] The microtubule-binding domain of the present invention can also be derived from the MAP6 protein (accession number: NP_149052), which includes the sequence SEQ ID N°12.

[0129] The microtubule-binding domain of the present invention can also be derived from the EB1 protein (accession number: NP_036457), which includes the sequence SEQ ID N°13.

[0130] Therefore, the microtubule-binding domains of the present invention may be selected from or comprise the group consisting of: Tau of sequences SEQ ID N°4 and SEQ ID N°5 to 8, MAP1A of sequence SEQ ID N°9, MAP2 of sequence SEQ ID N°10, MAP4 of sequence SEQ ID N°11, MAP6 of sequence SEQ ID N°12, EB-1 of sequence SEQ ID N°13, and / or any microtubule-binding domains derived from microtubule-associated proteins, and fragments and combinations thereof.

[0131] Methods for identifying polymeric tubule-binding moieties and / or microtubule-binding domains in proteins have been reported in this field. For reference, see: Cravchik et al.; Identification of a novel microtubule-binding domain in microtubule-associated protein 1A (MAP1A). J Cell Sci, 107 (Pt 3), 661-72, 1994.

[0132] According to one implementation, the decoy comprises:

[0133] - A polymeric tubulin-binding moiety containing one or more microtubule-binding domains (MBDs), and

[0134] - Bait section.

[0135] According to one embodiment, the decoy includes a connector (L) region located between the polymerized tubulin binding portion and the decoy portion.

[0136] According to one embodiment, the microtubule-binding domain is selected from the group consisting of: Tau of sequences SEQ ID N°4 and SEQ ID N°5 to 8, MAP1A of sequence SEQ ID N°9, MAP2 of sequence SEQ ID N°10, MAP4 of sequence SEQ ID N°11, MAP6 of sequence SEQ ID N°12, EB-1 of sequence SEQ ID N°13, and / or any microtubule-binding domain derived from microtubule-associated proteins, and fragments and combinations thereof.

[0137] According to one implementation, the candidate prey contains a fluorescent protein.

[0138] According to one embodiment, the method according to the invention includes step c', between step c. and step d., controlling the full entry of the compound into the cell. This optional step allows confirmation that the test compound may have been effective in the interaction between the decoy and prey. Cell treatments with various compounds involve the use of these molecules diluted in a solvent such as DMSO, and the duration of treatment can range from minutes to hours or even days. This flexibility in treatment duration allows researchers to evaluate the effects of the compound on cellular processes.

[0139] According to one embodiment, steps b and d further include quantifying the interaction between the one or more decoys and the one or more candidate prey, allowing comparisons between interactions before and after contact with the test compound, and allowing determination of the strength of the compound's destructive or stabilizing effects.

[0140] According to one implementation scheme, the cells in step b. are fixed cells or live cells.

[0141] The methods provided in steps a., b., and d. allow for the rapid identification of changes in interactions in living and immobilized eukaryotic cells (e.g., mammalian cells).

[0142] The nucleic acid sequence encoding the decoy is inserted into a plasmid to fuse with a given microtubule-binding domain. The nucleic acid interacting coupler can then be detected using antibodies targeting potential endogenous couplers in fixed cells or by screening with complementary oligonucleotides.

[0143] In both living and fixed cells, classic GFP marker plasmids encoding a set of prey associated with the interaction under study can be used to screen potential mates of a given bait and compare them with controls (e.g., tau-GFP protein only).

[0144] High-throughput analysis of protein and nucleic acid interactomes in fixed or live cells is also possible in 96-well, 384-well, and 1536-well plates.

[0145] If (i) the candidate prey is able to interact with the decoy, and if (ii) the decoy simultaneously binds to microtubules or tubulin polymerases in eukaryotic cells, then steps a., b., and d. can produce a “microtubule-like pattern” or a “tubulin polymerase-like pattern”.

[0146] Therefore, the detection of the “microtubule-like pattern” or “polymerized tubulin-like pattern” indicates the interaction between the one or more decoys and the one or more candidate preys in eukaryotic cells.

[0147] Specifically, the detection of candidate prey in eukaryotic cells can produce a “microtubule-like pattern” or a “polymerized tubulin-like pattern” if (i) the candidate prey can interact with the bait portion and if (ii) the polymerized tubulin binding portion binds simultaneously to microtubules or polymerized tubulin in eukaryotic cells.

[0148] Therefore, according to the described embodiment, detecting the “microtubule-like pattern” or “polymerized tubulin-like pattern” also indicates the interaction between the one or more bait portions in the eukaryotic cell and the one or more candidate prey.

[0149] According to one implementation, the one or more candidate prey and / or the one or more decoys are ribonucleic acid.

[0150] According to one implementation, the one or more candidate prey and / or the one or more decoys are deoxyribonucleic acid (DNA).

[0151] According to one embodiment, the one or more candidate prey and / or the one or more baits are proteins.

[0152] In general, the combination of the one or more candidate prey and the one or more decoys can be as follows:

[0153] i. The one or more baits: at least one protein / The one or more candidate prey: at least one protein

[0154] ii. The one or more lures: at least one protein / The one or more candidate prey: at least one ribonucleic acid

[0155] iii. The one or more decoys: at least one protein / The one or more candidate preys: at least one deoxyribonucleic acid

[0156] iv. The one or more decoys: at least one ribonucleic acid / The one or more candidate preys: at least one protein

[0157] v. The one or more decoys: at least one ribonucleic acid / The one or more candidate preys: at least one ribonucleic acid

[0158] vi. The one or more decoys: at least one ribonucleic acid / The one or more candidate preys: at least one deoxyribonucleic acid

[0159] vii. The one or more decoys: at least one deoxyribonucleic acid / The one or more candidate preys: at least one protein

[0160] viii. The one or more decoys: at least one deoxyribonucleic acid / The one or more candidate preys: at least one ribonucleic acid

[0161] ix. The one or more decoys: at least one deoxyribonucleic acid / The one or more candidate preys: at least one deoxyribonucleic acid

[0162] According to a preferred embodiment, the one or more baits are proteins.

[0163] According to a preferred embodiment, the one or more candidate prey are proteins or ribonucleic acid.

[0164] According to one embodiment of the invention, when the prey is a protein, the one or more candidate prey may comprise wild-type protein and mutant forms of the same protein.

[0165] According to one embodiment, the wild-type and mutant forms of the protein of interest comprise the sequence of the wild-type or mutant form of the protein of interest, a fluorescent tag, a linker, and one or more microtubule-binding domains. Preferably, the one or more microtubule-binding domains are selected from or comprise the group consisting of: Tau of sequences SEQ ID NO:4 and SEQ ID NO:5 to 8, MAP1A of sequence SEQ ID NO:9, MAP2 of sequence SEQ ID NO:10, MAP4 of sequence SEQ ID NO:11, MAP6 of sequence SEQ ID NO:12, EB-1 of sequence SEQ ID NO:13, and / or any microtubule-binding domains derived from microtubule-associated proteins, and fragments and combinations thereof.

[0166] The methods and uses of the present invention are suitable for evaluating the ability of compounds to disrupt or stabilize the interaction between one or more decoys and one or more candidate preys in eukaryotic cells, the eukaryotic cells comprising one decoy and one candidate prey, one decoy and more than one candidate prey, and more than one decoy and one candidate prey.

[0167] The method of the present invention is also suitable for evaluating the ability of compounds to disrupt or stabilize the interaction between multiple decoys and multiple candidate prey.

[0168] According to another embodiment, the method is used to evaluate the ability of a compound to disrupt or stabilize the interaction between a decoy and a candidate prey in eukaryotic cells.

[0169] According to the present invention, "eukaryotic cell" includes any eukaryotic cell that readily contains and / or expresses polymerized tubulin (such as microtubules).

[0170] According to a particular implementation, the eukaryotic cells are selected from the group comprising primary cells, stable cell lines, stem cells, and induced pluripotent stem cells (IPS). For example, the eukaryotic cells may be selected from a list comprising mammalian cells and insect cells, or a list consisting of mammalian cells and insect cells.

[0171] Specifically, eukaryotic cells can be selected from a list including or consisting of the following: HEK cells, NR cells, U2-OS cells, and HeLa cells, especially HeLa cells.

[0172] The cells in step b. can be fixed cells or live cells.

[0173] When the cells in step b. are living cells, it is preferable that the bait and / or candidate prey contain detectable tags, such as fluorescent tags.

[0174] The use of live cells in step b. allows for the detection of candidate prey along the dynamic polymers of tubulin, such as microtubules, when interacting with a decoy.

[0175] When the cells in step b. are fixed cells, the method requires an additional fixation step prior to step b. The methods for obtaining fixed cells are well known in the art.

[0176] In any case, microtubules appear as bright lines on the fluorescence image, which makes it easy to detect the binding of prey by using the fluorescence signal of the prey, for example, the appearance of long bright lines, resulting in the characterization of "microtubule-like patterns".

[0177] Candidate prey can be modified or unmodified. Candidate prey can also be homologous or heterologous.

[0178] The lure and / or prey candidate may also contain a detectable component. Specifically, the detectable component may be a fluorescent protein.

[0179] The detectable component can be selected from the group including the following: GFP, YFP, XFP, RFP, CFP, DsRED, mCherry, luciferase, anthocyanin dyes (such as Cy2, Cy3, or Cy5), luciferin, rhodamine, and Alexa fluor dye.

[0180] Specifically, fluorescent proteins can be selected from the group consisting of: GFP, YFP, XFP, RFP, CFP, DsRED, and mCherry.

[0181] Fluorescent proteins are well known in the field and can be found, for example, in Shaner & Steinbach & Tsien (A guide to choosing fluorescent proteins; Nat Methods; 2(12):905-9;2005).

[0182] The lure binds to polymeric tubulin in eukaryotic cells, thereby allowing one or more candidate prey to be located along the polymeric tubulin, thus enabling the detection of the interaction.

[0183] According to another embodiment, the present invention relates to a method as defined above, wherein in steps b. and d., the determination of the presence of the prey in the eukaryotic cell is performed using a detection method selected from the group consisting of: antibody binding, nucleic acid hybridization, and / or fluorescence measurement.

[0184] For example, steps b and d, which determine the occurrence of interactions, can be performed using video microscopy.

[0185] Steps b and d, which determine the occurrence of interactions, can also be implemented using computer-aided identification methods, such as those taught by Altinok et al. (Activity analysis in microtubule videos by mixture of hidden Markov models; Computer Vision and Pattern Recognition, IEEE Computer Society Conference, 2, 1662-1669; 2006).

[0186] Furthermore, the present invention relates to a method as defined above, which includes at least one step prior to steps b. and d. to depolymerize cellular tubulin using a microtubule depolymerization agent or cold exposure.

[0187] Microtubule disruptors, microtubule depolymerizers, or microtubule degrading agents, such as Nocodazole, vinblastine, colchicine, colchicine, podophyllotoxin, Rizhoxin, or vinorelbine, can also be used for this purpose.

[0188] Cold exposure is known in the art and generally involves steps to depolymerize microtubules by exposing them to low temperatures. Protocols associated with cold exposure are known in the art and are taught, for example, in Ochoa et al., Cold exposure reveals two populations of microtubules in pulmonary endothelia; Am. J. physiol. LungCell. Mol. Physiol; 300:L132-L138; 2011.

[0189] The polymerization and depolymerization steps defined above can be advantageously repeated over time, thereby generating a continuous alternation of decoy association and dissociation over time.

[0190] Therefore, for the in vitro method according to the present invention, the steps can be arranged as follows:

[0191] a. Providing eukaryotic cells expressing (i) one or more baits and (ii) one or more candidate prey, wherein the bait comprises a bait moiety and a polymerized tubulin-binding moiety,

[0192] Microtubule-depolymerizing agents or cold exposure can be used to depolymerize cellular microtubules.

[0193] b. Determine the occurrence of a first interaction between the one or more decoys and the one or more candidate preys in eukaryotic cells, wherein the decoy binds to polymeric tubulin in eukaryotic cells.

[0194] Optionally, microtubule-depolymerizing agents or cold exposure can be used to depolymerize cellular microtubules, and

[0195] b'. Determine the occurrence of a second interaction between the one or more decoys and the one or more candidate preys in eukaryotic cells, wherein the decoys bind to polymeric tubulin in eukaryotic cells.

[0196] c. Contact the eukaryotic cells with the compound;

[0197] Optionally, microtubule-depolymerizing agents or cold exposure can be used to depolymerize cellular microtubules, and

[0198] d. As in step b., determine the occurrence of a first interaction between the one or more decoys and the one or more candidate preys in eukaryotic cells, and quantify such interaction;

[0199] Optionally, microtubule-depolymerizing agents or cold exposure can be used to depolymerize cellular microtubules, and

[0200] d'. As in step b., determine the occurrence of a second interaction between the one or more decoys and the one or more candidate preys in eukaryotic cells, and quantify this interaction;

[0201] In step b, the decoy binds to polymeric tubulin in eukaryotic cells, thereby locating one or more candidate prey along the polymeric tubulin, thus detecting the first interaction.

[0202] In step b', the decoy binds to polymeric tubulin in eukaryotic cells, thereby causing the one or more candidate prey to be localized along the polymeric tubulin, thereby detecting the second interaction.

[0203] In step d, the decoy binds to polymeric tubulin in eukaryotic cells, thereby locating one or more candidate prey along the polymeric tubulin, thus detecting the first interaction.

[0204] In step d', the decoy binds to polymeric tubulin in eukaryotic cells, thereby causing the one or more candidate prey to be localized along the polymeric tubulin, thereby detecting the second interaction.

[0205] Of course, the depolymerization step can be repeated over time.

[0206] According to the method described, not only can the interaction between the bait and the prey be detected, but also the dynamics of polymerized tubulin in eukaryotic cells can be detected.

[0207] Because decoys and prey can interact in their living environment, the object of this invention is not only to allow the compound to influence the interaction between decoy and prey, but also to allow the detection of post-translational modifications that may occur in eukaryotic cells and can affect the efficacy of the compound. Therefore, the method of this invention may further include the step of detecting the presence of post-translational modifications, such as those on the decoy moiety and / or on the candidate prey. Post-translational modifications are well known in the art and include modifications such as phosphorylation events.

[0208] Therefore, this method is also effective for identifying post-translational events (such as altered phosphorylation of the decoy-prey interaction) in real time using, for example, video microscopy, as detailed in the example of the eIF2B-eIF2A interaction.

[0209] The present invention also relates to the use of polymerized tubulin binding moieties as a tool for evaluating the ability of compounds to disrupt or stabilize the interaction between one or more protein decoys and one or more candidate preys in eukaryotic cells.

[0210] The decoy binds to polymeric tubulin in eukaryotic cells, thereby allowing one or more candidate prey to be localized along the polymeric tubulin, thus detecting the ability of the compound to disrupt or stabilize the interaction between one or more protein decoys and one or more candidate prey.

[0211] Specifically, the present invention relates to the use of a polymeric tubulin-binding portion fused with one or more decoys as a tool for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more protein decoys and one or more candidate preys in eukaryotic cells. The decoy binds to polymeric tubulin in eukaryotic cells, thereby locating the one or more candidate preys along the polymeric tubulin, thereby evaluating the ability of a compound to disrupt or stabilize the interaction between one or more protein decoys and one or more candidate preys.

[0212] More specifically, the present invention relates to the use of at least one microtubule-binding domain fused with one or more decoys as a tool for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more protein decoys and one or more candidate preys in eukaryotic cells. The decoy binds to polymeric tubulin in eukaryotic cells, thereby localizing the one or more candidate preys along the polymeric tubulin, thus evaluating the ability of a compound to disrupt or stabilize the interaction between one or more protein decoys and one or more candidate preys.

[0213] According to the embodiments described, the present invention also relates to the use of multiple microtubule-binding domains (such as tandem microtubule-binding domains) fused with one or more decoys as a tool for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more candidate preys in eukaryotic cells.

[0214] Preferably, the polymerized tubulin-binding moiety and / or microtubule-binding domain can be used in the form of a decoy as further disclosed below.

[0215] bait

[0216] The present invention also provides decoys suitable for the methods and uses defined above.

[0217] Specifically, this paper describes a decoy comprising (i) a polymeric tubulin-binding moiety containing one or more microtubule-binding domains (MBDs), and (ii) a decoy moiety.

[0218] According to some implementation schemes, the "decoy portion" does not interact with or co-localize with polymerized tubulin or microtubules in eukaryotic cells.

[0219] According to some embodiments, the decoy portion may consist of an antibody or a fragment thereof, or a nucleic acid-binding protein. Therefore, according to some embodiments, a protein decoy comprises:

[0220] - A polymeric tubulin-binding moiety containing one or more microtubule-binding domains, and

[0221] - The decoy portion that does not interact with polymerized tubulin in eukaryotic cells.

[0222] According to some embodiments, the decoy portion may consist of an antibody or a fragment thereof, or a nucleic acid-binding protein. Therefore, according to some embodiments, a protein decoy comprises:

[0223] - A polymeric tubulin-binding moiety containing one or more microtubule-binding domains, and

[0224] - The decoy component, consisting of an antibody or its fragments or a nucleic acid-binding protein.

[0225] Barriers to interaction can be reduced lure accessibility due to proximity to the microtubule surface when fused with the polymeric tubulin binding site. To maximize lure accessibility to prey, the lure can be attached to a protruding domain, preferably an unstructured tail that allows for a nm-long gap between the lure and the microtubule or polymeric tubulin surface.

[0226] Therefore, according to a particular embodiment, a decoy suitable for the methods and uses of the present invention comprises: a polymerized tubulin binding portion comprising one or more microtubule-binding domains (MBDs) and a decoy portion.

[0227] Advantageously, the microtubule-binding domain (MBD) is selected from the group consisting of: Tau of sequence SEQ ID N°4, MAP1A of sequence SEQ ID N°9, MAP2 of sequence SEQ ID N°10, MAP4 of sequence SEQ ID N°11, MAP6 of sequence SEQ ID N°12, EB-1 of sequence SEQ ID N°13, or any other microtubule-binding domain derived from microtubule-associated proteins, and combinations thereof.

[0228] According to a preferred embodiment, the one or more baits are proteins.

[0229] According to some embodiments, the protein decoy comprises: a polymeric tubulin-binding moiety comprising one or more microtubule-binding domains selected from the group consisting of: Tau of sequences SEQ ID N°4 and SEQ ID N°5 to 8, MAP1A of sequence SEQ ID N°9, MAP2 of sequence SEQ ID N°10, MAP4 of sequence SEQ ID N°11, EB-1 of sequence SEQ ID N°13, and / or fragments and combinations thereof; and a decoy portion.

[0230] According to some embodiments, the protein decoy comprises: a polymeric tubulin-binding portion comprising one or more microtubule-binding domains selected from the group consisting of: Tau of sequences SEQ ID N°4 and SEQ ID N°5 to 8, MAP1A of sequence SEQ ID N°9, MAP2 of sequence SEQ ID N°10, MAP4 of sequence SEQ ID N°11, EB-1 of sequence SEQ ID N°13, and / or fragments and combinations thereof; and a decoy portion that does not interact with polymeric tubulin in eukaryotic cells.

[0231] According to some embodiments, the protein decoy comprises: a polymeric tubulin-binding moiety comprising one or more microtubule-binding domains selected from the group consisting of: Tau of sequences SEQ ID N°4 and SEQ ID N°5 to 8, MAP1A of sequence SEQ ID N°9, MAP2 of sequence SEQ ID N°10, MAP4 of sequence SEQ ID N°11, EB-1 of sequence SEQ ID N°13, and / or fragments and combinations thereof; and a decoy moiety consisting of an antibody or a fragment thereof or a nucleic acid-binding protein.

[0232] Therefore, the decoy may also include a linker (L) region located between the polymerized tubulin binding portion and the decoy portion.

[0233] The linker (L) region is typically an unstructured domain, especially an unstructured domain that allows for a nm-long spacing between the polymerized tubulin surface and the decoy, which is crucial for increasing the decoy's accessibility to the prey. Examples of linker regions are known in the art.

[0234] Protruding domains are present in microtubule-associated proteins (MAPs) such as MAP2 or Tau, and are involved in microtubule bundle formation and determining the spacing between microtubules. They can also interact with other cytoskeleton structures.

[0235] Preferably, to maximize the accessibility of the decoy to the prey, the decoy is connected to a junction region, which is a prominent structural domain from a MAP (such as Tau) or a fragment thereof.

[0236] Therefore, according to a preferred embodiment, the protein decoy comprises (i) a polymeric tubulin-binding moiety containing one or more microtubule-binding domains (MBDs), (ii) a protruding domain, and (ii) the decoy moiety.

[0237] The “protruding structural domain” suitable for the methods and uses of the present invention may comprise or consist of N-terminal segments of Tau.

[0238] According to a particular implementation, the "prominent domain" is the Tau prominent domain or a fragment thereof of sequence SEQ ID N°3.

[0239] The linker region or protruding domain can have different lengths, including any region or domain as defined above that is 1 to 150 amino acids in length, including lengths of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31. 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 11 7, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, and 150 amino acids.

[0240] When the bait contains (i) a polymeric tubulin-binding portion containing one or more microtubule-binding domains and (ii) a protruding domain, it is preferred that both portions are part of the same protein, especially part of the same microtubule-associated protein such as Tau.

[0241] The decoy may or may not contain a detectable component, such as a fluorescent protein.

[0242] Advantageously, the decoy may also contain any detectable portion as defined above in its N-terminal or C-terminal portion, such as a fluorescent label detectable using fluorescence microscopy.

[0243] In a non-limiting manner, the detectable portion may be a fluorescent protein, such as a protein selected from the group consisting of: GFP, YFP, XFP, RFP, CFP, DsRED, and mCherry.

[0244] Vectors and cell lines

[0245] This article also describes a carrier containing an expression box suitable for expressing decoys according to the uses and methods of the present invention.

[0246] Specifically, the present invention relates to a carrier containing an expression cassette suitable for expressing the protein decoy of the present invention.

[0247] Specifically, the present invention relates to a carrier containing an expression cassette suitable for expressing the present invention's RNA decoy.

[0248] Specifically, the present invention relates to a carrier containing an expression cassette suitable for expressing the present invention via a deoxyribonucleic acid decoy.

[0249] Specifically, the present invention relates to a carrier containing an expression cassette suitable for expressing the protein prey of the present invention.

[0250] Specifically, the present invention relates to a carrier containing an expression cassette suitable for expressing the RNA prey of the present invention.

[0251] Specifically, the present invention relates to a carrier containing an expression cassette suitable for expressing the deoxyribonucleic acid prey of the present invention.

[0252] Therefore, the carrier suitable for the uses and methods of the present invention is suitable for expressing decoys, and the carrier includes carriers comprising:

[0253] - An expression cassette encoding a decoy moiety fused to a polymeric tubulin-binding moiety containing one or more microtubule-binding domains (MBDs), or alternatively,

[0254] - Encodes a polymeric tubulin-binding moiety containing one or more microtubule-binding domains (MBDs) and is directly suitable for cloning at least one decoy moiety into a fusion decoy expression cassette.

[0255] If the expression cassette is suitable for introducing an insert encoding the decoy moiety within the same frame as the polymeric tubulin-binding portion, then according to standard cloning protocols, the expression cassette is directly suitable for cloning at least one decoy moiety into a fusion protein decoy. Standard cloning protocols include those comprising steps of restriction enzyme digestion and / or site-specific recombination.

[0256] Vectors may include one or more selection markers. Selection markers are gene markers introduced into cells, particularly bacteria or cultured cells, to confer traits suitable for artificial selection.

[0257] According to a preferred embodiment, the expression cassette encodes a polymeric tubulin-binding moiety comprising one or more microtubule-binding domains (MBDs) of microtubule-associated proteins derived from the present invention.

[0258] According to a preferred embodiment, the expression cassette encodes a polymeric tubulin-binding moiety comprising one or more microtubule-binding domains (MBDs) of the protein Tau derived from sequence SEQ ID N°14. A vector suitable for the uses and methods of the present invention may comprise:

[0259] - An expression cassette encoding a protein library containing a polymeric tubulin-binding moiety fused with one or more microtubule-binding domains (MBDs), or alternatively,

[0260] - An expression cassette that encodes a polymeric tubulin-binding moiety containing one or more microtubule-binding domains (MBDs) and is directly suitable for cloning nucleic acids of a protein library encoding the decoy moiety into a fusion decoy.

[0261] Examples of vectors suitable for expression in eukaryotic cells include the Gateway© pEF-Dest51 plasmid.

[0262] This article also describes cell lines transfected with the vectors described above and / or stably expressing baits suitable for the methods and uses of the present invention.

[0263] This article also describes a kit for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more preys in cells, the kit comprising decoys, carriers and / or cell lines suitable for the methods and uses of the present invention.

[0264] This document also describes a kit for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more preys in cells, the kit comprising decoys, carriers and / or cells, and combinations thereof, suitable for the methods and uses of the present invention.

[0265] This document also describes a kit as defined above, comprising a nucleic acid library encoding the bait or the prey, and / or a protein library containing the bait or the prey, and / or a deoxyribonucleic acid library encoding the bait and / or reagents suitable for detecting the presence of the bait or the prey in the cells, and combinations thereof.

[0266] This article also describes a kit for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more preys in eukaryotic cells, the kit comprising...

[0267] -The vector of the present invention, and / or the cell line of the present invention,

[0268] -Optionally, a protein library of one or more bait portions,

[0269] -Optionally, a reagent suitable for detecting the presence of the protein decoy and / or the candidate prey in eukaryotic cells.

[0270] This article also describes a kit for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more preys in eukaryotic cells, the kit comprising:

[0271] -The vector of the present invention, and / or the cell line of the present invention,

[0272] -A nucleic acid library or protein library encoding one or more decoy motifs.

[0273] -Optionally, a reagent suitable for detecting the presence of the protein decoy and / or the candidate prey in eukaryotic cells.

[0274] According to one embodiment of the invention, steps a and b are repeated prior to step c with different carrier constructs of the one or more decoys and the one or more candidate preys in order to determine the occurrence of interactions between the one or more decoys and the one or more candidate preys in eukaryotic cells and to select the correct combination for evaluation in the following steps to more closely approximate what occurs in vivo.

[0275] In a preferred embodiment, a vector library (pDEST plasmid) is generated that is specific to and exhaustive for the method according to the invention in order to characterize the interaction patterns between PPIs or RPIs, regardless of the system.

[0276] The pDEST vector is generated in a combinatorial manner to take into account all possible combinations of fusion for a given fluorophore.

[0277] In the case of PPI interactions between protein A (decoy) and protein B (prey), the following pDEST vector can be used, where pDON can be protein A (A_pDON) and protein B (B_pDON):

[0278] The four pDEST vectors at the N-terminus:

[0279] a. Decoy: A_pDON-GFP-MBD (SEQ ID NO 14)

[0280] b. Decoy: A_pDON-RFP-MBD (SEQ ID NO 14)

[0281] c. Prey: B_pDON-GFP

[0282] d. Prey: B_pDON-RFP

[0283] The four pDEST vectors in the C-terminus:

[0284] a. Decoy: MBD (SEQ ID NO 14)-GFP-A_pDON

[0285] b. Decoy: MBD (SEQ ID NO 14)-RFP-A_pDON

[0286] c. Prey: GFP-B_pDON

[0287] d. Prey: RFP-B_pDON

[0288] For this system, the following 8 plasmids were obtained:

[0289] Regarding decoys:

[0290] aA-GFP-MBD (SEQ ID N° 14)

[0291] bA-RFP-MBD (SEQ ID N° 14)

[0292] c.MBD (SEQ ID N° 14)-GFP-A

[0293] d.MBD (SEQ ID N° 14)-RFP-A

[0294] For the prey:

[0295] aB-GFP

[0296] bB-RFP

[0297] c.GFP-B

[0298] d.RFP-B

[0299] For the PPI A and B systems, 16 plasmids can be generated in detail to characterize the interaction patterns of protein pairs of two given fluorophores: GFP and RFP.

[0300] In this case, the combination is associated with the following factors

[0301] a. Two types of proteins (A / B)

[0302] b. Combinations of two types (bait / prey)

[0303] c. Two types of fusion (N-terminal / C-terminal)

[0304] d. Two types of fluorophores (GFP / RFP)

[0305] Steps a and b are repeated before step c. to determine the interaction between the one or more decoys and the one or more candidate preys in eukaryotic cells, and the correct combination is selected for evaluation in the following steps, with the advantage of more closely resembling the situation occurring in vivo, in the non-limiting case of PPI interactions:

[0306] a. Improved cell microscopy results for identifying and characterizing binding patterns of interactions (PPI (decoy / prey) or RPI: decoy only);

[0307] b. Select the best system for molecular screening to identify regulation (stability or disruption of bait / prey interaction);

[0308] c. It can generate more carriers for each fluorophore under consideration;

[0309] d. Reduce the number of validation steps in classical molecular biology (e.g., restriction enzymes), and thus increase the robustness and speed of obtaining plasmids of interest;

[0310] e. In the context of target identification, this method is applied to identify protein couples.

[0311] This repetitive step relates to any combination of bait and prey (DNA, RNA, or protein) according to the invention.

[0312] General Scheme for High-Throughput Analysis and Evaluation Systems

[0313] In the case of protein decoys, a cDNA sequence encoding a known protein, or "decoy," is inserted into a plasmid. This plasmid directs the synthesis of the decoy within the cell. The decoy is fused with a microtubule-binding protein (tau, MAP2) or a domain linked to a protruding structure to enhance the decoy's accessibility to the prey. The decoy may also be fused with a fluorescent label such as a fluorescent protein (GFP, RFP, ...).

[0314] The cDNA sequence encoding the "prey" (peptide or protein) is inserted into a plasmid to guide its expression within cells. The prey may be fused with a fluorescent label such as a fluorescent protein (i.e., GFP or XFP) to be detected in living cells via fluorescence microscopy.

[0315] Cell sample arrays in 6- to 96-well plates were co-transfected with different prey (see...). Figure 1 This allows for high-throughput identification of various interacting mates of a given decoy. The automated detection system enables the identification of prey bound to microtubules in order to obtain a Boolean output (interacting or non-interacting).

[0316] 4) Testing:

[0317] -i) For detection in live or fixed cells ( Figure 2 and Figure 3 When interacting with a decoy, co-expression of fluorescent decoys and prey allows for the detection of prey along dynamic microtubules via fluorescence video microscopy. Time-lapse imaging of less than one minute (15 images per minute) is sufficient to detect microtubule structures with high sensitivity. Sensitivity is enhanced in live cells compared to fixed cells because the highly dynamic microtubules allow for the detection of prey bound to the decoy via differential fluorescence images of the prey (image at time t0 minus image at time t0+Δt) (Desforges et al.; An intercellular polyamine transfer via gap junctions regulates proliferation and response to stress in epithelial cells. Mol Biol Cell 24, 1529-1543; 2013), a classic technique for highlighting microtubule movement, which can be used here to eliminate the fluorescent background of non-interacting prey.

[0318] To improve the accessibility of decoys to prey and enhance detection sensitivity, reversible drugs such as nocodazole can be used to depolymerize microtubules in living cells. Nocodazole allows the decoy to freely explore the cytoplasm and interact with the decoy far from the microtubule surface. Once the depolymerizing drug is washed out, the microtubules rapidly regrow from the centromere (<2 minutes). These patterns are captured using video microscopy, utilizing the fluorescence of the prey, whenever an interaction occurs between the decoy and prey. Confocal fluorescence microscopy or total internal fluorescence microscopy (TIRF) can be used to enhance sensitivity by reducing the fluorescence signal-to-noise ratio.

[0319] -ii) For detection in fixed cells ( Figure 3This method allows for the fluorescence detection of natural prey (endogenous proteins, endogenous RNA, endogenous DNA, or other biomolecules) brought to the microtubules due to their interaction with the bait. Detection can be performed using classical optical fluorescence microscopy after labeling the prey with specific antibodies (selected for proteins that potentially interact with the bait and using a negative control), or after in situ hybridization when the aim is to detect the interaction of nucleic acids with the bait (RNA, especially including miRNA and mRNA). This choice is also relevant when the fluorescent labeling of the prey (GFP, RFP, or other labels) might hinder prey-bait interactions, or when prey overexpression might bias the results.

[0320] For detection in fixed cells co-expressing decoys and fluorescent prey, fluorescent detection of prey brought to the microtubules by their interaction with the decoy can be performed directly using classical optical microscopy due to their fluorescent labeling.

[0321] In any case, the microtubes appear as bright lines on the fluorescence image, which makes it easy to detect the binding of prey on the appearance of the µm-long bright lines by using the fluorescence signal of the prey.

[0322] Stable cell lines expressing decoys can also be used to improve the efficiency of co-expression and facilitate the study of interactions between a given protein decoy. If decoy expression is toxic, an inducible expression system can be developed in stably transfected cell lines to achieve significant expression at a given time.

[0323] After step c., which involves contacting the eukaryotic cells with the compound, step d. is processed as discussed in this section for steps a. and b., and the results are compared in step e. to infer in step f. whether the compound disrupts or stabilizes the interaction.

[0324] Test compounds (step c.) and potential candidates for treating diseases

[0325] The in vitro method according to the invention includes step c. of contacting the eukaryotic cells with the compound, ultimately resulting in an alteration in the interaction between the decoy and the prey, and allowing inference that the compound acts to disrupt or stabilize the interaction between the one or more decoys and one or more candidate prey (step f.).

[0326] Contact between the eukaryotic cells and the compound is preferably performed using an automated robot with 96-well, 384-well, or 1536-well plates. The eukaryotic cells are treated with various compounds diluted in a solvent such as DMSO. The treatment duration can range from minutes to hours or even days.

[0327] The contact between the eukaryotic cells and the compound is preferably carried out by diluting the compound in a solvent, more preferably DMSO. The compound is diluted to different concentrations depending on the test compound.

[0328] The test compound was then brought into contact with the eukaryotic cells.

[0329] The duration of contact between the test compound and the eukaryotic cells ranges from several minutes to several days.

[0330] According to one embodiment of the present invention, the duration of contact between the test compound and the eukaryotic cells is selected from: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, or 59 minutes.

[0331] According to one embodiment of the present invention, the duration of contact between the test compound and the eukaryotic cells is selected from: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23 or 24 hours.

[0332] According to one embodiment of the present invention, the duration of contact between the test compound and the eukaryotic cells is selected from: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 or 31 days.

[0333] The compound can be natural or a chemical substance. In one embodiment, the compound is a small molecule.

[0334] This compound can be a known substance or a substance library based on in vivo, in vitro, or computer data.

[0335] In one particular embodiment of the invention, the method is repeated with different concentrations of the compound in order to determine the IC50 of the test combination of the compound disrupting or stabilizing one or more decoys and one or more candidate preys.

[0336] In one particular embodiment of the invention, the method according to the invention allows for screening molecular libraries to identify potential candidates and to determine these relevant candidates using IC50.

[0337] According to embodiments of the present invention, the compound may be a drug candidate, provided that the interaction between the one or more decoys and the one or more candidate preys is characteristic of a specific pathology.

[0338] This invention allows for the evaluation of the ability of such compounds to disrupt or stabilize the interaction between one or more decoys and one or more candidate preys in eukaryotic cells. Since the combination of prey and decoy can be a characteristic of an interaction (or the absence of an interaction) present in a specific pathology, this method allows for the identification of drug candidates that will be able to act on the interaction (or the absence of an interaction) and improve the treatment of said pathology.

[0339] In examples of the invention not limited to the combinations covered by this application, the applicant has been able to identify compounds that affect RPI interactions (protein bait and RNA prey) and PPI interactions (protein bait and protein prey), demonstrating the efficacy and potential of the methods according to the invention.

[0340] According to one embodiment, the present invention relates to a compound identified by an in vitro method according to the invention, the compound being used as a drug.

[0341] In one particular embodiment of the invention, the one or more decoys and one or more candidate prey are characteristics of cancer.

[0342] In a preferred embodiment of the invention, the cancer is selected from the group consisting of or including the following: colorectal cancer, malignant brain tumor, lung cancer, drug-resistant cancer, and pancreatic cancer.

[0343] In this embodiment, the method and use according to the invention can infer whether the effect of the test compound is disruptive or stabilizing interaction, and thus infer whether one or more test compounds can be of interest as potential drug candidates in cancer signatures of a combination of decoy and prey.

[0344] According to one implementation scheme, the compound is used to treat cancer in a subject.

[0345] In a particular embodiment of the invention, the one or more lures, preferably protein lures, and one or more candidate prey are characterized by protein disease.

[0346] According to one implementation scheme, the compound is used to treat a subject's protein disease.

[0347] In a preferred embodiment of the invention, the protein disease is selected from the group consisting of or including Alzheimer's disease, Parkinson's disease, Huntington's disease, amyotrophic lateral sclerosis, frontotemporal dementia, TDP-43 or FUS protein diseases.

[0348] In a more preferred embodiment of the invention, the protein disease is TDP-43 protein disease, preferably amyotrophic lateral sclerosis (ALS) or frontotemporal degeneration (FTLD), and the protein of interest is TDP-43.

[0349] According to one implementation scheme, the mutation forms of TDP-43 are selected from the group consisting of or including the following: A90V, D169G, N267S, G287S, G290A, S292N, G294A / V, G295C / R / S, G298S, M311V, A315T / E, A321G / V, Q331K, S332N, G335 D. M337V, Q343R, N345K, G348C / R / V, N352S / T, G357R / S, R361S / T, P363A, G368S, Y 347X, G376D, N378D / S, S379C / P, A382P / T, I383V, G384R, W385G, N390S / D and S393L.

[0350] The method according to the present invention has been able to identify compounds associated with TDP-43 proteinopathy.

[0351] Many mutations in the C-terminal domain of TDP-43 facilitate the transition of TDP-43 from a liquid state (where the protein exists in a reversible liquid subcellular compartment) to a solid aggregate state (where the protein accumulates in the cell to form solid and irreversible aggregates).

[0352] The method according to the WO2016012451 patent family has been used to study mutations involving the transition of TDP-43 to a solid aggregate state, a transition that is characteristic of ALS pathology.

[0353] Since the mutant forms of TDP-43 exhibit various defects in cell testing, this approach aims to find compounds capable of restoring these defects.

[0354] 134 compounds from a library of kinase or phosphatase inhibitors have been tested using the method according to the invention with the system TDP43-GFP-MBD (SEQ ID No. 14) [wt] / SE3-RFP-MBD (SEQ ID No. 14).

[0355] After the first screening, 13 compounds that increased protein stratification (i.e., the solid aggregate state indicating the mutant form of TDP-43) or conversely increased protein mixing (i.e., the liquid state indicating the mutant form of TDP-43) were selected without any toxicity issues.

[0356] Thirteen selected compounds were then tested in three cell assays: enrichment in wtTDP-43 stress granules and its subcellular localization in the presence of the tested compounds; and the effect of the tested compounds on the alternative splicing of the cystic fibrosis transmembrane transport regulator (CFTR) gene by wtTDP-43. Finally, the effects of the thirteen compounds on the aggregation of the TDP-43 G146A mutant were tested.

[0357] Kinase inhibitors increased protein stratification and aggregation of mutant G146A, while phosphatase inhibitors increased protein mixing and reduced aggregation of mutant G146A. Of these compounds, four showed effects in at least two of the four assays: three kinase inhibitors (Ro-methanesulfonate, PKRi, and 5-iodotuberculin) and one phosphatase inhibitor (9,10-phenanthroquinone). Therefore, these results demonstrate that the phosphatase inhibitors identified by this method can restore defects observed in ALS, such as the accumulation of mutant TDP-43.

[0358] These results show that the method according to the invention can be used to test mutant forms of proteins of interest, particularly proteins that accumulate in subjects with protein diseases, and to identify compounds that can be used to treat said protein diseases.

[0359] Example

[0360] Example 1: Targeting Proteins: Protein-Protein Interactions to Identify Regulators in the Cellular Environment

[0361] We selected the widely studied system p53 / MDM2 (Moll and Petrenko, 2003; Nayak et al., 2018; Zhao et al., 2015) as a proof-of-concept to investigate the performance of the MT bench technique using a high-content screening system to screen PPI inhibitors. High-content screening (HCS) is an image-based approach for drug discovery to identify molecules (biologics or small molecules) that alter phenotypes by simultaneously using several cellular parameters as readouts (Fraietta and Gasparri, 2016; Mattiazzi Usaj et al., 2016). An analytical workflow was developed to measure colocalization on individual cells from HCS images by extracting correlation scores. The correlation scores were calculated using the Pearson correlation coefficient (PCC) between the enhanced texture intensities from prey and decoy. We obtained a good Z' factor of 0.69 (between 0.5 and 1, the assay is considered excellent) (Zhang, 1999; Zhang et al., 2000), reflecting the robustness of this technique in screening small inhibitors of p53 / MDM2 with high content. We then validated that the interaction detected on the microtubules depended on specific residues at the MDM2-p53 binding site. Having demonstrated the effectiveness of the MT bench assay in quantifying the p53 / MDM2 interaction, we began screening a small group of 10 molecules for hit identification, which had known activity against MDM2, as previously characterized by the HTRF assay (Gicquel et al., 2018). False negatives and false positives identified by the HTRF assay were successfully detected, and this was also independently confirmed by measuring the accumulation of p53 and p53-related proteins. The MT bench assay was also used to calculate IC50 from HCS images. We obtained cellular IC50 values ​​of 74 nM and 426 nM for the known p53 / MDM2 inhibitors RG7112 and Nutlin-3a, respectively (Vu et al., 2013). These values ​​are consistent with those found in the literature from in vitro HTRF assays (Liu et al., 2019; Neocoritis et al., 2019a; Vu et al., 2013) and from cellular nanoBRET assays (Itatani and Olsen).

[0362] We also used the results to characterize the inhibition constants of all the tested molecules that allow for the generation of cellular structure-activity relationships (SARs). In summary, the results presented in this study validate that the MT bench technique in an HCS setting is a robust method for screening and characterizing PPI regulators in cells.

[0363] Methods and materials:

[0364] Generation of genetic constructs

[0365] The cloning of genes of interest was performed using the gateway system from Invitrogen. A set of pDEST plasmids derived from pDEST-CMV-N-EGFP and pDEST-CMV-C-EGFP were designed using synthetic genes purchased from Twist Bioscience to generate a combination of N-terminal and C-terminal fusion protein constructs with prey (the protein of interest fused with an eGFP protein) and decoy (the protein of interest fused with an RFP fluorophore and a microtubule-binding domain (MBD (SEQ ID N°14)) corresponding to the longest isotype of human Tau protein as described above (Boca et al., 2015)). Human p53 and MDM2 full-length cDNA were amplified by PCR using primers containing attB1 and attB2 sites and with a kozak sequence but without a stop codon to generate a "fusion" entry clone in pDONR221 for recombination into the pDEST-CMV-C vector (for C-terminal fusion proteins), or amplified without ATG but with a stop codon to generate a "near" entry clone for recombination into the pDEST-CMV-N vector (for N-terminal fusion proteins).

[0366] The construct with C-terminal fusion of p53 (SEQ ID N°15) as the decoy (p53-FL_C_RFP-MBD) and the construct with C-terminal fusion of MDM2 (SEQ ID N°21) as the prey (MDM2-FL_C_GFP) were used on the MT bench. The p53 mutants were obtained by subcloning synthetic genes purchased from Twist Bioscience to generate mutant constructs: p53F19A (SEQ ID N°17) C_RFP-MBD and p53W23A (SEQ ID N°16) C_RFP-MBD with single mutations; p53F19AW23A (SEQ ID N°18) C_RFP-MBD and p53W23AL26A (SEQ ID N°19) C_RFP-MBD with double mutations; and p53F19AW23AL26A (SEQ ID N°20) C_RFP-MBD with triple mutations. All generated plasmids were purified using the NucleoSpin plasmid DNA purification kit from MACHEREY-NAGEL, and the integrity of the different DNA sequences was verified by Eurofins Genomics sequencing.

[0367] Production of cell plates

[0368] The MT bench was constructed using the human osteosarcoma (U-2 OS) cell line, which served as a reference for cell imaging. U-2 OS cells were cultured at 37°C in a humidified atmosphere with 5% CO2 in Durbeco Modified Eagle Medium (DMEM, Life Technologies) supplemented with 10% FBS (fetal bovine serum, Life Technologies) and 1% penicillin / streptomycin.

[0369] Cell plate production (inoculation, transfection, and processing) is automated using the BRAVO liquid processor from Agilent, equipped with a 96-LT (large tip) tip. U-2 OS cells were seeded at a density of 18,000 cells / well in black 96-well cell carrier ultra (PerkinElmer) plates and incubated for 24 hours in a humidified incubator at 37°C and 5% CO2. For the positive control (untreated conditions), 1.2 μg of human p53-FL (SEQ ID N°15)_C_RFP-MBD and 0.3 μg of human MDM2-FL (SEQ ID N°21)_C_GFP plasmid were used, or for the negative control (non-interacting conditions / treated conditions), 1.2 μg of human p53-F19AW23AL26A (SEQ ID N°20)_C_RFP-MBD and 0.3 μg of human MDM2-FL (SEQ ID N°21)_C_GFP plasmid were used, and co-transfected in optiMEM with 0.5 μg of Lipofectamine 2000 (Invitrogen) for 18 hours.

[0370] For hit detection, cells were treated in culture medium at 37°C for 2 hours with 0.1% DMSO (control wells) and two concentrations (1 μM and 10 μM) of RG7112, Nutlin-3a, and eight other compounds (Rac-7b'', Rac-4a'', Rac-4c', Rac-4d', Rac-8', (-)-8', (+)-8', and Rac-MG-1133) diluted from 50 mM or 10 mM stock solutions. For IC50 detection... 50 Cells were treated at 37°C for 2 hours by adding 0.1% DMSO (control well) and 10 of the same compounds at concentrations ranging from 0.001 μM to 25 μM or 0.01 μM to 100 μM to the culture medium.

[0371] Eighteen hours after transfection, cells were fixed for 10 minutes at -20°C with 100% methanol in ice-cold water for 10 minutes, washed with PBS, and then further fixed for 10 minutes at room temperature with freshly prepared 4% paraformaldehyde (PFA) in PBS. This double fixation (methanol / PFA) was used as the optimal method for revealing microtubule and colocalization events. After washing with PBS, the nuclei were stained with DAPI (0.1 μg / mL) for 5 minutes at room temperature, and the samples were held in PBS for fluorescence microscopy imaging. Imaging was performed using a 40× 1.1NA water immersion objective on an Opera Phenix HCS imaging system from PerkinElmer and with a Harmony microscope. ® The software acquires cell fluorescence signals.

[0372] Image and statistical analysis

[0373] To quantify the level of colocalization between protein decoys and putative protein prey fused with MBD, we developed an analytical workflow adapted from a previously described method (French et al., 2008). Following image acquisition, Harmony was used... ® The software uses sequential building blocks for image segmentation (finding cell nuclei, finding cytoplasm, and selecting populations by thresholding the maximum and average intensities of GFP and RFP, as well as cell area). The two images are then filtered using a SER Ridge filter to highlight microtubule structures in the cytoplasm.

[0374] To quantify the degree of colocalization between fluorophores, we used the Pearson correlation coefficient (PCC) (Benesey et al., 2009). We also used the Mander overlap coefficient (MOC) (Manders et al., 1993) (data not shown). Although the two coefficients are mathematically similar, PCC appears to quantify colocalization more accurately by using deviations from the mean (Adler and Parmryd, 2010). PCC is a confirmed correlation score for measuring colocalization and is calculated as follows:

[0375]

[0376] Where R i and G i It is the pixel intensity value, and R av and G av These are the average values ​​for the red and green channels, respectively. The values ​​range from +1 (perfectly correlated) to -1 (perfectly inversely correlated), where 0 indicates no co-localization. Therefore, the PCC value reflects the co-localization of decoy and prey proteins into the cytoplasm, translating interactions on the microtubule network.

[0377] In the case of hit identification, the raw data (PCC score) are normalized to allow comparisons between experimental plates (Kevorkov and Makarenkov, 2005). This normalization is performed by extrapolating PCC values ​​relative to controls using the mean of the negative control wells and the mean of the positive control wells.

[0378] The Z' factor is correlated with the separation band between the distributions of positive and negative control values. It is calculated using the following formula:

[0379]

[0380] Where σC+ and σC- are the standard deviations of the positive and negative controls, respectively; μC+ and μC- are their mean values ​​(Zhang, 1999; Zhang et al., 2000). When samples are used instead of negative controls, the Z factor is the equivalent of the Z' factor. The quality of HTS assays is classified according to the obtained Z' factor and Z factor values. A value of 0.5 is a generally accepted threshold for high-quality screening.

[0381] The mean ± SD of four independent experiments are presented as a histogram. R statistical software and GraphPadPrism were used. ® The software performs data calculations, statistical analysis, and graphing. In GraphPad Prism... ® The Student's t-test was used for statistical comparison. The following notations were used: *p<0.1, **p<0.01, ***p<0.001, ****p<0.0001.

[0382] Western blotting

[0383] Under a humidified atmosphere of 37°C and 5% CO2, in DMEM (Life Technologies) supplemented with 10% FBS (Life Technologies) and 1% penicillin / streptomycin, at a concentration of 0.5 × 10⁻⁶... 6 The confluence of cells / well was measured in human osteosarcoma (U-2 OS) cultured in 6-well plates. After 24 hours of culture, the compound was added to a final concentration of 10 μM. After 6 hours of treatment, cells were lysed with RIPA buffer (Sigma) supplemented with a protease inhibitor mixture (Sigma). The cells were then analyzed using a BCA protein assay kit (Pierce). ™ After protein dosing, an equal volume of protein lysate (10 μg) was denatured and loaded into the wells of a 12% SDS-PAGE gel. Proteins were separated by electrophoresis (120 V, 45 min), and then trans-blot was used. ® Turbo ™Proteins were transferred to a PVDF membrane (Amersham Hybond-P Membrane, GE Heathcare) using a blotting system (100V, 1 hour 15 minutes). The membrane was blocked for 2 hours at room temperature with 5% skim milk in 0.1% Tween-PBS (T-PBS). The membrane was then probed with the following primary antibodies: mouse monoclonal anti-human p53 (DO-1, #sc-126, Santa Cruz Biotechnology Inc.), rabbit anti-human p21 (C-19, #sc-397, Santa Cruz Biotechnology Inc.), and mouse monoclonal anti-actin (#A5441, Sigma). Horseradish peroxidase (HRP) conjugated goat anti-mouse IgG (#PO447, Dako) and goat anti-rabbit IgG (#AP187P, Millipore) were used as secondary antibodies. The blots were visualized using the ECL Western Blot Kit (Bio-Rad) and then visualized and quantified using the ChemiDoc MP Imaging System (Bio-Rad).

[0384] result:

[0385] High-content screening of p53 / MDM2 interaction using cell-based microtubule scaffold assay

[0386] In order to perform MT bench technology on the HCS system ( Figure 4 (A) We used the well-known PPI, p53-MDM2 (Moll and Petrenko, 2003; Nayak et al., 2018; Zhao et al., 2015), as a model. To analyze the interaction between p53 and MDM2, we transiently expressed both the p53 and MDM2 fusion proteins in U-2OS cells. We were able to transfer this assay from microscopy to the HCS system, allowing for robust quantification of the binding event.

[0387] The decoy protein, full-length p53 (SEQ ID N°15), is fused with RFP (red fluorescent protein) and the MBD (microtubule-binding domain) (SEQ ID N°14) from MAP (microtubule-associated protein) to carry p53 onto microtubules and allow observation of the decoy along the microtubule network in the cell using the red channel. Conversely, the prey protein MDM2 is not carried onto microtubules but is fused only with GFP (green fluorescent protein) to allow detection using the green channel via fluorescence microscopy. When the red and green channels are combined, the two fluorescent signals colocalize along the same microtubules, corresponding to the interaction between decoy p53 and prey MDM2. Figure 4A). Although the native sublocalization of p53 and MDM2 is primarily in the nucleus, this technique's ability to relocalize nucleoproteins in the cytoplasm allows for the detection of p53 / MDM2 interactions along microtubules. As a negative control, a plasmid containing full-length p53 as a decoy (p53-FL (SEQ ID N°15)_C_RFP-MBD) was mutated at three hotspot residues (Phe19, Trp23, and Leu26), eliminating the interaction with MDM2 (Bista et al., 2013). We then observed a loss of colocalization at the microtubules, with MDM2 relocalizing to the nucleus ( Figure 4 B). Interestingly, a single mutation in one of these three hotspot residues is sufficient to eliminate more than 90% of the interactions detected on microtubules. Figure 5 ).

[0388] To provide interaction scores, we developed an image analysis workflow to utilize Harmony from PerkinElmer. ® The software measures the co-location of p53 and MDM2 on the microtube, as described in the Materials and Methods section. Figure 6 To this end, we measured the Pearson correlation coefficient (PCC) in the cytoplasm of each co-transfected cell type (Adler and Parmryd, 2010; Benesey et al., 2009), and then calculated the mean PCC per well to give a statistically reliable score. Figure 6 After detecting the interaction between p53 and MDM2 in a cellular background, we used a well-characterized MDM2 inhibitor, RG7112 (Liu et al., 2019; Vu et al., 2013), to investigate whether small molecule-induced perturbations could be quantified using the MT bench assay. When cells were treated with progressively increasing concentrations of RG7112 ([0.05 μM], [0.5 μM], [5 μM], and [50 μM]), we observed a dose-dependent decrease in the p53 and MDM2 interaction score using the MT bench assay. Figure 4 C). The dose-dependent response of RG7112 is characterized by weak inhibition of the interaction at low concentrations ([0.05 μM]) and strong inhibition of the interaction at high concentrations ([5 μM] and [50 μM]), followed by moderate inhibition of the interaction at moderate concentrations ([0.5 μM]). Figure 4 C). Furthermore, we observed a gradual disappearance of MDM2 on the microtubule network in the green channel, reflecting a gradual loss of the small molecule-induced interaction between p53 and MDM2. Interestingly, using the MT bench assay, the presence of p53 / MDM2 inhibitors not only triggered the loss of interaction but was also associated with the gradual relocation of MDM2 to the nucleus.

[0389] Next, we tested whether the MT benchtop assay could be used to screen PPI inhibitors with HCS. We then measured the widely used Z' factor in a 96-well plate format, a statistical parameter validating the suitability of the assay for high-throughput screening (Zhang, 1999) (see Methods). When using the p53 triple mutant (p53-F19AW23AL26A) as a negative control, the resulting Z' factor was 0.69, higher than the threshold of 0.5 for considering the assay superior (Figure 15A). The system was also performed in a 384-well plate format, with Z' = 0.65 (…). Figure 7 A). Notably, the Z' factor calculated after using G3BP1 protein as a negative control decoy was 0.59. This value is in the same range as the p53 triple mutant that eliminates PPIs, meaning that, in the absence of prior knowledge about hotspot residues that block specific PPIs, we can use another unrelated protein as a negative control. Figure 7 B- Figure 7 C). Furthermore, we measured the Z factor when the negative control corresponded to a sample in which the characterized inhibitor had achieved complete disruption of the interacting complex (Zhang et al., 2000). Here, we measured the minimum colocalization score for 5 μM RG7112 and obtained a Z' factor of 0.67 when this condition was used as a negative control (Fig. 15B). In summary, these data demonstrate that the system p53 / MDM2 setup using the MT bench technique in HCS is robust, reproducible, and sensitive for hit identification.

[0390] discuss:

[0391] Compared to other methods, the direct visualization and quantification of full-length protein-protein interactions in human cells offers several advantages. Notably, we demonstrated that the MT benchtop technique allows for the identification of hotspot residues and small-molecule modulators of PPIs, representing strategically important targets for drug discovery. Furthermore, we were able to measure dose-response effects and extract their cellular viability (IC50). 50 This type of cellular structure-activity relationship (SAR) information can pave the way for the optimization of regulators by using cheminformatics tools.

[0392] A key feature of the MT bench assay is its ability to reduce the number of false hits compared to HTRF, likely due to its high sensitivity and the fact that the interactions involve full-length proteins and are present in human cells. Indeed, based on the screening reported here, the molecule Rac-MG-1133 was not identified as a potential p53 / MDM2 disruptor in HTRF, but it significantly inhibited p53 / MDM2 interactions in cells. Therefore, Rac-MG-1133 can be considered a false negative. The MT bench assay can distinguish the effects of enantiomers from racemic mixtures, as exemplified by the molecule Rac-8. The dextrorotatory enantiomer (+)-8' induced strong inhibition in both assays, while the levorotatory enantiomer (-)-8' was a false positive in the MT bench assay, most likely because this enantiomer is ineffective in cells but still shows inhibition in in vitro HTRF.

[0393] The robustness of the MT bench assay lies in its basis in single-cell analysis. Therefore, the output per well is an average of hundreds of events, resulting in a precise quantitative score of the interaction between the two proteins, and consequently, its perturbation by modulators. In fact, the reproducibility of the MT bench assay has been demonstrated using non-binding proteins as decoys, non-interacting mutants as controls, or similar Z' factors that result in complete disruption of the interaction (Figures 15A-15B and 15B). Figure 7 C).

[0394] In addition to the variety of negative controls available, the simplicity and flexibility of techniques that utilize full-length proteins in cells allow for the screening of protein interactions without prior structural data or knowledge about the interaction interfaces. Another advantage of using full-length proteins is the ability to identify ortho- and allosteric regulators. The MT bench can also be used to confirm interactions between two mating bodies and to identify hotspots in cells, which provides valuable information for discovering PPI inhibitors using virtual screening methods (Guo et al., 2014).

[0395] Here, we demonstrate that the MT scaffold is a reliable cell-based technique for the discovery of PPI inhibitors and that it can provide a solution for the discovery of new chemical scaffolds with improved ADMET properties (Neochoritis et al., 2019b), and contribute to the clinical development of MDM2 / MDMX inhibitors for cancer therapy.

[0396] In summary, we have demonstrated that the MT bench is a robust, quantitative, and sensitive cell-based screening assay. This technology, combined with computer expertise, can significantly streamline early drug discovery for challenging targets such as PPIs in a time-efficient manner. By reducing the risk of false positives and false negatives during the hit identification phase and characterization of the mechanism of action of PPI modulators, the MT bench can facilitate lead optimization of small molecules with properties suitable for preclinical studies.

[0397] Example 2: Targeting Proteins: Protein-Protein Interactions to Identify Regulators and Determine IC50 Numbers in Cells according to .

[0398] Methods and materials:

[0399] Generation of genetic constructs

[0400] Cloning of the genes of interest was performed using the gateway system from Invitrogen. A set of pDEST plasmids derived from pDEST-CMV-N-EGFP and pDEST-CMV-C-EGFP were designed using synthetic genes purchased from Twist Bioscience to generate a combination of N-terminal and C-terminal fusion protein constructs with prey (the protein of interest fused with an eGFP protein) and decoy (the protein of interest fused with an RFP fluorophore and a microtubule-binding domain (MBD) corresponding to the longest isotype of human Tau protein as described above (Boca et al., 2015) (SEQ ID N°14)). Human p53 and MDM2 full-length (SEQ ID N°21) cDNA were amplified by PCR using primers containing attB1 and attB2 sites and with a kozak sequence but without a stop codon to generate a "fusion" entry clone in pDONR221 for recombination into the pDEST-CMV-C vector (for C-terminal fusion proteins), or amplified without ATG but with a stop codon to generate a "near" entry clone for recombination into the pDEST-CMV-N vector (for N-terminal fusion proteins).

[0401] Constructs of the C-terminal fusion with RFP-MBD (p53-FL (SEQ ID N°15)-RFP-MBD (SEQ ID N°14)) using p53 as bait and constructs of the C-terminal fusion with GFP (MDM2-FL-GFP) using MDM2 (SEQ ID N°21) as prey were used on the MT bench. p53 mutants were obtained by subcloning synthetic genes purchased from Twist Bioscience to produce mutant constructs: p53F19A-RFP-MBD and p53W23A-RFP-MBD with single mutations; p53F19AW23A-RFP-MBD and p53W23AL26A-RFP-MBD with double mutations; and p53F19AW23AL26A-RFP-MBD with triple mutations. All generated plasmids were purified using the NucleoSpin plasmid DNA purification kit from MACHEREY-NAGEL, and the integrity of the different DNA sequences was verified by Eurofins Genomics sequencing.

[0402] Production of cell plates

[0403] The MT bench was used with the human osteosarcoma (U-2 OS) cell line. U-2 OS cells were cultured at 37°C in a humidified atmosphere with 5% CO2 in DMEM (Life Technologies) supplemented with 10% FBS (fetal bovine serum) and 1% penicillin / streptomycin.

[0404] Cell plate generation (seeding, transfection, and processing) was automated using the BRAVO liquid processor from Agilent equipped with a 96-LT (large tip) nozzle. U-2 OS cells were seeded at a density of 18,000 cells / well in black 96-well cell carrier ultra (PerkinElmer) plates and incubated for 24 hours in a humidified incubator at 37°C and 5% CO2. Then, the cells were co-transfected for 18 hours with the specified amount of plasmid of interest in optiMEM using 0.5 μg Lipofectamine 2000 (Invitrogen).

[0405] For IC50, cells were treated at 37°C over a 2-hour period by adding 0.1% DMSO (control wells) and 10 concentrations of RG7112 ranging from 0.001 μM to 25 μM to the culture medium.

[0406] Eighteen hours after transfection, cells were fixed for 10 minutes at -20°C with 100% methanol in ice-cold water for 10 minutes, washed with PBS, and then further fixed for 10 minutes at room temperature with freshly prepared 4% paraformaldehyde (PFA) in PBS. This double fixation (methanol / PFA) was used as the optimal method for revealing microtubule and colocalization events. After washing with PBS, the nuclei were stained with DAPI (0.1 µg / mL) for 5 minutes at room temperature, and the samples were held in PBS for fluorescence microscopy imaging. Imaging was performed using a 40× 1.1 NA water immersion objective on an Opera Phenix HCS imaging system from PerkinElmer and with a Harmony microscope. ® The software acquires cell fluorescence signals.

[0407] Image and statistical analysis

[0408] To quantify the enrichment levels of protein prey on protein baits fused with MBD, we developed an analytical workflow adapted from a previously described method (French et al., 2008). Image acquisition and statistical analysis of cellular fluorescence signals were performed using a 40× water immersion objective with a numerical aperture of 1.1 on the Perkin Elmer high-content imaging system Opera Phenix Plus, allowing us to obtain good resolution in confocal mode. Several fields of view were taken for each well, resulting in analysis of thousands of cells per well in 96-well or 384-well plate formats with the strongest statistical significance. The analytical workflow, using SImA, incorporates sequential building blocks for image segmentation, selection of populations of interest, and computation of signal enrichment on microtubules (MTs). ® Software extracts data. Enrichment on identified spots is calculated using the RFP (red fluorescent protein) channel, which corresponds to the signal of decoy proteins forced to localize at the MT due to their fusion with the microtubule-binding domain (MBD). Spots representing MT segments are selected based on their shape and the intensity of the RFP channel signal (corresponding to the presence of decoys on the MT). Another protein is detected using GFP (green fluorescent protein), and these can be brought to the MT due to their potential interaction with the decoys. From SImA ® The software extracts the calculated RFP and GFP intensities on the spots and in the cytoplasm, and then processes them to measure the slope of prey protein enrichment on the MT (mean spot intensity divided by mean cytoplasmic intensity versus mean bait spot intensity).

[0409] Using Spotfire ®Data analysis software was used to perform calculations, statistical analysis, and plotting. The raw data were normalized using min-max normalization (with the positive and negative controls as max and min) to allow for comparisons between experimental plates (Kevorkov and Makarenkov, 2005).

[0410] Results and analysis:

[0411] To perform the MT bench technique on the HCS system, we used the well-known PPI p53-MDM2 (Moll and Petrenko, 2003; Nayak et al., 2018; Zhao et al., 2015) as a model. To analyze the interaction between p53 and MDM2, we transiently expressed both the p53 and MDM2 fusion proteins in U-2OS cells.

[0412] The decoy protein, full-length p53, is fused with RFP (red fluorescent protein) and the MBD (microtubule-binding domain) from MAP (microtubule-associated protein) to carry p53 onto microtubules and allow observation of the decoy along the microtubule network in the cell using the red channel. Conversely, the prey protein MDM2 is not carried onto microtubules but is fused only with GFP (green fluorescent protein) to allow detection using the green channel via fluorescence microscopy. Figure 4 A).

[0413] When the red and green channels merge, the two fluorescent signals co-localize along the same microtubules, corresponding to the interaction between the decoy p53 and the prey MDM2. Although the natural sublocalization of p53 and MDM2 is primarily in the nucleus, this technique's ability to relocalize nucleoproteins in the cytoplasm allows for the detection of p53 / MDM2 interactions along microtubules. A plasmid containing full-length p53 as a decoy (p53-FL-RFP-MBD) mutation was used. When the red and green channels merge, the two fluorescent signals co-localize along the same microtubules, corresponding to the interaction between the decoy p53 and the prey MDM2. Although the natural sublocalization of p53 and MDM2 is primarily in the nucleus, this technique's ability to relocalize nucleoproteins in the cytoplasm allows for the detection of p53 / MDM2 interactions along microtubules. A plasmid containing full-length p53 as a decoy (p53-FL-RFP-MBD) with mutations at three hotspot residues (Phe19, Trp23, and Leu26) eliminated interaction with MDM2 (Bista et al., 2013). We then observed loss of colocalization at the microtubule site, with MDM2 relocalizing to the nucleus. Figure 4 B).

[0414] To quantitatively estimate protein-protein interactions, we developed an image analysis pipeline to measure the enrichment of MDM2 on microtubules via its interaction with p53, as described in the Materials and Methods section, and we tested whether the MT bench assay could be used to screen PPI inhibitors with HCS.

[0415] We first measured the enrichment of MDM2 into microtubules when different p53 mutants were overexpressed. Figure 8 ).

[0416] We evaluated the enrichment level under each condition ( Figure 9 Expression of the p53 mutant resulted in less enrichment on microtubules, closer to the enrichment levels of the negative control. Under conditions expressing the single mutant p53-L26A-RFP-MBD or the triple mutant p53-F19AW32AL26A-RFP-MBD, MDM2 was never detected at microtubules, indicating that these mutants completely eliminated protein-protein interactions. These data demonstrate that the MT benchtop technique allows for the identification of hotspot residue regulators of PPIs.

[0417] After detecting the interaction between p53 and MDM2, we used a well-characterized MDM2 inhibitor, RG7112 (Liu et al., 2019; Vu et al., 2013), to explore whether small molecule-induced perturbations could be quantified using the MT bench assay. We determined the IC50 of this molecule in cells at 10 concentrations using the MT bench assay. The calculated IC50 for RG7112 was 0.074 ± 0.033 µM, consistent with the IC50 reported by the HTRF assay (IC50 = 0.018 µM) (Vu et al., 2013). Some cellular data for RG7112 using the MTT assay have also been reported (IC50 = 0.4 µM) (Vu et al., 2013). (These results demonstrate that the MT bench technique is sensitive enough to detect perturbations in cells, even at the lowest inhibitor concentrations, allowing us to obtain more accurate IC50 data in cells.)

[0418] Example 3: Targeting RNA:protein interactions to identify regulators and determine IC50 data in cells .

[0419] Methods and materials:

[0420] Generation of genetic constructs

[0421] The genes of interest were cloned using the gateway system from Invitrogen. A set of pDEST plasmids derived from pDEST-CMV-N-EGFP and pDEST-CMV-C-EGFP (pDEST-N-MBD-EGFP and pDEST-CMV-C-EGFP-MBD) were designed using synthetic genes purchased from Twist Bioscience to generate combinations of constructs in which the bait protein is fused to EGFP at the N-terminus or C-terminus and then fused to the microtubule-binding domain (MBD) corresponding to the longest isotype of human Tau protein as described above (Boca et al., 2015). Human YB1FL (full-length) cDNA was amplified by PCR using primers containing attB1 and attB2 sites and with a kozak sequence but without a stop codon to generate a "fusion" entry clone in pDONR221 for recombination into the pDEST-C-GFP-MBD vector (for C-terminal fusion proteins), or amplified without ATG but with a stop codon to generate a "near" entry clone for recombination into the pDEST-CMV-N vector (for N-terminal fusion proteins). YB1∆CSD (SEQ ID N°23) cDNA (with amino acids 52-129 corresponding to the CSD domain of YB1 deleted) was synthesized by Eurofins Genomics and then constructed in a similar manner to YB1 FL (SEQ ID N°22).

[0422] All generated plasmids were purified using the NucleoSpin plasmid DNA purification kit from MACHEREY-NAGEL, and the integrity of the different DNA sequences was verified by Eurofins Genomics sequencing.

[0423] Production of cell plates

[0424] MT bench was used with HeLa cells (ATCC CCL-2). HeLa cells were cultured at 37°C in a humidified atmosphere with 5% CO2 in Durbeco Modified Eagle Medium (DMEM, Life Technologies) supplemented with 10% FBS (fetal bovine serum, Life Technologies) and 1% penicillin / streptomycin.

[0425] Cell plate generation (seeding, transfection, and processing) was automated using the BRAVO liquid processor from Agilent equipped with a 96-LT (large tip) nozzle. HeLa cells were seeded at a density of 5000 cells / well in black 384-well CellCarrier Ultra (PerkinElmer) plates and incubated for 24 hours in a humidified incubator at 37°C and 5% CO2. Then, the cells were co-transfected for 18 hours with 0.5 µg of the plasmid of interest in 0.1 µg of Lipofectamine 2000 (Invitrogen) in OptiMEM.

[0426] Cells were then fixed for 10 minutes at -20°C with 100% methanol for 10 minutes, washed with PBS, and then further fixed for 10 minutes at room temperature with freshly prepared 4% PFA in PBS. After fixation, cells were incubated at 37°C for 2 hours with oligo-dT-[Cy3] diluted in SSC 2×, 1 mg / ml yeast tRNA, 0.005% BSA, 10% dextran sulfate, and 25% formamide for RNA visualization. Washing was performed using 4× SSC buffer followed by 2× SSC buffer (0.88% sodium citrate, 1.75% NaCl, pH 7.0). Nuclei were stained with DAPI (0.1 μM) for 5 minutes at room temperature.

[0427] After washing with PBS, the cell nuclei were stained with DAPI (0.1 µg / mL) for 5 minutes at room temperature, and the samples were held in PBS for fluorescence microscopy imaging. The images were taken on a PerkinElmer HCS imaging system, Opera Phenix Plus, using a 40× 1.1 NA water immersion objective and a Harmony microscope. ® The software acquires cell fluorescence signals.

[0428] Image analysis

[0429] To quantify the enrichment levels of mRNA prey on protein baits fused with MBD, we developed an analytical workflow adapted from a previously described method (French et al., 2008). Image acquisition and statistical analysis of cellular fluorescence signals were performed using a 40× water immersion objective with a numerical aperture of 1.1 on the Perkin Elmer high-content imaging system Opera Phenix Plus, allowing for good resolution in confocal mode. Taking 20 fields of view per well resulted in well-by-well analysis of thousands of cells for the strongest statistical significance. The analytical workflow, using SImA, employed sequential building blocks for image segmentation, selection of populations of interest, and computation of signal enrichment on microtubules (MTs). ®Software extracts data. Enrichment on identified spots is calculated using the GFP (green fluorescent protein) channel, which corresponds to the signal of the decoy protein forced to localize at the MT due to its fusion with the microtubule-binding domain (MBD). Spots representing MT segments are selected based on their shape. mRNA is detected using Cy3-labeled poly(dT) and can be carried onto the MT due to their potential interaction with the decoy. From SImA... ® The software extracts the calculated GFP and Cy3 intensities from the spots and cytoplasm, and then processes them to measure the relative enrichment of mRNA on the MT (mean spot intensity divided by mean cytoplasmic intensity (GFP) versus mean bait spot intensity divided by mean cytoplasmic intensity (Cy3)).

[0430] Using Spotfire ® Data analysis software was used to perform calculations, statistical analysis, and plotting. The raw data were normalized using min-max normalization (with the positive and negative controls as max and min) to allow for comparisons between experimental plates (Kevorkov and Makarenkov, 2005).

[0431] Statistical analysis

[0432] To determine whether there is a significant difference in relative enrichment between the conditions, we performed a Student's t-test. This statistical test allows us to compare the means of two groups and determine whether the difference between them is statistically significant. We set the following hypotheses for the t-test:

[0433] Null hypothesis (H0): There is no significant difference in relative enrichment between conditions.

[0434] Alternative hypothesis (H1): There are significant differences in the relative enrichment between the conditions.

[0435] For the t-test to be reliable, certain conditions must be met. First, observations within each group should be independent of each other. In our experiment, we used multiple replicates for each condition, which helped ensure the independence of observations.

[0436] Next, the data should be normally distributed within each group. We used histograms and normal probability plots to check the normality of the data and found that the data is approximately normally distributed.

[0437] We also used the F-test to check for equal variances and found no significant difference in variances between the two groups. This is important because the t-test assumes equal variances, and failure to meet this assumption can affect the reliability of the test.

[0438] Finally, we checked the sample sizes of the two groups and found that they were equal. This is important because the t-test assumes equal sample sizes, and failure to meet this assumption can affect the reliability of the test.

[0439] In summary, we believe that the conditions for a t-test are met in our experiments, and we can trust the results of the test to accurately compare the relative enrichment of the two conditions.

[0440] Based on the results of the t-test, we can infer that the null hypothesis should be rejected. The p-value of this test is less than 0.001, indicating that the relative enrichment difference between the two experiments is statistically significant. This suggests that there is a real difference between the two groups, and not merely random variation.

[0441] We used geometric mean, normalized with a negative control as a reference, to compare the relative enrichment of different constructs.

[0442] To calculate the IC50 curve and values, the original data were normalized using min-max normalization (with the positive and negative controls as max and min, Kevorkov and Makarenkov, 2005).

[0443] Results and analysis:

[0444] While N-terminal and C-terminal fusion tags are useful tools for studying protein-RNA interactions, their specific localization and interaction with RNA can influence the results of these studies. We first examined the effect of GFP fusion protein localization on mRNA enrichment on YB1. Figure 10 ).

[0445] Our results show that YB1 and RNA can still interact regardless of the location of the GFP tag (N-terminus or C-terminus). Figure 11 ).

[0446] However, the enrichment was significantly different, with lower enrichment observed in the N-terminal GFP-tagged constructs. The underlying YB1:RNA interaction mechanisms explaining this difference require further investigation.

[0447] Next, we examined the effect of losing the cold shock domain (CSD) (SEQ ID N°23) of YB1, a conserved RNA-binding domain for drug discovery (El Hage et al., under review). Figure 12 ).

[0448] Our results show that the absence of the RNA-binding domain leads to a significant reduction in RNA binding and an enrichment of N-terminal and C-terminal GFP-tagged proteins. These results indicate that the presence of the RNA-binding domain is essential for this significant enrichment.

[0449] Similar to the PPI system, we attempted to explore whether the MT bench assay could be used to quantify perturbations induced by small molecules. Preliminary NMR and cellular results showed that the C8 compound reduced RNA:YB1 interactions. We determined the IC50 of this molecule in HeLa cells at 19 concentrations using the MT bench assay: 31.3 μM ( Figure 13 ).

[0450] These combined results demonstrate that the MT bench technique is sensitive enough to detect changes in protein / RNA interactions. Combining this technique with modifications / recombinations / mutations of proteins expressed in cells (via gateway vectors) will allow for the characterization and quantification of protein / RNA interactions within cells.

[0451] Sequence List Free Text

[0452]

[0453]

[0454]

[0455]

[0456]

Claims

1. An in vitro method for evaluating the ability of a compound to disrupt or stabilize the interaction between one or more decoys and one or more candidate preys in eukaryotic cells, said in vitro method comprising the following steps: a. Providing eukaryotic cells expressing (i) one or more baits and (ii) one or more candidate prey, wherein the bait comprises a bait portion and a polymerized tubulin-binding portion; b. Detect the interaction by determining the presence of an interaction between one or more decoys and one or more candidate preys in the eukaryotic cell, wherein the decoys bind to tubulin in the eukaryotic cell, thereby causing the one or more candidate preys to be localized along the tubulin; c. Contact the eukaryotic cells with the compound; d. As in step b., determine the occurrence of interactions between the one or more decoys and the one or more candidate preys in the eukaryotic cells, and quantify such interactions; e. Compare the occurrence of the interactions generated in steps b and d; f. Infer whether the compound acts to disrupt or stabilize the interaction between one or more decoys and one or more candidate prey.

2. The in vitro method according to claim 1, wherein the decoy comprises: - A polymeric tubulin-binding moiety containing one or more microtubule-binding domains (MBDs), and - Bait section.

3. The in vitro method according to any one of claims 1 or 2, wherein the decoy comprises a connector (L) region located between the polymeric tubulin binding portion and the decoy portion.

4. The in vitro method according to any one of the preceding claims, wherein the candidate prey comprises a fluorescent protein.

5. The in vitro method according to any one of the preceding claims, wherein determining the presence of the prey in the eukaryotic cells in steps b. and d. is a detection method selected from the group consisting of: antibody binding, nucleic acid hybridization, and fluorescence measurement.

6. The in vitro method according to any one of the preceding claims, wherein the cells in step b. are fixed cells or live cells.

7. The in vitro method according to any one of the preceding claims, wherein the method is repeated with different concentrations of the compound to determine the IC50 of the compound for a test combination of one or more decoys and one or more candidate prey.

8. The in vitro method according to any one of the preceding claims, wherein steps a. and b. are repeated prior to step c. with different carrier constructs of the one or more decoys and the one or more candidate preys to determine the occurrence of interactions between the one or more decoys and the one or more candidate preys in the eukaryotic cells, and to select the appropriate combination for evaluation in the following steps to more closely approximate what occurs in vivo.

9. The in vitro method according to any one of the preceding claims, wherein the one or more candidate prey and / or the one or more decoys are selected from: ribonucleic acid and / or deoxyribonucleic acid and / or protein.

10. The in vitro method according to any one of claims 1 to 9, wherein one or more baits are proteins.

11. The in vitro method according to any one of claims 1 to 10, wherein one or more candidate prey are ribonucleic acid.

12. The in vitro method according to any one of claims 1 to 10, wherein one or more candidate prey are proteins.

13. The in vitro method according to any one of the preceding claims, wherein the one or more decoys and one or more candidate prey are characterized by protein diseases or cancer.

14. A compound identified by an in vitro method according to any one of the preceding claims, said compound being used as a drug.

15. Use of a polymeric tubulin-binding portion fused with one or more decoys as a tool for evaluating the ability of a compound to disrupt or stabilize the interaction between the one or more decoys and one or more preys in eukaryotic cells, wherein the decoys bind to polymeric tubulin in the eukaryotic cells, thereby positioning the one or more preys along the polymeric tubulin, and the one or more decoys and the one or more preys come into contact with the compound, thereby allowing evaluation of the disruption or stabilization of the interaction.

Images